PDA

View Full Version : Top v Bottom or Colour



shaun
30-03-2005, 10:04 PM
Spun off from Q 3 in the Doeberl Cup Arbiting Quiz

As I said in the quiz there are no right and wrong answers and I think that this is most certainly the case here. While Charles Z and Cathy Rogers think that Top V Bottom is part of the system, I have always thought that colour takes preference. Having looked at the rules, both on the FIDE website and in Reubens book, it is my opinion that the rules are so badly worded that we could all be right.
Some observations:
At the start of the pairing process you group players into score groups and then sort them into rating order. You then split them into 2 sub-groups, top half and bottom half. <This indicates that top v bottom is part of the system>
In the rules section A9 refers to transpositions (ie re-ordering within a subgroup) and exchanges (swapping players between sub-groups). However this section does not say when to apply them, just a reference to "sound pairings".
In section B2 (Relative Criteria) part (b) it says that "as many players as possible should recieve their colour preference". It does not mention top v bottom as one of the pairing criteria <This indicates that either colour is more important or that top v bottom is so fundamental it wasn't explicitly stated>
In section C6 the method described is to pair the first player in S1 with the first player in S2, the scond in S1 with the second in S2 etc etc. It then states "if p pairings are obtained in complience with B1 and B2 the pairing of this scoregroup is now considered complete". When the rules were first published the clause was badly written leaving it unclear whether you only had to fulfill B1 and not B2. <To me this makes clause B2(b) an integral part of the pairing process>
If p pairings are not obtained you firstly transpose players within S2, repeating step C6 as you go, and if you exhaust all the pairings without succes, you exchange players between S1 and S2 and repeat C6 and C7. <And I think this is the root cause of the confusion. While a computer can mindlessly carry out these steps, for an arbiter to carry out this process exactly as written would take quite some time. Instead humans tend to take shortcuts which may result in a different pairing than the pairing system would arrive at>

Now my reading of the rules would indicate colour is more important than top v bottom but I think there is enough wiggle room in the wording for the opposite conclusion to be also drawn.

But while thinking about this another question occured to me: Under what circumstances would you need to pair within a sub-group? If all of S1 had the same colour preference and all of S2 had the opposite preference then their wouldn't be a problem. Even if both sub-groups wanted the same colour you could still pair top v bottom, with the bottom group simply getting screwed on colour. My guess is that this problem only occurs if there are some topvbottom pairings that cannot occur because the players have already played or the colour balance would end up +3 or -3.

One final observation. On the FIDE web site is the pairing system that preceded the Dutch system. It is listed as a FIDE approved pairing system, so I assume that could be used instead (and indeed it is the system I tend to use for manual pairings). But be warned. In the description is an example of swapping players around to find the best pairings and the example pairs players from the same sub-group together.

Rhubarb
01-04-2005, 01:03 PM
Shaun, if you get a chance can you post the original SP pairings for the last round and explain why they were overruled.

Edit: Actually it'll probably be too much trouble. I don't suppose you or Cathy or Ian can remember the original pairings down to board 4.

Kevin Bonham
01-04-2005, 01:58 PM
One of the problems with the rules as written is the lack of instruction on how to deal with conflicts between section B and other sections. This also comes up when dealing with double-downfloat situations (a known Swiss Perfect issue).

I take it that section B tells you the overarching principles and the other sections including A and C fill in the fine details of how to apply them. This means that section B overrides sections A and C in any case of apparent conflict between them. Actually there is really no conflict if you take this interpretation, only if you take the interpretation that B doesn't override the others.

B clearly says that the colour differences shall be as small as possible so I have always taken that to mean that you exchange between top and bottom halves as a last resort to improve colour allocation. Whether this is how the rules should be written is another question:

Towards the end of a tournament a given score group usually includes both higher-rated players playing badly and lower-rated players playing above themselves. If everyone in a group is playing about the same strength then it makes more sense to deliver colour parity than to worry about seedings which may be somewhat rubbery as an indicator of form on the weekend in question.

I have some degree of bias on this one because I do believe that I have good and bad weekends and that my results vary more from weekend to weekend than sheer chance would indicate (meaning that my score after 6 rounds may be a better predictor of my next result than my rating) - although I haven't tested this view statistically. However I also have no reason to personally care about colour allocation because I score the same with either colour.

pax
01-04-2005, 02:06 PM
Shaun, if you get a chance can you post the original SP pairings for the last round and explain why they were overruled.

Was it a manual pairing, or invocation of "last round special rules"?

Rhubarb
01-04-2005, 02:41 PM
Was it a manual pairing, or invocation of "last round special rules"?I believe it was a manual override.

shaun
01-04-2005, 02:47 PM
To answer both Pax and Greg ... (NB I don't have the SP files with me as I type this).
It was changed because of the "last round special rules", which the version of Swiss Perfect I was using did not follow. It had avoided the obvious upfloat (Lane up to Rogers) because Lane had floated up 2 rounds before (IIRC). Now at the start of the tournament we had anounced we were dropping that restriction for all rounds. However even if we hadn't made that anouncement, the pairing was still wrong because under the last round rules that restriction is dropped for players on 50% or over. So in this case it was a bug in the program. And again if I recall correctly it had Bjelobrk floating up, Smerdon v Lane and Canfell v Rej.
(Now corrected to Canfell v Lane, Smerdon v Rej)

pax
01-04-2005, 03:05 PM
And again if I recall correctly it had Bjelobrk floating up, Smerdon v Lane and Canfell v Rej.

Blimey, keg, are you sure you wanted to know that? That's a slap in the face from the devil of the draw.

Rhubarb
01-04-2005, 03:15 PM
Blimey, keg, are you sure you wanted to know that? That's a slap in the face from the devil of the draw.Well yes but Tomek had just beaten Dazza. I'd be pretty confident with White though.

If SP chooses to upfloat Bjelobrk, why wouldn't it make boards 3 and 4 Canfell-Lane, Smerdon-Rej? Doesn't that solve top half v bottom half as well as colours?

antichrist
01-04-2005, 05:09 PM
...
Random events have been conspiring against me lately to give me a preponderance of Blacks, especially in important games. There was the zonal, the Oz masters, the NSWCA October weekender (where SP wasn't overruled when it would have favoured me for once - there's a whole thread on this here) and now I have 4 out of 6 Blacks in the C of S after preparing with White in round 6, thinking it would be virtually impossible to be Black against Ayvazyan since I was half a point ahead (I subsequently found out that if two players have 3 Bs and 2 Ws, sequence is given a higher priority than rank)....


If you had hung around long enough at end of Round 5 C of S would you have known your draw and colour for Round 6 as is computer generated or is it 2 rounds in one day?

In one of the small comps I put on just to hurry and simplify things I declared that all colours will be decided by "hands behind back" prior to play, I got away with it for a few rounds but then Fred F jacked up. Is it too crude a manner? I thought it was terrific but then I would.

shaun
01-04-2005, 06:32 PM
Well yes but Tomek had just beaten Dazza. I'd be pretty confident with White though.

If SP chooses to upfloat Bjelobrk, why wouldn't it make boards 3 and 4 Canfell-Lane, Smerdon-Rej? Doesn't that solve top half v bottom half as well as colours?

Having just arrived home and repaired Rd7 (without the intervention) the pairings you suggest are in fact the ones it gave.

Bill Gletsos
01-04-2005, 07:19 PM
To answer both Pax and Greg ... (NB I don't have the SP files with me as I type this).
It was changed because of the "last round special rules", which the version of Swiss Perfect I was using did not follow.
Do you mean that when you asked SP to pair round 7 that even though it asked you the question Do you want to apply special last round pairing rules? (Press F1 for Help) it did not actually apply them, or did it not ask you that question.

BTW for the benefit of others not familiar with SP, SP will only ask that question if the Tournament setup screen has the number of rounds set. If you leave it at the default of 11 rounds it wont ask until you pair round 11.

Rhubarb
01-04-2005, 10:47 PM
If you had hung around long enough at end of Round 5 C of S would you have known your draw and colour for Round 6 as is computer generated or is it 2 rounds in one day?Rounds 5 & 6 were on the same day. During the week I'd worked out that if I won Round 5 I would almost certainly play Ayvazyan. As I would be half a point ahead of him and neither of us were going to cop 3 Blacks in a row I thought that rank would be the next priority and I would get my due colour, so I duly prepared with White. My fault for not knowing SP's rules as sequence comes next in this scenario. From memory I had BWBWB while Ayvazyan had WBBWB so he got the White.


In one of the small comps I put on just to hurry and simplify things I declared that all colours will be decided by "hands behind back" prior to play, I got away with it for a few rounds but then Fred F jacked up. Is it too crude a manner? I thought it was terrific but then I would.Why stop there? Make the players draw out of a hat to see which piece to move.

BBlooksee
01-04-2005, 11:06 PM
i never worry what color i got . it makes no diff . anyways wheres the proof that at a tourni like ours white wins more than black . in the data bases it does but does it happen too in doeberl .

Duff McKagan
01-04-2005, 11:51 PM
I know it is good to know that our arbiters are looking after our best interests. Now can Shaun or someone actually crystallise the whole problem with SP and tell us in two paragraphs, 1. what should SP do and 2. what doesn't SP do that makes over riding it a necessity.

I suppose that a third question might arise, How is the pairing process supposed to be transparent when arbiters are fiddling with them without explanations? At least if SP is used and never over ridden an unhappily paired player can be told to argue with the computer.

Rhubarb
01-04-2005, 11:54 PM
i never worry what color i got . it makes no diff . anyways wheres the proof that at a tourni like ours white wins more than black . in the data bases it does but does it happen too in doeberl .Sure, I imagine for a lot of people it doesn't make much difference. But for someone like myself who likes the initiative, I'd much prefer to wind up with White in the last round and try and take someone out.

And spare a thought for Solo. He's quite obviously stronger with White, imo, but he copped two Blacks to finish as well.

Garvinator
02-04-2005, 01:20 AM
I know it is good to know that our arbiters are looking after our best interests. Now can Shaun or someone actually crystallise the whole problem with SP and tell us in two paragraphs, 1. what should SP do and 2. what doesn't SP do that makes over riding it a necessity.
The easiest solution is for fide to clarify once and for all about the pairings rules.

The problem is not so much with sp itself, it is that the dutch pairing rules are not crystal clear. In one section they say that rating counts most, then in others it is colours that matter. Therefore depending on which particular paragraphs an arbiter believes is most important, can influence how players are paired.

Also I get the feeling that sp is not consistent either in how it operates (gut feeling, no real basis but others might have the same feeling).

Oepty
02-04-2005, 02:25 PM
I know it is good to know that our arbiters are looking after our best interests. Now can Shaun or someone actually crystallise the whole problem with SP and tell us in two paragraphs, 1. what should SP do and 2. what doesn't SP do that makes over riding it a necessity.

I suppose that a third question might arise, How is the pairing process supposed to be transparent when arbiters are fiddling with them without explanations? At least if SP is used and never over ridden an unhappily paired player can be told to argue with the computer.

It is my understanding that the rules are not clear enough to make it absolutely clear how pairing are to be made in every situation. This means there are various, well at least 2, ways of interpreting the rules. SP has been written to follow one interpretation and some of Australia's arbiters like Charles Z and it appears also Cathy Rogers use a different interpreation. This leaves the question as to whether to override the pairings that SP makes in a tournament. It is done in some tournaments not other. It appears kegless has recently been on the bad end of this recently which is a pity.

Now as to what the difference is. I am not entirely sure, but I believe it is whether you should always pair the top half of a score group against the bottom half of the score group, even when it means the less players in the score group receive their colour preferences. I think that Charles Z from working with him in Adelaide for the Australian Champs believes that this should be done. Others, which seems to include Shaun, seem to think otherwise.

Now I would not be surprised if I have made a total mess of the last paragraph. If I have then I am sorry.

Scott

Duff McKagan
02-04-2005, 04:15 PM
I am not entirely sure, but I believe it is whether you should always pair the top half of a score group against the bottom half of the score group, even when it means the less players in the score group receive their colour preferences. I think that Charles Z from working with him in Adelaide for the Australian Champs believes that this should be done. Others, which seems to include Shaun, seem to think otherwise.

Scotty please not, thanks guy for that explanation. You take a risk when you answer anything here :)

What is worrying is that a player like kegless who is always alert to changes in pairing ;) could ask at the beginning of a tournament, what will the arbiters do in the last round. If it is good enough to advertise entry fee, prizes, rounds etc. what about a statement to the effect that colour over-rides group of vica verca. It is actually about the important matters of transparency and consistancy and percieved fairness.

The ACF probably needs to step in and say what the pairing standard will be.

Cheers

pax
02-04-2005, 04:28 PM
The ACF probably needs to step in and say what the pairing standard will be.


The ACF has no jurisdiction over such matters. It is entirely in the hands of organisers to choose which pairing method to use.

If they were to recommend anything, I suspect it would be SP with standard (automatic) last round rules invoked (since nearly everyone is using SP now).

shaun
02-04-2005, 04:32 PM
Scotty please not, thanks guy for that explanation. You take a risk when you answer anything here :)

What is worrying is that a player like kegless who is always alert to changes in pairing ;) could ask at the beginning of a tournament, what will the arbiters do in the last round. If it is good enough to advertise entry fee, prizes, rounds etc. what about a statement to the effect that colour over-rides group of vica verca. It is actually about the important matters of transparency and consistancy and percieved fairness.

The ACF probably needs to step in and say what the pairing standard will be.

Cheers

Such a good idea that we did this at the start of this years Doeberl Cup.

shaun
02-04-2005, 04:51 PM
Do you mean that when you asked SP to pair round 7 that even though it asked you the question Do you want to apply special last round pairing rules? (Press F1 for Help) it did not actually apply them, or did it not ask you that question.

BTW for the benefit of others not familiar with SP, SP will only ask that question if the Tournament setup screen has the number of rounds set. If you leave it at the default of 11 rounds it wont ask until you pair round 11.

Yes. We replied "YES" to "apply special last round rules", but it still floated the wrong player based on previous float histories. This may be because of the version of SP I was running "Swiss Perfect 97 Release Build #0373a" which may have predated the latest version of the pairing rules, but even the rules published in Ruebens "Orange" 1997 Rule Book say ignore float histories in the last round for players on or above 50%

Oepty
02-04-2005, 05:32 PM
Scotty please not, thanks guy for that explanation. You take a risk when you answer anything here :)

What is worrying is that a player like kegless who is always alert to changes in pairing ;) could ask at the beginning of a tournament, what will the arbiters do in the last round. If it is good enough to advertise entry fee, prizes, rounds etc. what about a statement to the effect that colour over-rides group of vica verca. It is actually about the important matters of transparency and consistancy and percieved fairness.

The ACF probably needs to step in and say what the pairing standard will be.

Cheers

To fair to kegless, I think he was just frustrated that in his last few tournaments, whenever the arbiter has gone with SP it has meant he has more blacks than whites and whenever the arbiter has chosen to intervene it has meant more he has had more blacks than whites. A bit understandable, and he just let this frustration out a little bit at Doeberl without blaming anybody. Whether he knew which way things were going to go before the start at Doeberl is something I don't know.

Scott

Garvinator
02-04-2005, 05:34 PM
Without really wanting to speak for kegless, I am sure he is aware the arbiters will over ride sp and re pair players. Kegless has had Charles Z as an arbiter many times.

arosar
02-04-2005, 06:36 PM
Quick question: aren't there other pairing programs around? I seem to recall Gijssen recently trialling one. What programs do they use in other countries?

AR

antichrist
02-04-2005, 06:43 PM
The ACF has no jurisdiction over such matters. It is entirely in the hands of organisers to choose which pairing method to use.

If they were to recommend anything, I suspect it would be SP with standard (automatic) last round rules invoked (since nearly everyone is using SP now).

Which would not interfere with 100% manual pairings of course?

Garvinator
02-04-2005, 08:32 PM
Quick question: aren't there other pairing programs around? I seem to recall Gijssen recently trialling one. What programs do they use in other countries?

AR
yes there are other pairing programs. The main one I think is swiss sys that I have seen used in Australia.

I am not sure though what would happen for sending a tournament for rating if the tournament was run using swiss sys.

pax
02-04-2005, 09:14 PM
I am not sure though what would happen for sending a tournament for rating if the tournament was run using swiss sys.

Simple. The organiser or the ratings officer would need to transcribe the results into SP. If it happened frequently, someone could easily write a conversion tool.

Denis_Jessop
02-04-2005, 09:29 PM
The ACF has no jurisdiction over such matters. It is entirely in the hands of organisers to choose which pairing method to use.

If they were to recommend anything, I suspect it would be SP with standard (automatic) last round rules invoked (since nearly everyone is using SP now).

The ACF has in effect endorsed SP several years ago by buying a licence for it that enables chess clubs, State Associations etc to use it for free. Moreover the principles re ratings require the results to be submitted via SP files (doing otherwise incurs an extra fee).

But I agree that that endorsement is not to the exclusion of other Computer progams or manual pairing or even a FIDE-recognised Swiss system other than the Dutch (or probably even a Swiss or other pairing system not recognised by FIDE).

DJ

Denis_Jessop
02-04-2005, 09:47 PM
Quick question: aren't there other pairing programs around? I seem to recall Gijssen recently trialling one. What programs do they use in other countries?

AR

There are several I think. Some time ago Geurt Gijssen was very enthusiastic about Swiss Master 5 written for the Royal Dutch Chess Federation. I have tried it briefly with interesting results. In the current Canberra Chess Club Championship, it paired Rd1 in the same way as SP (one would hope so!!). In Rd 2 it made the same pairings as SP for all the winners in Rd1 but quite different pairings from SP for those on 0. (I assume it also follows the Dutch System Swiss Rules :rolleyes: ) I haven't yet gone into why that happened. But I must say that SP is a far easier program to use.

Recently, in his March Chess Cafe Column, Geurt wrote about the Aeroflot Open for which he was Chief Arbiter and says that there they used another (recently developed) Dutch program called Swiss Under Windows 5.1 in which there was a bug that had hurriedly to be fixed. After that all apparently went well.

SwissSys, mentioned by Garvin, is another but I have never used it and have the impression that it is a bit old hat (but I may be wrong).

DJ

PS A couple of years ago FIDE put out its own Swiss pairing program but I don't know what happened to it.

Bill Gletsos
03-04-2005, 12:25 AM
Yes. We replied "YES" to "apply special last round rules", but it still floated the wrong player based on previous float histories. This may be because of the version of SP I was running "Swiss Perfect 97 Release Build #0373a" which may have predated the latest version of the pairing rules, but even the rules published in Ruebens "Orange" 1997 Rule Book say ignore float histories in the last round for players on or above 50%I think Build #0415b (the latest on the SP website) gets it correct. Once someone sends me the SP files for Doeberl I'll check out what that version generates.

Garvinator
03-04-2005, 12:35 AM
SwissSys, mentioned by Garvin, is another but I have never used it and have the impression that it is a bit old hat (but I may be wrong).
i have mucked around with swisssys, but have not bought the full version, so I dont know of its complete features or tried to run a full trial on it. It looks ok thought, but until i had run a few tournaments on it and tried out different features, cant really tell.

I have heard on the grapevine that with swisssys that you can specify how much of a rating group you want the program to pair players with etc.

ursogr8
03-04-2005, 08:21 AM
I know it is good to know that our arbiters are looking after our best interests. Now can Shaun or someone actually crystallise the whole problem with SP and tell us in two paragraphs, 1. what should SP do and 2. what doesn't SP do that makes over riding it a necessity.
The easiest solution is for fide to clarify once and for all about the pairings rules.

The problem is not so much with sp itself, it is that the dutch pairing rules are not crystal clear. In one section they say that rating counts most, then in others it is colours that matter. Therefore depending on which particular paragraphs an arbiter believes is most important, can influence how players are paired.

Also I get the feeling that sp is not consistent either in how it operates (gut feeling, no real basis but others might have the same feeling).

hi gg''

But I thought the original poster was asking basically
> if SP doesn't get it quite right
>> and manual intervention doesn't get it quite right
why don't we do away with manual intervention?
The benefits of 'hands off' are reducing controversy and delay. What are the benefits of 'hands on'?

Qn directed to you because you were a reliable observer at Mt B.

regards
starter

antichrist
03-04-2005, 10:57 AM
Rounds 5 & 6 were on the same day. During the week I'd worked out that if I won Round 5 I would almost certainly play Ayvazyan. As I would be half a point ahead of him and neither of us were going to cop 3 Blacks in a row I thought that rank would be the next priority and I would get my due colour, so I duly prepared with White. My fault for not knowing SP's rules as sequence comes next in this scenario. From memory I had BWBWB while Ayvazyan had WBBWB so he got the White.

Why stop there? Make the players draw out of a hat to see which piece to

move.

You (and myself) like the perceived advantage of White, "hands behind back" takes away the certainity of such an advantage and in that manner is more fair. One can't sort of gang up in preparation with certainity.

Denis_Jessop
03-04-2005, 06:42 PM
The easiest solution is for fide to clarify once and for all about the pairings rules.

The problem is not so much with sp itself, it is that the dutch pairing rules are not crystal clear. In one section they say that rating counts most, then in others it is colours that matter. Therefore depending on which particular paragraphs an arbiter believes is most important, can influence how players are paired.

Also I get the feeling that sp is not consistent either in how it operates (gut feeling, no real basis but others might have the same feeling).

I think that SP is quite consistent in the way it operates but that it consistently applies the Dutch Swiss Rules in certain circumstances that have become well-known in a way that is not the way most arbiters would interpret them (see more below).

The problem with using a computer pairing program is that first one must know what rules it is using - the FIDE Dutch System is the commonest - and then whether the program is applying them properly. To decide the latter matter one must be able to interpret the rules being used. As Garvin says, the FIDE Dutch rules are far from clear and even I, as a lawyer with many years experience in statutory interpretation, (essentially the same thing) have real trouble with parts of them.

It would be good if FIDE could "clarify once and for all about the pairings rules" but I think that, if that were possible, it would already have been done.

You will recall the preface to the Chess Laws saying that they "cannot cover all possible situations". With the Laws it is left to arbiters to apply the laws correctly at their discretion in light of their assumed competence etc.

In the case of Swiss pairing rules the need for precision is even greater but I think that the same proviso as is in the Preface to the Chess Laws applies. This means that it may well be impossible to draft Swiss Rules that cover every case. Equally, as far as I know, you can't successfully program the exercise of discretions into a computer.

This means that arbiters have the final say and the pairing program is only a guide, albeit an excellent one if the program is well written. I believe that this principle is well understood by most arbiters, hence the actions that Charles Z and others take to vary SP pairings from time to time.

SP has some well-known situations in which it makes pairings that most arbiters would not make and these are usually the cases in which arbiters override it. I have done so not long ago where the pairings looked extremely odd and a better set of pairings was possible without infringing any of the various bits of the Dutch Rules. SP seems especially vulnerable in this regard in the last one or two rounds - that is, it is not only the "special last round rules" that have a bearing on the matter.

DJ

Duff McKagan
05-04-2005, 10:00 AM
It would be good if FIDE could "clarify once and for all about the pairings rules" but I think that, if that were possible, it would already have been done.

Denis, you have plenty of faith in FIDE which is probably misplaced.



You will recall the preface to the Chess Laws saying that they "cannot cover all possible situations". With the Laws it is left to arbiters to apply the laws correctly at their discretion in light of their assumed competence etc.

Discretion is important I agree, but pairing is such an easy chore. The rules for pairing are only a set of either/or instructions. If there is are two conflicting instructions in the rules there has to be a consistent ruling as to which instruction is to be followed. The ACF is the only thing that can do that at the moment in Australia.

As I say discretion is OK for a "deliberate touch" ruling but pairing should be made black and white.


it may well be impossible to draft Swiss Rules that cover every case. Equally, as far as I know, you can't successfully program the exercise of discretions into a computer.

Pairing is not rocket science.


This means that arbiters have the final say and the pairing program is only a guide

Not acceptable, Denis. The ACF should not allow a situation to continue where arbiters can make arbitery decisions were the pair process can be made perfectly transparent and repeatable by a ruling from the ACF stating the colour takes preference over group.

Final round pairings are extremely important because of the white players' advantage. If the arbiters make ad hoc discretionary reversal of colours the whole process starts to lose its legitimacy. The ACF has a duty to fix this problem.

Cheers

Ian Rout
05-04-2005, 10:22 AM
Final round pairings are extremely important because of the white players' advantage. If the arbiters make ad hoc discretionary reversal of colours the whole process starts to lose its legitimacy. The ACF has a duty to fix this problem.

Cheers
I don't think there's any suggestion that arbiters make arbitrary allocation of colour. The rules for who gets what colour when two players are paired are fairly clear-cut, if mildly complex on occasions, as far as I can see. The issue under discussion is more how the players are paired at all.

Duff McKagan
05-04-2005, 10:31 AM
I don't think there's any suggestion that arbiters make arbitrary allocation of colour. The rules for who gets what colour when two players are paired are fairly clear-cut, if mildly complex on occasions, as far as I can see. The issue under discussion is more how the players are paired at all.

Sorry Ian, I should have given more of an explanation in what I said. :( The reversal of colour can be a result of re-pairing.

Ian Rout
05-04-2005, 10:31 AM
PS: Why do people say something is "not rocket science" to mean it's not difficult, whereas rocket science is? Given that rockets have been routinely launched for years, and even took humans to the moon and back as long ago as the 1960s, it would seem that rocket science is at best in the mid-range of difficulty.

Duff McKagan
05-04-2005, 10:43 AM
PS: Why do people say something is "not rocket science" to mean it's not difficult, whereas rocket science is? Given that rockets have been routinely launched for years, and even took humans to the moon and back as long ago as the 1960s, it would seem that rocket science is at best in the mid-range of difficulty. :lol: Perhaps its because its an expression that came from the 1960's and lives on in the lexicon, like every scandal is a something-gate.

Rincewind
05-04-2005, 11:27 AM
:lol: Perhaps its because its an expression that came from the 1960's and lives on in the lexicon, like every scandal is a something-gate.

Actually Watergate was in 1972, but I think you're right. ;)

Garvinator
05-04-2005, 11:28 AM
:lol: Perhaps its because its an expression that came from the 1960's and lives on in the lexicon, like every scandal is a something-gate.
i was going to call this pairinggate ;) :lol: :whistle:

ursogr8
05-04-2005, 11:50 AM
i was going to call this pairinggate ;) :lol: :whistle:
hi gg''

Could you have a look at post #32.

starter

Garvinator
05-04-2005, 12:01 PM
hi gg''

But I thought the original poster was asking basically
> if SP doesn't get it quite right
>> and manual intervention doesn't get it quite right
why don't we do away with manual intervention?
The benefits of 'hands off' are reducing controversy and delay. What are the benefits of 'hands on'?

Qn directed to you because you were a reliable observer at Mt B.

regards
starter
i am not sure how to answer this, probably why I didnt answer it the first time. I am someone who has no hesitation in repairing, but I am a little lost about why a repairing was actually done, but I dont want to do any questioning of the Doeberl organisers and arbiters.

ursogr8
05-04-2005, 12:29 PM
i am not sure how to answer this, probably why I didnt answer it the first time. I am someone who has no hesitation in repairing, but I am a little lost about why a repairing was actually done, but I dont want to do any questioning of the Doeberl organisers and arbiters.

Can you answer it in the context of Mt B. then?

starter

Garvinator
05-04-2005, 01:06 PM
Can you answer it in the context of Mt B. then?

starter
what do you mean, what context? I guess I am not on your train of thought yet, can you please give me a direct question to answer please. Sorry to be difficult. I would just like a direct question.

I will even wait around for your question so I can answer it promptly ;) if i can answer it that is.

pax
05-04-2005, 01:22 PM
Not acceptable, Denis. The ACF should not allow a situation to continue where arbiters can make arbitery decisions were the pair process can be made perfectly transparent and repeatable by a ruling from the ACF stating the colour takes preference over group.

Final round pairings are extremely important because of the white players' advantage. If the arbiters make ad hoc discretionary reversal of colours the whole process starts to lose its legitimacy. The ACF has a duty to fix this problem.

Why do you put this down to the ACF? The ACF has absolutely no power over what pairing system organisers choose to use.

The ACF can recommend a pairing system (and, indeed they do), but they have no penalties to apply if organisers choose to ignore them or pair at random.

They cannot (for example) refuse to rate such tournaments as long as the conditions of the games themselves are deemed appropriate.

Oepty
05-04-2005, 02:34 PM
I don't see why the ACF would want to take a side in the debate. All taking a side is going to do is to annoy the other side of the debate. I don't think it is something worth the risk of pushings some of arbiters out of arbiting.
There is also the question as to whether the ACF would actually be able to come to an agree on the issue.

Also I am trying to understand the correct way to do pairings, difficult I know. If I put up some pairing examples would people be prepared to say how they would do the pairings, and more importantly WHY?

Scott

Garvinator
05-04-2005, 02:55 PM
Also I am trying to understand the correct way to do pairings, difficult I know. If I put up some pairing examples would people be prepared to say how they would do the pairings, and more importantly WHY?

Scott
i would be willing to do this, but i am not sure what it is going to acheive, especially as those who are on here are in the colour camp and those who are in the top v bottom are not on here.

Therefore the answers you will be getting are biased in favour of the colour group.

I think I am the only one on here who is in the top half v bottom half group.

1min_grandmaster
05-04-2005, 06:22 PM
Here is a Q that I don't think has been asked, and if answered, may help a lot in deciding whether to prioritise top vs botom or colour. "Between [top vs bottom] and [colour], which is more important *and why*?"

This is my answer, which may not necessarily be the best, but it is the logic I use to justify my decision.

I think colours are more important, because there is never a restriction between 2 players not being able to meet because of ratings. However, there is a possibility of 2 players not being able to meet because they have both had 2 of the same colour in a row (e.g. Kasparov and Fischer have both had black in last 2 rounds). Hence I think it makes more sense to try to 'sort out' their colours ASAP rather than possibly having this problem in a later round.

What do others think? I would be happy to hear what others have to say, perhaps someone has a better idea.

shaun
05-04-2005, 07:27 PM
Here is a Q that I don't think has been asked, and if answered, may help a lot in deciding whether to prioritise top vs botom or colour. "Between [top vs bottom] and [colour], which is more important *and why*?"

This is my answer, which may not necessarily be the best, but it is the logic I use to justify my decision.

I think colours are more important, because there is never a restriction between 2 players not being able to meet because of ratings. However, there is a possibility of 2 players not being able to meet because they have both had 2 of the same colour in a row (e.g. Kasparov and Fischer have both had black in last 2 rounds). Hence I think it makes more sense to try to 'sort out' their colours ASAP rather than possibly having this problem in a later round.

What do others think? I would be happy to hear what others have to say, perhaps someone has a better idea.

There are various stories about why the Dutch System operates the way it does, but the one I heard a number of years ago is:
When looking at a new pairing system the designers consulted a number of leading GM's about what they wanted in a pairing system. The majority wanted colour to have priority as at the level they played having White or Black was quite significant, especially in the monster swiss' which began to attract large, GM heavy fields where the difference in strength between the players wasn't that great.

But when I mentioned this story to Cathy Rogers she said that this story did not seem credible. Interestingly she also said that she liked the Dutch pairing system because it *didn't* give preference to colour over seeding. As Cathy lives in The Netherlands a lot of the time and I would assume that the Dutch system would be used exclusively there, it appears that even there the implmentation of the rules favours seeding over colour. Hopefully Cathy can shed some light on this.

As for avoiding barred pairings in later rounds, evening of colours is a sensible precaution. When I started out as a DOP I used the USCF rule book and had a rough and ready pairing system which paired top v bottom in seeding order, with no transpostions for colour. The higher ranked player got the colour they were due while their opponent just had to make do. Consequently by about Rd 5 of an event there were lots of players with 2 whites or 2 blacks in a row, making later pairings more difficult than they should have been.

Bill Gletsos
05-04-2005, 08:03 PM
It is my understanding that the program called Petunia (it is now called Swiss Master) is officially endorsed by the FIDE Swiss Pairings Committee.

Stewart Reubens notes on page 34 of the Chess Organiser's Handbook Second Edition that "I have never heard comments about the Petunia Program for the Dutch System giving the wrong pairings".

An intersting point he makes on the same page is "They are all based on the premise that throughout the tournament all players on all score groups are paired top half v bpottom half in rating order after taking into account colour equalisation and then colour alternation." (the bolding of after is mine)

Petunia/Swiss Master is also the program the Geurt swears by.

It would be interesting to compare the draw done for the Doeberl by SP and by Swiss Master 5.1.

N.B. On the FIDE web site under the section on Swiss rules, Regualtions for swiss tournaments, Miscellaneous it notes "The FIDE Swiss Rules pair the players in an objective and impartial way, and different arbiters following the pairing rules should arrive at identical pairings. "

1min_grandmaster
05-04-2005, 08:34 PM
Stewart Reubens notes ... "They are all based on the premise that throughout the tournament all players on all score groups are paired top half v bpottom half in rating order after taking into account colour equalisation and then colour alternation."


This makes complete sense to me. Colour equalisation is most important because it is important to avoid having players with a big imbalance by the end of the tournament. Next comes colour alternation because if you disregard it you may get barred pairings later on. Ratings are of least consideration because the problem of a large difference in ratings between 2 opponents is minimal.



"The FIDE Swiss Rules pair the players in an objective and impartial way, and different arbiters following the pairing rules should arrive at identical pairings. "


This also makes complete sense to me. Now I wish that all arbiters can read this and agree with it! No more arguments!

Garvinator
05-04-2005, 08:40 PM
what i am finding disappointing at the moment is that we have one ia who reads on here regularly and hasnt commented and another ia who can post on here when its necessary and we havent heard anything from either of them.

shaun
05-04-2005, 08:43 PM
An intersting point he makes on the same page is "They are all based on the premise that throughout the tournament all players on all score groups are paired top half v bpottom half in rating order after taking into account colour equalisation and then colour alternation." (the bolding of after is mine)


So I guess that ends the debate then. Colour has preference over seeding. Charles Z isn't going to be happy.

PS Page 34 also contains the quote "The Australian, British, and USCF all have their own variations". I'm not sure Australia ever had a distinct variation (as opposed to just applying the rules differently), unlike NZ, who do have their own variation (reverse pairings in the middle rounds).

Kaitlin
05-04-2005, 08:45 PM
Isn't like tossing a coin at the start of the game like even :ponder:

arosar
05-04-2005, 08:46 PM
I always thought seeding had preference and colours follow. Anyway, as a player the only bit of pairing rule I ever cared about was that I didn't get 3 blacks in a row!

AR

Garvinator
05-04-2005, 09:40 PM
So I guess that ends the debate then. Colour has preference over seeding. Charles Z isn't going to be happy.

PS Page 34 also contains the quote "The Australian, British, and USCF all have their own variations". I'm not sure Australia ever had a distinct variation (as opposed to just applying the rules differently), unlike NZ, who do have their own variation (reverse pairings in the middle rounds).
i remember before though that Geurt or Stewarts opinions have not always been the last word on the matter. I think Bill has said this a few times on other rules.

Bill Gletsos
05-04-2005, 09:49 PM
i remember before though that Geurt or Stewarts opinions have not always been the last word on the matter. I think Bill has said this a few times on other rules.I think what you will find I have said is that they dont always agree.

I note that last back in Sept 2003 Cathy Rogers wrote on the old ACF BB "The opinion I wanted was from British IA Stewart Reuben who is my pick for the best arbiter in the world."

I am inclined to take Reubens view over Geurts, as Geurt seems to lack any experience with events other than those at the highest level.

In the past when I have wanted a definitive answer to a question regarding the Laws of Chess, I've asked Stewart Reuben.

Denis_Jessop
05-04-2005, 10:29 PM
Denis, you have plenty of faith in FIDE which is probably misplaced.



Discretion is important I agree, but pairing is such an easy chore. The rules for pairing are only a set of either/or instructions. If there is are two conflicting instructions in the rules there has to be a consistent ruling as to which instruction is to be followed. The ACF is the only thing that can do that at the moment in Australia.

As I say discretion is OK for a "deliberate touch" ruling but pairing should be made black and white.



Pairing is not rocket science.



Not acceptable, Denis. The ACF should not allow a situation to continue where arbiters can make arbitery decisions were the pair process can be made perfectly transparent and repeatable by a ruling from the ACF stating the colour takes preference over group.

Final round pairings are extremely important because of the white players' advantage. If the arbiters make ad hoc discretionary reversal of colours the whole process starts to lose its legitimacy. The ACF has a duty to fix this problem.

Cheers

I was debating with myself whether this posting was, with all due respect to Duff, worthy of a reply.

The problems are that it quite misconceives the problem, then engages in some muddled thinking and ends with an "the ACF should fix it" exhortation of a typical kind in which it is not clear just what is to be fixed or how the poster thinks it might be done.

Statements like "pairing is easy" are not to the point. (Wrong pairings are a piece of cake but right ones are not :P ) I wonder how many tournaments Duff has directed using manual pairings according to the Dutch system ad how easy he found it to get the right pairing strictly according to the rules?

Of course one can always manually pair the players. But the question that I was addressing was the use of computer pairings and how they are to be treated. There is no doubt that if a computer is used the arbiters are still finally responsible and the computer is merely a guide. It may be an excellent one and I can say from having used manual and later SP pairings that when using SP the number of queries from players about the legality of the draw dropped dramatically - and not because I knew nothing about the Swiss Rules when pairing manually ;)

To say that pairing is not rocket science adds nothing to the debate but may suggest a limited experience in running Swiss events. As I'm sure you know there are several different sets of Swiss rules to begin with. It has already been said, correctly, in this Forum that the most commonly-used rules (and those used by SP and probably most other computer programs) are the FIDE Dutch Rules and that these are quite difficult to interpret. It's not a question of rocket science, if you insist, but one of interpretation of language. A Canberra colleague of mine even suggests that the rules were written not for manual pairing but as a basis for a computer program though wht a computer programmer would find them easier to understand than any other reader is not clear. They may also suffer from having been written in English by people whose native language is not English (let alone not by experienced drafters of rules) though I wouldn't be sure about that.

As others have pointed out, the ACF cannot insist that the FIDE rules be interpreted in a certain way. Moreover even if it could, the preference of colour over group (or not) is not the right question. The real question at issue here is how does SP apply the Dutch rules and is it the right way. No one is suggesting that arbiters should, or do, make arbitrary decisions. Those debated cases of arbiters making changes turn on the arbiters being of the view that SP has wrongly applied the Swiss Rules. Those are decisions based on the arbiter's skilled interpretation of the rules as they apply to a particular case, not to some general application.

Finally, I am a bit puzzled by the reference to discretionary reversal of colours. The reversal of colours has nothing to do with colour over group. The issue is the actual pairings. As far as I know the recorded cases of arbiters varying the SP draw relate to this issue. The reasons for varying pairings may relate to issues other than mere colour such as downfloating. It would be a gross oversimplification to suggest that the ACF or anyone else could make a simple statement about the rules covering the various possibilities, hence my reference to the Preface to the Laws.

DJ

Duff McKagan
06-04-2005, 02:02 AM
I was debating with myself whether this posting was, with all due respect to Duff, worthy of a reply.
With all due respect, it was worth a reply ipso facto you have replied.


The problems are that it quite misconceives the problem,
No, I have not misconceive the problem, I have identified a problem that you put in the too hard basket.


... then engages in some muddled thinking
I quite thought that of all people, you Denis, would not have allowed yourself to adopt the same disparaging remarks oft employed by the regular bullies who infect this BB.


...and ends with an "the ACF should fix it" exhortation of a typical kind in which it is not clear just what is to be fixed or how the poster thinks it might be done.
Utterly false. I stated that "The ACF should not allow a situation to continue [by ruling] the colour takes preference over group." As to how it should be done, it is called leadership.


Statements like "pairing is easy" are not to the point. (Wrong pairings are a piece of cake but right ones are not :P ) I wonder how many tournaments Duff has directed using manual pairings according to the Dutch system ad how easy he found it to get the right pairing strictly according to the rules?
In excess of 100 events by manual pairing, Denis. I recall three instances where I could find no way to avoid an legal pairing.


There is no doubt that if a computer is used the arbiters are still finally responsible and the computer is merely a guide.
With all due respect, there is doubt. The arbiter relies on a chip and software in the clock and so should it be for pairings.


To say that pairing is not rocket science adds nothing to the debate but may suggest a limited experience in running Swiss events.
It may also suggest that I know pairing is easy. I do not for one moment think that you find pairing difficult. Therefore, I cannot understand why you defend tampering to prioritise group ahead of colour.


A Canberra colleague of mine even suggests that the rules were written not for manual pairing but as a basis for a computer program though wht a computer programmer would find them easier to understand than any other reader is not clear. They may also suffer from having been written in English by people whose native language is not English (let alone not by experienced drafters of rules) though I wouldn't be sure about that.
A bemusing anacdote, but it contributes little to the problem at hand. ie. arbiters fiddling with pairings because SP did not give them the answer that they would like. They then arbitrarily arbitrate an arbiters' arbitration.


As others have pointed out, the ACF cannot insist that the FIDE rules be interpreted in a certain way.
Not true. The ACF can insist on anything it likes. The crunch comes when people decide to abide or not.


No one is suggesting that arbiters should, or do, make arbitrary decisions. Those debated cases of arbiters making changes turn on the arbiters being of the view that SP has wrongly applied the Swiss Rules.
Quite nonsensical. I am not "no one" and I have said that arbiters make arbitrary decisions.


Those are decisions based on the arbiter's skilled interpretation of the rules as they apply to a particular case, not to some general application.
If arbiters were as skilled as you appear to think they are, then they would be in accord. However, they are not in accord, which suggests varying degrees of skill. The conclusion I draw from that is that arbiters need some guidance, which I will call, a ruling by the ACF.


With all due respect, Denis, could you spend less time debating with yourself over whether or not to reply, and more time on fixing the problem.

Cheers,
Duff

Garvinator
06-04-2005, 11:35 AM
Duff,

Can you please show all of us from the dutch pairing rules which sections support your cause and disprove the sections that dont support your cause?


Secondly, the acf council would not be able to make any decent binding decision for the following reason:

They would be advised by arbiters all over the country and they would have different views, most likely either from the top v bottom or colour group and both would have arguments that support their cause.

The acf council would then be in a difficult position in trying to find an acceptable solution. When in a situation like this, the normal way to find a solution on this is to compromise, unfortunately there really isnt a compromise, it is one or the other.

Garvinator
06-04-2005, 11:39 AM
For everyone,

Is this way of pairing legal and would it be a compromise from the situation we currently have.

1) Eight players on same score
2) Divide players into top and bottom by rating
3) Check who has played who
4) Pair top seed of the eight with someone from the bottom half who has the most opposite colour preference ie if the top seed is on +2, then you are looking for someone in the bottom half who is closest to -2.
5) Check float status
6) Continue with pairing till all legal pairings have been found.

Would this work?

Oepty
06-04-2005, 03:09 PM
A simple question. If player A, the higher ranked player, has had the following colours WWB and you were pairing them against player B with WBW what pairing should be made? Why?

Scott

Garvinator
06-04-2005, 03:49 PM
A simple question. If player A, the higher ranked player, has had the following colours WWB and you were pairing them against player B with WBW what pairing should be made? Why?

Scott
In this situation, I believe that player A would be white.

Reasoning is as follows:

Both players are +1. In the most previous round, Player A was black, Player B was white, therefore under the pairing rules, they alternate, therefore Player A is white.

E. Color Allocation Rules
For each pairing apply (with descending priority):

E1. Grant both color preferences. (both players are +1, therefore they both should be black, which obviously cant be done)

E2. Grant the stronger color preference. (both players are +1, so neither has a stronger colour preference)

E3. Alternate the colors to the most recent round in which they played with different colors. ( Player A was black in the last round and Player B was white, therefore the colours alternate, meaning Player A is now white)

E4. Grant the color preference of the higher ranked player. In the first round all even numbered players in S1 will receive a color different from all odd numbered players in S1.

Oepty
06-04-2005, 04:05 PM
In this situation, I believe that player A would be white.

Reasoning is as follows:

Both players are +1. In the most previous round, Player A was black, Player B was white, therefore under the pairing rules, they alternate, therefore Player A is white.

E. Color Allocation Rules
For each pairing apply (with descending priority):

E1. Grant both color preferences. (both players are +1, therefore they both should be black, which obviously cant be done)

E2. Grant the stronger color preference. (both players are +1, so neither has a stronger colour preference)

E3. Alternate the colors to the most recent round in which they played with different colors. ( Player A was black in the last round and Player B was white, therefore the colours alternate, meaning Player A is now white)

E4. Grant the color preference of the higher ranked player. In the first round all even numbered players in S1 will receive a color different from all odd numbered players in S1.

That sounds right, thankyou.
Scott

Ian Rout
06-04-2005, 04:29 PM
Yes, that's right. The rule about who gets what colour in a pairing is generally clear, it's who plays who that is difficult.

Note that this outcome is independent of what the other pairings are; if And B are supposed to play, then for whatever reason the draw is re-done and is different but A and B are still playing, then they will have the same colours as previously.

Ian Rout
06-04-2005, 04:42 PM
Footnote: Two common errors by inexperienced (i.e. have not actually read the rules) operators in manually pairing -

a) Looking at last round only, if totals are the same. The entire history should be considered (E3).

b) Giving due colour to higher rated rather than higher ranked (E4). In the case of an upfloat the higher rated could be lower ranked.

Garvinator
06-04-2005, 05:18 PM
which situation is more correct out of the two following:

Eight players, seven are on the same score, the eighth player is half a point lower.

Would the pairings be:

a) 1 v 4, 2 v 5, 3 v 6, 7 v 8
b) 1 v 5, 2 v 6, 3 v 7, 4 v 8

It is a straight pairings question, no colours to worry about or float issues.

Oepty
06-04-2005, 06:08 PM
Footnote: Two common errors by inexperienced (i.e. have not actually read the rules) operators in manually pairing -

a) Looking at last round only, if totals are the same. The entire history should be considered (E3).

b) Giving due colour to higher rated rather than higher ranked (E4). In the case of an upfloat the higher rated could be lower ranked.

Ian. What actually happened was I was doing some manual pairings at home and then doing them with swiss perfect to see if I was getting the same pairings. I got the same pairings but two pairings reversed. I just forgot about E3 and could not understand why the difference occurred.
Scott

Oepty
06-04-2005, 06:15 PM
which situation is more correct out of the two following:

Eight players, seven are on the same score, the eighth player is half a point lower.

Would the pairings be:

a) 1 v 4, 2 v 5, 3 v 6, 7 v 8
b) 1 v 5, 2 v 6, 3 v 7, 4 v 8

It is a straight pairings question, no colours to worry about or float issues.

a is correct. This is because the first 7 players consitute a score group in them selves. This means that they should be paired first, with the best pairing possible before 8 is even considered for pairings. The player who is left unpaired is then paired with 8 if that is possible.
b is a possible pairing if there is colour issues or float issues or players have played each other previously
Scott

Garvinator
06-04-2005, 07:46 PM
Freddy,

Are you going to give us some of the pairings you were going to ask about?

Ian Rout
06-04-2005, 08:24 PM
Ian. What actually happened was I was doing some manual pairings at home and then doing them with swiss perfect to see if I was getting the same pairings. I got the same pairings but two pairings reversed. I just forgot about E3 and could not understand why the difference occurred.
Scott
Sorry - just a general observation, not a reference to anyone in particular.

Denis_Jessop
06-04-2005, 11:40 PM
With all due respect, it was worth a reply ipso facto you have replied.


No, I have not misconceive the problem, I have identified a problem that you put in the too hard basket.


I quite thought that of all people, you Denis, would not have allowed yourself to adopt the same disparaging remarks oft employed by the regular bullies who infect this BB.


Utterly false. I stated that "The ACF should not allow a situation to continue [by ruling] the colour takes preference over group." As to how it should be done, it is called leadership.


In excess of 100 events by manual pairing, Denis. I recall three instances where I could find no way to avoid an legal pairing.


With all due respect, there is doubt. The arbiter relies on a chip and software in the clock and so should it be for pairings.


It may also suggest that I know pairing is easy. I do not for one moment think that you find pairing difficult. Therefore, I cannot understand why you defend tampering to prioritise group ahead of colour.


A bemusing anacdote, but it contributes little to the problem at hand. ie. arbiters fiddling with pairings because SP did not give them the answer that they would like. They then arbitrarily arbitrate an arbiters' arbitration.


Not true. The ACF can insist on anything it likes. The crunch comes when people decide to abide or not.


Quite nonsensical. I am not "no one" and I have said that arbiters make arbitrary decisions.


If arbiters were as skilled as you appear to think they are, then they would be in accord. However, they are not in accord, which suggests varying degrees of skill. The conclusion I draw from that is that arbiters need some guidance, which I will call, a ruling by the ACF.


With all due respect, Denis, could you spend less time debating with yourself over whether or not to reply, and more time on fixing the problem.

Cheers,
Duff

All very interesting but I remain unmoved. Our differences have a quite simple basis so it seems. Duff believes that the SP program accurately implements the FIDE Dutch Rules and thus its pairings are immutable. I and the other arbiters who vary the pairings believe SP does not do so, and with good reason in my experience.

DJ

PS Sorry if you thought I was being rude. Iwasn't - just wait until I am :evil:

Denis_Jessop
07-04-2005, 12:05 AM
It is my understanding that the program called Petunia (it is now called Swiss Master) is officially endorsed by the FIDE Swiss Pairings Committee.

Stewart Reubens notes on page 34 of the Chess Organiser's Handbook Second Edition that "I have never heard comments about the Petunia Program for the Dutch System giving the wrong pairings".

An intersting point he makes on the same page is "They are all based on the premise that throughout the tournament all players on all score groups are paired top half v bpottom half in rating order after taking into account colour equalisation and then colour alternation." (the bolding of after is mine)

Petunia/Swiss Master is also the program the Geurt swears by.

It would be interesting to compare the draw done for the Doeberl by SP and by Swiss Master 5.1.

N.B. On the FIDE web site under the section on Swiss rules, Regualtions for swiss tournaments, Miscellaneous it notes "The FIDE Swiss Rules pair the players in an objective and impartial way, and different arbiters following the pairing rules should arrive at identical pairings. "


Bill, for a minor example of SP v SM5 (not the Doeberl) please see my posting #29 on this thread. I still haven't tried to unravel the problem!

DJ

ursogr8
07-04-2005, 08:01 AM
All very interesting but I remain unmoved. Our differences have a quite simple basis so it seems. Duff believes that the SP program accurately implements the FIDE Dutch Rules and thus its pairings are immutable. I and the other arbiters who vary the pairings believe SP does not do so, and with good reason in my experience.

DJ


Denis

I have been watching this thread with interest, but with no expertise on pairing rules. I am getting a bit of a free education, thanks to the elucidatory posts of yours and quite a few others.

One point that DMcK was making intrigued me, but you seem to have read it differently. I thought he was saying that SP did not accurately implement the FIDE Dutch Rules; and neither did the arbiting community...judging on the conflict and interpretation going on in this thread. If this is an fair summary of this one point of the discussion then why don't we just accept SP pairings; this would have the benefit of speed and reduced controversy compared with manual intervention.
Or asking the question another way, what tangible benefits come from manual intervention, that would outweigh the down-sides of delay and controversy?

regards
starter

arosar
07-04-2005, 09:50 AM
I always thought that sometimes manual intervention is required because SP produces such 'inequitable' pairings - particularly in later rounds - no matter that those pairings may be as per the rules. :doh: :hmm:

AR

shaun
07-04-2005, 09:53 AM
Or asking the question another way, what tangible benefits come from manual intervention, that would outweigh the down-sides of delay and controversy?

regards
starter

A more general answer than just this specific question.

Apart from the missaplication of the last round pairing rules (which may be version specific), Swiss Perfect has always followed the Dutch rules correctly. However there are two things an Arbiter must/can do. The thing an arbiter must do is to be able to explain how the pairings were arrived at, even if they were paired by computer. This at least means that the arbiter should check the pairings after the computer has generated them, and before they are posted. The thing an arbiter can do is to alter the pairing rules, as we did, and annouced, at the Doeberl Cup. This clearly involves manual intervention by the arbiter in this case, as Swiss Perfect doesn't have an option to modify the pairing rules it uses.
So why do Arbiters fiddle with pairings? Mainly becasue they don't like some aspects of the Dutch pairing rules. Possibly this dislike causes them to interpret the rules to suit their likes, but you would have to ask them. And while my personal opinion is that topvbottom gives better pairings (tournament result wise) than pairing for colour, this is not how the rules are written.

1min_grandmaster
07-04-2005, 10:29 AM
And while my personal opinion is that topvbottom gives better pairings (tournament result wise) than pairing for colour, this is not how the rules are written.
Why does top vs bottom give better pairings? I would have thought it causes more problems in later rounds with possible barred pairings. Hence I think prioritising colour gives better pairings.

arosar
07-04-2005, 10:43 AM
Look, what's the fundamental principle behind Swiss? Isn't it top vs bottom?

AR

shaun
07-04-2005, 11:01 AM
Why does top vs bottom give better pairings? I would have thought it causes more problems in later rounds with possible barred pairings. Hence I think prioritising colour gives better pairings.

Because if the ultimate aim of a chess tournament is to find a winner (and prize winners) then in the final rounds (or in fact any round after the half way point IMO) having players who "should" play each other based on seeding is preferrable than them not playing due to a colour preference issue.
(I should have prefaced my previous remarks with the fact I was mainly referring to the last/second last rounds)

shaun
07-04-2005, 11:04 AM
Look, what's the fundamental principle behind Swiss? Isn't it top vs bottom?

AR

No. The ultimate aim is to find a winner. How that is best done is the cause of the debate. Some players argue that the process is flawed if they have to play a harder opponent that the "seedings" would suggest, while others would argue that the process is flawed if they have to play with a colour (usually Black) that the colour history doesn't suggest.

Denis_Jessop
07-04-2005, 12:48 PM
A more general answer than just this specific question.

Apart from the missaplication of the last round pairing rules (which may be version specific), Swiss Perfect has always followed the Dutch rules correctly. However there are two things an Arbiter must/can do. The thing an arbiter must do is to be able to explain how the pairings were arrived at, even if they were paired by computer. This at least means that the arbiter should check the pairings after the computer has generated them, and before they are posted. The thing an arbiter can do is to alter the pairing rules, as we did, and annouced, at the Doeberl Cup. This clearly involves manual intervention by the arbiter in this case, as Swiss Perfect doesn't have an option to modify the pairing rules it uses.
So why do Arbiters fiddle with pairings? Mainly becasue they don't like some aspects of the Dutch pairing rules. Possibly this dislike causes them to interpret the rules to suit their likes, but you would have to ask them. And while my personal opinion is that topvbottom gives better pairings (tournament result wise) than pairing for colour, this is not how the rules are written.

Hi Shaun

I basically agree with the proviso that what you say assumes that the Dutch rules are capable of one interpretation only. I don't think that is the case. Moreover are your comments about SP's correctly following the Dutch rules confined to colour preference v. top/bottom? To clarify, what do you say about SP's predilection for downfloating players to a group further than that immediately below when it could make a valid pairing in the group below? In last year's ACT Championship it make a quite bizarre set of pairings for part of one round that turned on this quirk.

DJ

Denis_Jessop
07-04-2005, 01:11 PM
Denis

I have been watching this thread with interest, but with no expertise on pairing rules. I am getting a bit of a free education, thanks to the elucidatory posts of yours and quite a few others.

One point that DMcK was making intrigued me, but you seem to have read it differently. I thought he was saying that SP did not accurately implement the FIDE Dutch Rules; and neither did the arbiting community...judging on the conflict and interpretation going on in this thread. If this is an fair summary of this one point of the discussion then why don't we just accept SP pairings; this would have the benefit of speed and reduced controversy compared with manual intervention.
Or asking the question another way, what tangible benefits come from manual intervention, that would outweigh the down-sides of delay and controversy?

regards
starter


Hi Starter

I must say, on reflection, that I am not sure what Duff was saying on this point, hence my original reference to thinking methods that he didn't like!

My problem with his argument is that it seems to assume that the Dutch rules are perfectly clear when they aren't. His reference to arbiters taking different views of the same matter reflecting on their degree of skill does not necessaily follow. It may be, and is more likely to be, that the rules are capable of more than one interpretation and that the correct or intended meaning is far from clear. This is a situation with which lawyers are only too familiar. He is also mistaken in asserting that I am supporting group over colour. What I am intending to say is that no such general assertion/instruction can be made which is one reason why the ACF should not attempt it. Moreover in my view SP does other odd things in this area especially regarding downfloats.

I thought that Duff was saying that the program should prevail but perhaps not. On the other hand I don't believe that his solution is workable, as any general statement of principle would only add confusion to an already difficult set of rules. How Duff finds them easy is a mystery but it would take ages and lots of individual examples to illustrate this point.

Cheers
DJ

shaun
07-04-2005, 02:03 PM
Hi Shaun

I basically agree with the proviso that what you say assumes that the Dutch rules are capable of one interpretation only. I don't think that is the case.
If you strictly follow the Dutch rules, and the method prescribed in them, you will only arrive at 1 pairing. And I believe there is only one interpretation of the rules, unless you go through some linguistic gymnastics.



Moreover are your comments about SP's correctly following the Dutch rules confined to colour preference v. top/bottom? To clarify, what do you say about SP's predilection for downfloating players to a group further than that immediately below when it could make a valid pairing in the group below? In last year's ACT Championship it make a quite bizarre set of pairings for part of one round that turned on this quirk.

DJ

While I vaguely remember the case I would have to see the exact pairing information before I could comment. However I wouldn't be suprised if the SP pairing were indeed the correct ones, and the alternative pairing resulted from an alternative reading of the rules.

Bill Gletsos
07-04-2005, 02:37 PM
Bill, for a minor example of SP v SM5 (not the Doeberl) please see my posting #29 on this thread. I still haven't tried to unravel the problem!Yes, I did take note of that post.

Based on FIDE's statement that "The FIDE Swiss Rules pair the players in an objective and impartial way, and different arbiters following the pairing rules should arrive at identical pairings." it would seem reasonable to believe that all programs that are endorsed by the FIDE Swiss Pairings Committee would all generate the same parings.

Pentunia/Swiss Master is endorsed. Cost is 50 euro.
According to its website Swiss Manager is FIDE approved. It was used for the last Olympiad in Spain and the World Youth in Greece. Cost of Pro version is 199 euro.

Swiss-Chess for Windows is FIDE endorsed. Price is 135 euro.

SVBoss is also endorsed by FIDE but it doesnt look like its been updated since Nov 2001.

It would be intersting to see what pairings they give.

Bill Gletsos
07-04-2005, 02:49 PM
FWIW Reuben in the First Edition of his Chess Organiser's Handbook provides a 60 player 10 round simulated Swiss event based on the FIDE Dutch rules.

A lot of answers about how the FIDE Dutch rules actually are applied and hence how to correctly interpret them can be learnt by manually working through all 10 rounds of that example.

1min_grandmaster
07-04-2005, 08:15 PM
Because if the ultimate aim of a chess tournament is to find a winner (and prize winners) then in the final rounds (or in fact any round after the half way point IMO) having players who "should" play each other based on seeding is preferrable than them not playing due to a colour preference issue.
I don't quite understand the part where you say:

players who "should" play each other based on seeding
It seems to mean that if someone is lower seeded, but they are on the same score as someone higher seeded in the final rounds, then the higher seeded players "should" win, so they "should" have the right to play the lower seeded players. I don't really agree with this. By prioritising colours, you are still equally likely to find a winner, which as you stated is the aim.

Duff McKagan
08-04-2005, 07:25 AM
Denis, it might appear that I am singling you out but I hope you do not take all this personally. However, I am not actually defending SP or colour over group. I am actually trying to promote consistency between arbiters.


My problem with his argument is that it seems to assume that the Dutch rules are perfectly clear when they aren't.
It does not matter if SP is not 100% right in interpreting the Dutch rules exactly as you or anyone else thinks they should be, 99% would also be OK.


It may be, and is more likely to be, that the rules are capable of more than one interpretation and that the correct or intended meaning is far from clear.

I repeat, it does not matter if SP is not 100% right in interpreting the Dutch rules exactly as you or anyone else thinks they should be, 99% would also be OK.


What I am intending to say is that no such general assertion/instruction can be made which is one reason why the ACF should not attempt it.

As a lawyer, you of all people, would be aware that the court has a duty to make a decision. Reserving a decision for years is an abject avoidance of responsibility. For the sake of fairness and consistency in tournament governance across the country, the ACF should say "SP rules OK".


On the other hand I don't believe that his solution is workable, as any general statement of principle would only add confusion to an already difficult set of rules.

Totally untrue. "Confusion" is reduced by consistency. Inconsistency bought about by arbiters rejigging the pairings in an arbitrary manner leads inexorably toward confusion. I need only point at this thread and say QED.

Cheers :)

pax
08-04-2005, 09:38 AM
As a lawyer, you of all people, would be aware that the court has a duty to make a decision. Reserving a decision for years is an abject avoidance of responsibility. For the sake of fairness and consistency in tournament governance across the country, the ACF should say "SP rules OK".

I have asked this question before, Duff. How would you expect the ACF to enforce such a rule?

ursogr8
08-04-2005, 10:44 AM
I have asked this question before, Duff. How would you expect the ACF to enforce such a rule?

pax

I think you ask a reasonable question. It is similar to a question asked by Denis Jessop on other issues.
And I think DMcK should be required to post here some ideas on the thorny question of how to encourage/endorse/enforce a standard.

Bill's earlier post indicated that FIDE themselves were endorsing three other products. In some cases endorsement by the national/international authority is sufficient for standardisation to start to occur.
If SP were so wide of the mark then perhaps we would need to re-visit our use of it. But I have seen nothing posted here to measure the downside of using SP without manual intervention.
I don't have the answers, I am first to admit. But the manual intervention in (pairings made by) software, where it seems there is controvery in how to intervene, does seem undesirable for the delays it causes.

starter

Duff McKagan
08-04-2005, 10:46 AM
I have asked this question before, Duff. How would you expect the ACF to enforce such a rule?

Sorry pax, didn't answer you. :oops:

Actually, I am sure that when push comes to shove, the ACF cannot enforce it. However, there is an implied top end standing that the ACF has not ever used, to the best of my knowledge. I think that if the ACF were to simply issue a public directive saying that it wishes all arbiters to use the pairing generated by SP without manual alteration, and in the long run, it would become the standard practice.

The ACF holds some moral sway and political power primarily by virtue of it being *the* national federation. Players knowing of the ACF directive, would know they had the moral support of the ACF behind them, if they wanted to challenge a manual pairing. Those emboldened players would then be able to provide so much "pressure" on recalcitrant arbiters, that arbiters generally, would likely decide not to venture down that road in the first place.

Cheers

jase
08-04-2005, 02:54 PM
Both colour preference and top v bottom are important criteria for all Swiss pairing systems. Which should hold greater sway is as much a philosophical point of difference as a statistical one – since both are essential in the quest to determine the winners of tournaments. To some of the points made thus far, and some of my own thoughts on the topic:


In section B2 (Relative Criteria) part (b) it says that "as many players as possible should recieve their colour preference". It does not mention top v bottom as one of the pairing criteria <This indicates that either colour is more important or that top v bottom is so fundamental it wasn't explicitly stated>

The fundamental nature of top v bottom in the Swiss style of pairings is outlined in the previous section A6 Subgroups :

To make the pairing, each score bracket will be divided into two subgroups, to be called S1 and S2.
In a heterogeneous score bracket S1 contains all players moved down from a higher score bracket.
In a homogeneous score bracket S1 contains the higher half (rounding downwards) of the number of players in the score bracket.
The number of players in S1 will be indicated by "p", indicating the number of pairings to be made.
In both cases S2 contains all other players of the score bracket.
In both S1 and S2 players are ordered according to A2.

I learnt the Swiss pairing system from Cathy Rogers and Peter Parr around 1990. At that time the principle of top v bottom held sway over colour alternation. This scenario had been the case for decades – even before ratings were used to seed players, Swiss pairings were done using Buchholtz scores [and indeed, a modified version of such a system remains in use at events like Olympiads] with top v bottom being an underlying principle.

I have always considered top v bottom an excellent way to determine winners of tournaments: the strong half line up against the weaker half, with the form players rising to the top. The system is extremely efficient. Manually doing the pairings throughout the 1990s, I rarely encountered situations where a violation of colour rules [+/-2] forced me to make a radical alteration to pairings.

Statistically there is always a demand for Black in the top half of a tournament, and a demand for White in the lower half, since White is scoring better. This imbalance used to be monitored by restricted transpositions and exchanges within a ratings band. I have done pairings for Australian Championships/Opens alongside Cathy Rogers, John Frew, Gary Bekker, and Manuel Weeks, and found a consistent pairings philosophy with each.

I have seen several estimations of the statistical advantage for White over Black, but all hover around the mark of 55% White, 45% Black. Therefore much of this debate centres around giving players equal opportunity at both colours to balance this ledger.


When looking at a new pairing system the designers consulted a number of leading GM's about what they wanted in a pairing system. The majority wanted colour to have priority

There was a zonal tournament about 5 years ago where the Swiss Perfect pairings were completely different to how I would have done them manually. I wrote to the program designer, who emailed me a file written by Geurt Gijssen to explain his Dutch Pairing System. In this file it explains that a survey was undertaken of leading players, in which they expressed a desire for colour alternation.

The hitch is that these top players are meeting opponents with very similar ratings. When GM Joel Benjamin was in Sydney in 1999 he told me he hadn’t played anyone rated under 2000 in about 5 years! The ratings spread of elite tournaments is a very narrow band, compared with local tournaments, where the spread from top to bottom usually exceeds 1,000. Examining FIDE’s ratings tables, the estimated 10% difference between White and Black is the equivalent of about 40 rating points. However making exchanges between subgroups to satisfy colour alternation, in local tournaments, will often mean that your transposed/exchanged opponent has altered by over 100 points.

Arbiters used to work by a 100 point rule: transpositions and exchanges were done so long as the ratings differential between the original and revised pairings did not exceed 100 points.

When the Gijssen’s Dutch Pairing System came in vogue, my reading of C Pairing Procedures was consistent with what I’d learnt from Cathy and Peter. My interpretation was that Top v Bottom was an over-riding principle, but care needed to be taken to ensure that players had a fair distribution of Whites and Blacks. The Zonal tournament mentioned above shook that foundation:

There were 8 players on a scoregroup. The rating range of the 8 players was, as I recall, over 500 points. Each had received equal Whites and Blacks. The officials were using Swiss Perfect; on this occasion, the only possible pairings that would ensure colour alternation would be to pair the two highest rated players. And this is precisely what Swiss Perfect did. I thought this was ludicrous, given that ALL players would have a colour imbalance after this round anyway. I would have kept top v bottom and allowed one or two players to have the same colour as their previous round – I would not have considered such drastic exchanges necessary merely to ensure colour alternation.

However it has become increasingly clear, from amendments to the wording of the Dutch System, and the opportunities I’ve had to talk to the likes of Geurt Gijssen and Stewart Rueben, that the algorithms used by Swiss Perfect meet the aims of the Dutch System.

I do think that both subgroupings and colour preferences are important criteria, however I remain at odds with colour taking such strong preference. I think that most exchanges between S1 and S2 in Australian tournaments will greatly exceed the statistical 10% discrepancy between White and Black, therefore I prefer to maintain subgroups where feasible. Relative Criteria B.6 seems to acknowledge that determining the winners of tournaments is best achieved by pairing top v bottom, doing away with most of the colour preferences for the last round.

I do recognise the appeal of consistent standards, and would be comfortable with enforcing a blanket policy on this issue if players so desired. Adhering to the Dutch rules, as per Swiss Perfect [and Swiss Master, I am sure, since Geurt Gijssen uses that program] saves a lot of time for arbiters, and provides greater certainty, however I think for tournaments with large rating spreads it creates more anomalies than is necessary.

Garvinator
08-04-2005, 05:56 PM
I do recognise the appeal of consistent standards, and would be comfortable with enforcing a blanket policy on this issue if players so desired.
By saying players, wouldnt this then have a blanket policy decided by people who, on the whole, have very little knowledge of how the pairings systems work and why re pairings are sometimes done.

All that would happen is that players would want the arbiters to stand by what the computer has paired because they got a harder or unfavourable pairing after a manual repairing.

Garvinator
08-04-2005, 05:59 PM
Hello Jason,

In an earlier post of mine, I proposed a possible compromised solution to this debate, but didnt receive a reply from anyone.

What do you think about my compromised solution below, do you think it could work or is it rubbish :uhoh:

Is this way of pairing legal and would it be a compromise from the situation we currently have.

1) Eight players on same score
2) Divide players into top and bottom by rating
3) Check who has played who
4) Pair top seed of the eight with someone from the bottom half who has the most opposite colour preference ie if the top seed is on +2, then you are looking for someone in the bottom half who is closest to -2.
5) Check float status
6) Continue with pairing till all legal pairings have been found.

Would this work?

jase
08-04-2005, 06:40 PM
By saying players, wouldnt this then have a blanket policy decided by people who, on the whole, have very little knowledge of how the pairings systems work and why re pairings are sometimes done.

All that would happen is that players would want the arbiters to stand by what the computer has paired because they got a harder or unfavourable pairing after a manual repairing.


Two points here. To the first, I think this very thread demonstrates that players have acquired a lot of knowledge of how pairings systems work.

On the second, electing to accept the Dutch interpretation of the Swiss principles, ie accepting Swiss Perfect/Swiss Master etc pairings [subject to the satisfaction that said pairings are shown to be in accordance with the pairing rules] does not mean that you get "a harder or unfavourable pairing". It does mean the pairing is different to the system I outlined in my previous post, but it does not follow that the pairing is necessarily more difficult.

Sometimes it will work in your favour, and other times it will not. It just results in a different set of pairings to determine the outcomes of a tournament - outcomes which I think are more skewed than if top v bottom were prioritised, at least over alternation, - but the differences may be not be significant enough to outweigh the benefits of uniformity.

Regarding your compromised pairing method, it appears to me to be a rudimentary version of what is in place already, with some of the finer points of pairing rules omitted. It seems quite arbitrary to find the top seed "someone" to play. If more than one player meets this criteria, I need a mechanism to determine who plays who.

Pairing manually I like to set my pairing cards out - ie 1 v 5, 2 v 6, 3 v 7, 4 v 8 - and then go through a detailed routine [eg check for repeat pairings, assess colour preferences, float status etc], making transpositions from the bottom up where necessary [ie the smallest transposition I could make would be players 7 and 8].

Your system would work, but its less efficient than what's in place now. And different arbiters will not arrive at the same pairings.

I hope you don't find my reply "disappointing" :hmm:

Kevin Bonham
08-04-2005, 06:59 PM
Players knowing of the ACF directive, would know they had the moral support of the ACF behind them, if they wanted to challenge a manual pairing.

I don't see why players would need the moral support of anything.

I think that an arbiter who overrides the pairings manually on this issue when they have not announced such an intention prior to the tournament really is asking for an appeal. Furthermore if I was hearing such an appeal I think I would uphold it, because the arbiter should not override the program without notice because of their interpretation, but should only do so when the program is categorically wrong.

If an arbiter announces before the tournament starts that such a subjective override will be applied, that's fine by me, assuming it is consistent.

Perhaps all we're waiting for is a case where a player does appeal against an override and is successful.

Garvinator
08-04-2005, 07:13 PM
Two points here. To the first, I think this very thread demonstrates that players have acquired a lot of knowledge of how pairings systems work. a point that has been made to me many times is that this bb does not generally represent the greater chess community. Therefore I would doubt that even though this thread has been a good thread for expanding knowledge of pairing rules, only a small number of the chess playing population are reading this thread.


On the second, electing to accept the Dutch interpretation of the Swiss principles, ie accepting Swiss Perfect/Swiss Master etc pairings [subject to the satisfaction that said pairings are shown to be in accordance with the pairing rules] does not mean that you get "a harder or unfavourable pairing". It does mean the pairing is different to the system I outlined in my previous post, but it does not follow that the pairing is necessarily more difficult. I was not saying that re pairing does/does not give harder or more unfavourable pairings, what I was more trying to highlight was that players and people in general generally tend to remember negative experiences more often.


Regarding your compromised pairing method, it appears to me to be a rudimentary version of what is in place already, with some of the finer points of pairing rules omitted. It seems quite arbitrary to find the top seed "someone" to play. If more than one player meets this criteria, I need a mechanism to determine who plays who. my compromised idea was a five second writing based on thoughts i have had and discussions on here. It wasnt any anyway meant to be a binding set of rules etc.


Pairing manually I like to set my pairing cards out - ie 1 v 5, 2 v 6, 3 v 7, 4 v 8 - and then go through a detailed routine [eg check for repeat pairings, assess colour preferences, float status etc], making transpositions from the bottom up where necessary [ie the smallest transposition I could make would be players 7 and 8]. I am pleased to see this as this is very similiar to how i manually pair.


Your system would work, but its less efficient than what's in place now. And different arbiters will not arrive at the same pairings. Answered above :cool:


I hope you don't find my reply "disappointing" :hmm:not at all.

Oepty
08-04-2005, 07:46 PM
Sorry - just a general observation, not a reference to anyone in particular.

Why Sorry? Nothing you said insulted me or anything of the like. I thought they were perfectly fair comments. Just adding some background that's all.
Scott

Oepty
08-04-2005, 07:47 PM
Freddy,

Are you going to give us some of the pairings you were going to ask about?

Maybe, maybe not. I might simply just not have time.
Scott

Duff McKagan
09-04-2005, 07:02 AM
If an arbiter announces before the tournament starts that such a subjective override will be applied, that's fine by me, assuming it is consistent.

Kevin, you are assume that it is consistent. Unfortunately for your argument, it is not consistant. Different arbiters do different things in the same circumstances. That means that pairing procedure nation wide is inconsistently applied. Announcing manual overrides at the beginning does not make the our pairing system any more consistent, it simply says we reserve the right to fiddle with a consistent program.


Perhaps all we're waiting for is a case where a player does appeal against an override and is successful.

Why aren't I surprised with this abrogation of responsibility. The ACF never seems to make the rules or guide Australian chess. Here we have a typical example of the ACF hands off mind set, just let players go to the time trouble and effort of appealing and thereby set a precedent, when all that was required was for the ACF to issue a directive in the first place.

Cheers

Garvinator
09-04-2005, 08:57 AM
Here we have a typical example of the ACF hands off mind set, just let players go to the time trouble and effort of appealing and thereby set a precedent, when all that was required was for the ACF to issue a directive in the first place.

Cheers
you seem to believe that the acf has some overall power of responsibility. They dont have any such power. The acf cannot tell any state or territory ;) how to run their tournaments or what pairing program to use. Each state, territory and/or club is free to make whatever decisions it considers best for its tournaments and general well being.

The only time the acf has any real power is in acf events and even then its power is quite often limited.

Instead of blaming the acf for manual overrides of sp, perhaps you should be complaining about fide for not clarifying once and for all exactly how the dutch pairing rules are to be interpreted.

arosar
09-04-2005, 10:04 AM
Why aren't I surprised with this abrogation of responsibility. The ACF never seems to make the rules or guide Australian chess. Here we have a typical example of the ACF hands off mind set, just let players go to the time trouble and effort of appealing and thereby set a precedent, when all that was required was for the ACF to issue a directive in the first place.


Mate, we have to always strike a balance between the convenience that IT systems offer and their inherent weaknesses. So while SP can do away with much of the laborious task of manual pairings - it does produce, at times, poor outcomes. This is when the arbiter must intervene. Here we have a fairly good sample of the exactitude of software working hand-in-hand with human judgement.

I think that if the ACF does what you want it to do, we lose this balance and end up with a poorer situation overall.

Personally, I've always limited my concerns to matters OTB. I'll play anyone - some beginner or GM Rogers. But, of course, I'm a bloody patzer myself, so pairings hardly matter.

AR

Duff McKagan
09-04-2005, 02:21 PM
Here we have a fairly good sample of the exactitude of software working hand-in-hand with human judgement.

I think that if the ACF does what you want it to do, we lose this balance and end up with a poorer situation overall.


Amiel :) How are you going. Will you be taking pics at the NSW Open? Then maybe you could ruin this site by posting pics of me OTB.

Anyways, what you say about using software as a guide and wet ware for detail is the most compelling argument anyone has put up so far. In fact I would say that it is also the only argument for intervention that holds water.

The solution then, is to modify the current SP program or right one that pairs group over colour. A mod SP might be easier but expensive because of copyright, while righting a new program is a huge amount of work.

Unfortunately, until one of these two things are done, we have a situation where some arbiter will and some won't re-pair. It is inconsistent and therefore will result in some unfair breaks for some players.

This is why I think the ACF should resolve the matter for time being at least, by using the moral authority that comes from being our *national federation*.

I have probably said way too much already so I think I might stop banging on a bout it.

Cheers

Duff McKagan
09-04-2005, 02:30 PM
The acf cannot tell any state or territory ;) how to run their tournaments or what pairing program to use.

The ACF can tell the arbiters to do anything the ACF wants. The issue is, will the arbitors take any notice.


Instead of blaming the acf for manual overrides of sp, perhaps you should be complaining about fide for not clarifying once and for all exactly how the dutch pairing rules are to be interpreted.

Are you suggesting that Australian arbiters will actually listen to the FIDE before they listen to the ACF? I hope not, because currently the arbiters are doing what ever the hell they want. I say the ACF should provide guidance or declair a policy in support of standardised consistancy in pairing.

Cheers

jase
09-04-2005, 04:55 PM
Duff,

Paraphrasing and repitition doesn't add to your argument. Amiel answered about 6 of your posts on this thread in 6 lines - see post #101. You display little understanding of the administrative structure of Australian chess.


The ACF can tell the arbiters to do anything the ACF wants. The issue is, will the arbitors take any notice.

The ACF might also like to advise the Reserve Bank about monetary policy. And I guess the issue would be whether the Reserve would take any notice...


Are you suggesting that Australian arbiters will actually listen to the FIDE before they listen to the ACF? I hope not, because currently the arbiters are doing what ever the hell they want.

The ACF is a member of FIDE, and as such follows this body's rulings and direction. Regardless of any ACF directive, any FIDE-ratified event must satisfy not ACF guidelines, but FIDE. Arbiters are not doing "what ever the hell they want". Arbiters follow the guidelines of Swiss Pairing Systems. As approved by FIDE, there are several systems that use the Swiss principles, but derive different pairings. And really Duff, this is okay. It really is.

The aim is to find the winner of the tournament, and unless you have a double round robin, there will be inherent imperfections in any system. The one consistency that is essential is that an arbiter applies the same philosophy, in accordance with the fundamentals of Swiss Pairings, throughout an event.

Take a load off, because you are derailing what was previously an active and interesting thread.

Rhubarb
09-04-2005, 05:13 PM
unless you have a double round robin, there will be inherent imperfections in any system.Yes, the last time I copped even an equal deal on colour was a double round robin, last year's NSW Grade Matches.

Take a load off, because you are derailing what was previously an active and interesting thread.Would you rather have me continue whingeing, jase? :cool:

Duff raised some interesting points. From what I've understood, some of the situations in which manual overrides are deemed necessary are too complicated to fully explain in detail to all the players before the tournament begins, which does leave arbiters open to allegations of bias - even when there is a total consensus amongst the arbiters in terms of interpretation - precisely because an override has occurred. This is really what Duff is getting at, I think.

jase
09-04-2005, 06:03 PM
Any constructive debate, as this thread was before Duff started repeating himself, is interesting. You whingeing is occasionally, err, entertaining?

Generally speaking, when pairing you're not aware of who the players are - just their numbers and their stats [eg recent floats, past opponents].
To extend your logic, an arbiter without a computer pairing program at his disposal is left "open to allegations of bias".

As for an "equal deal on colour", the tournaments you play in have uneven number of rounds. You do the math. Are you implying that you always get more Blacks than Whites? Or is that you don't get White in the last round? or something else? Your phrasing is imprecise.

Whilst the mechanisms for manually overriding pairings might be somewhat laborious to detail to an expectant crowd of 200, I don't think think it takes much to explain under what circumstances the practice would be employed, and why. Do players want this explained to them at the commencement of a tournament?

Without going through the logical process of pairing a round manually, I find it quite difficult to spot every difference between the Swiss pairing procedures that I was taught, and the Dutch system. When the program pairs the next round, you need to open other windows to pursue colour histories and floaters. Alternatively you can print out this information.

One bone I will throw you, in line with your "whingeing": were the arbiters at Doeberl doing these verifications for every round, and every board? Are they checking for previous floats for pairings that involve players on different scores? I have a speculation about where the majority of manual overrides occur for certain arbiters, but I might be advised to run a couple of simulations first ... :hmm:

Rhubarb
09-04-2005, 06:57 PM
Any constructive debate, as this thread was before Duff started repeating himself, is interesting. You whingeing is occasionally, err, entertaining?You can be sure I won't do it again. It's too easy to be misunderstood.

Generally speaking, when pairing you're not aware of who the players are - just their numbers and their stats [eg recent floats, past opponents].
To extend your logic, an arbiter without a computer pairing program at his disposal is left "open to allegations of bias".Well of course they are, but correct me if I'm wrong, you stated that we should use both the technology available and the arbiter's discretion. Starter (and Duff) have asked as a possible solution why not completely do away with human intervention, but as far as I can tell not one single arbiter participating in this thread has agreed with them. As such, all arbiters may find themselves open to allegations of bias as far as pairings go, since pairing rule changes will involve manual intervention.


As for an "equal deal on colour", the tournaments you play in have uneven number of rounds.No kidding.

You do the math.It's called maths in this country.

Are you implying that you always get more Blacks than Whites?No.

Or is that you don't get White in the last round?No.

or something else? Your phrasing is imprecise.This thread was split from the Doeberl Cup Arbitering Hypotheticals thread (also in the Arbiters Corner forum). Have you really not read that thread, jase? There's been a hell of a lot of cross-referencing between the two threads, so I made the assumption that you, being an arbiter, had read it and since you were so upset by Duff allegedly repeating himself I didn't want to be accused of doing the same myself. In the first thread, I stated
I'm sick of getting Black. In my last six tournaments I've been "screwed on colour" as Shaun would have it. When I mentioned that the last time I got an even colour spread was in a double round robin (my seventh last tournament), I did so in agreement with your point that this is the only tournament format that doesn't contain inherent imperfections, not that I think all tournaments should be double round robins. If you had read the other thread, of course, you would realise I am merely cursing my luck.


Whilst the mechanisms for manually overriding pairings might be somewhat laborious to detail to an expectant crowd of 200, I don't think think it takes much to explain under what circumstances the practice would be employed, and why. Do players want this explained to them at the commencement of a tournament? [EDIT: Certainly the players would want which pairing rules are going to be changed explained, as was done at Doeberl. A printout posted everywhere would be sufficient for me, perhaps with a couple of relevant examples, with an announcement at the start for people to read the printouts if they're interested.]


Without going through the logical process of pairing a round manually, I find it quite difficult to spot every difference between the Swiss pairing procedures that I was taught, and the Dutch system. When the program pairs the next round, you need to open other windows to pursue colour histories and floaters. Alternatively you can print out this information.
OK, but this is not my problem.


One bone I will throw you, in line with your "whingeing": were the arbiters at Doeberl doing these verifications for every round, and every board? Are they checking for previous floats for pairings that involve players on different scores? I have a speculation about where the majority of manual overrides occur for certain arbiters, but I might be advised to run a couple of simulations first ... :hmm:I have no idea. Not being an arbiter, why on earth would I pay attention to whether the arbiters are doing these verifications for every board and every round? It's only the last round or two I take a more than superficial interest in how the pairings were arrived at. So I might, for example, have a very civil discussion with Shaun after the last-round pairings come out.

ursogr8
09-04-2005, 10:11 PM
<snip> Starter (and Duff) have asked as a possible solution why not completely do away with human intervention, but as far as I can tell not one single arbiter participating in this thread has agreed with them.
<snip>.


hi kegless

Nice posts of yours on this thread...enjoyable reading.
Btw, there have been a few times that I have not had the numbers in an ISSUE on the bb. I think the GURUs marketing puff was the lonelyist.
But of this manual_intervention in SP, it does seem to be simply a values clash between

> the arbiters desire to follow the rules accurately
versus
> the chess promoters desire to have pairings up early without delay and controversy.

regards
starter

Rhubarb
09-04-2005, 11:23 PM
But of this manual_intervention in SP, it does seem to be simply a values clash between
> the arbiters desire to follow the rules accurately
versus
> the chess promoters desire to have pairings up early without delay and controversy.Point taken, starter.

BTW, does Box Hill ever run tournaments with SP but without intervention? When you're not quartiling, intermingling and quarantining, that is.

ursogr8
09-04-2005, 11:47 PM
Point taken, starter.

BTW, does Box Hill ever run tournaments with SP but without intervention? When you're not quartiling, intermingling and quarantining, that is.

hi kegless

There is a subtle sting in your question that is a bit chewy on the back palate. Actually, my liqueur glass tonight has Renmano 1976 Show Reserve Sweet White that is moderating. (Hope this doesn't get me RIP'd).

Now, turning to your question.......... quartiling, intermingling and quarantining all happen before we press the drop-down menu item called pairing...(not to be confused with paring nor parring).
So, no, we don't intervene in that sense.

But, when we get latecomers, like last night, or forget to enter some-one into the player list, like last night, or some go home because they didn't listen to the thrice-announced 5 rounds not 4, like last night,.........then I might fiddle a bit. At a pinch this could be seen as ''''''' intervening.



starter

ps
Bottle #4222 Individually bottled.
Do you want the text on the other side of the bottle?

Rhubarb
10-04-2005, 12:48 AM
There is a subtle sting in your question that is a bit chewy on the back palate.No disrespect intended, I can assure you.

Actually, my liqueur glass tonight has Renmano 1976 Show Reserve Sweet White that is moderating.I must try that myself. ;)

Now, turning to your question.......... quartiling, intermingling and quarantining all happen before we press the drop-down menu item called pairing...Never having used SP before, I wasn't sure this was the case. OK, so after you've made your initial adjustments, you go with SP without overriding.

Bottle #4222 Individually bottled.
Do you want the text on the other side of the bottle?I'm sure I'm missing something now...

Regards,
kegless

jase
10-04-2005, 01:53 AM
correct me if I'm wrong, you stated that we should use both the technology available and the arbiter's discretion. I correct you. You are wrong.

My initial post on this thread outlined the differences between Swiss Perfect's pairing formula and the way that many leading arbiters were taught to do the pairings. I expressed a preference for the 'old' style . I do not believe arbiters who manually override SP are doing so on a wholly consistent basis, for reasons I touched on in my reply to you.

Duff is seeking consistency from all arbiters by having them not manually adjust SP pairings. You have engaged in and appear to be in support of this argument. Yet you choose to dismiss the points I raised concerning consistency with "not my problem" and "why on earth would I pay attention". This has lead you to overlooking what I have inferred.


Starter (and Duff) have asked as a possible solution why not completely do away with human intervention, but as far as I can tell not one single arbiter participating in this thread has agreed with them. As such, all arbiters may find themselves open to allegations of bias as far as pairings go, since pairing rule changes will involve manual intervention.
To the best of my recollection [having not run a tournament for about a year now] I have not manually adjusted SP pairings for a couple of years. The program produces pairings in line with the Dutch system, and that's fine with me. I don't think the Dutch System is as effective as pairing rules I was taught, but I'd prefer one or the other, not a bit of both [what I'd love is a computer pairing program that places more importance on top v bottom].

You then go on contradict yourself concerning last round colour allocations:


Jase: Or is that you don't get White in the last round?
Kegless: No.

I'm sick of getting Black.
Yes, I was aware you were cursing your luck. But perhaps before these 6 tournaments you quote you were on a hot streak of last round Whites? The two crosstables of old State Championships I was able to google today both had you + White.

[i]Rosencrantz: The law of averages, if I've got this right, means that if six monkeys were thrown up in the air long enough, they would land on their tails about as often as they would land on their -
Guildenstern: Heads

Greg, I took some time before posting on the Doerberl/Arbitering thread, because
1. I've been unwell for a while, and
2. I wanted to try to ensure any points I made were clear and well written.

I do get the impression you've not grasped much at all of what I've written.


But of this manual_intervention in SP, it does seem to be simply a values clash between

> the arbiters desire to follow the rules accurately
versus
> the chess promoters desire to have pairings up early without delay and controversy.
It might also be the arbiter's desire to follow the rules they were taught .

I think SP is following the Dutch rules quite accurately. However my preference, as outlined earlier on this thread, is for the 'old' system.

Having pairings up early and without controversy is also a desire of the arbiter. And you get crosstables, standings, buchholtz scores, etc etc. If it weren't for these benefits I'd probably choose a pack of index cards and a pen over the Dutch/SP system.

Rhubarb
10-04-2005, 04:20 AM
You know, jase, we probably wouldn't be having this discussion if you'd read this thread from the start in conjunction with the other Doeberl thread (http://www.chesschat.org/showthread.php?t=2202). Since I consider myself to be something of a guest in these arbiting threads, it is perhaps my fault that you erroneously got the impression that I supported Duff's position from an isolated reading of #106 in this thread.


I correct you. You are wrong. So you don't want to take advantage of the technology?


Having pairings up early and without controversy is also a desire of the arbiter. And you get crosstables, standings, buchholtz scores, etc etc. If it weren't for these benefits I'd probably choose a pack of index cards and a pen over the Dutch/SP system.So you do want to take advantage of the technology?
I'm confused, Guildenstern, should we be using SP or not? If yes, please explain to me how I am "wrong".


My initial post on this thread outlined the differences between Swiss Perfect's pairing formula and the way that many leading arbiters were taught to do the pairings. I expressed a preference for the 'old' style . I do not believe arbiters who manually override SP are doing so on a wholly consistent basis, for reasons I touched on in my reply to you. If it is true that arbiters are manually overriding SP in an inconsistent fashion, then this is good news for Duff's case.


Duff is seeking consistency from all arbiters by having them not manually adjust SP pairings. You have engaged in and appear to be in support of this argument.Incorrect, as stated above.


Yet you choose to dismiss the points I raised concerning consistency with "not my problem" and "why on earth would I pay attention". This has lead you to overlooking what I have inferred.Perhaps you should have saved the points for Duff.


To the best of my recollection [having not run a tournament for about a year now] I have not manually adjusted SP pairings for a couple of years. The program produces pairings in line with the Dutch system, and that's fine with me. I don't think the Dutch System is as effective as pairing rules I was taught, but I'd prefer one or the other, not a bit of both [what I'd love is a computer pairing program that places more importance on top v bottom].Okay, thanks for your view.


You then go on contradict yourself concerning last round colour allocations: Huh? The first of your two questions did not apply solely to the last round:

JL: "Are you implying that you always get more Blacks than Whites?"
GC: "No."

Note that I'm answering that I don't "always" get more Blacks. However, I later went on to explain that the last 6 tournaments I have got more Blacks. ("I'm sick of getting Black".)

JL: "Or is that you don't get White in the last round?"
GC: "No."

OK, I didn't want to confuse the issue since you weren't yet recognising my main point, but the last round colour is something of an adjunct to the discontent. It is far worse to start with White and finish with two Blacks (for 7 rounds, WBWBWBB) than it is to stay in sequence BWBWBWB. Conversely, when you get an extra White, it's much better to go BWBWBWW than stay in sequence WBWBWBW (which even so would make a nice change for me). If you can't work out why, then it means you still haven't actually read what I've been saying on the two threads. Note there is nothing for you or any arbiter to address here; it's just my observation that it is most desirable to finish with 2 Whites and least desirable to finish with 2 Blacks.


Yes, I was aware you were cursing your luck. But perhaps before these 6 tournaments you quote you were on a hot streak of last round Whites? The two crosstables of old State Championships I was able to google today both had you + White. This is the part where you start patronising a Pure Mathematics and Statistics graduate. May I draw your attention to this quote (http://www.chesschat.org/showpost.php?p=51776&postcount=33). In particular: "The last thing I said to Shaun when I was discussing this with him was I hoped that it would eventually even out one day. (And no doubt you won't see me whinging when it does. :) )"


[i]Rosencrantz: The law of averages, if I've got this right, means that if six monkeys were thrown up in the air long enough, they would land on their tails about as often as they would land on their -
Guildenstern: Heads Very good. Since you've related this story to me many times, I assume you're doing it for the benefit of the audience.


Greg, I took some time before posting on the Doerberl/Arbitering thread, because
1. I've been unwell for a whileI hope you're feeling better.

2. I wanted to try to ensure any points I made were clear and well written.They were. It's unfortunate you thought that I was attacking them. It's interesting that none of the arbiters besides gg have responded to your points yet.


I do get the impression you've not grasped much at all of what I've written.I get the impresssion you haven't even read an entire thread relevant to my position.

Duff McKagan
10-04-2005, 07:58 AM
I think SP is following the Dutch rules quite accurately. However my preference, as outlined earlier on this thread, is for the 'old' system.

Good one. You do "what ever the hell [you] want" :lol:

SP and SwissMaster,
Old versions, new versions.
Want good pairing faster,
Not opaque perversions.

Sorry Jase, but consistancy is more important than perfection.

As an aside, you are probably right about me taking the discussion a bit off topic by proposing that the ACF should or should have a policy on pairing. But hey, who need solutions when we can just roll along with the "Law of Averages*". :P

*There is no such "law" in stats. There is, however, a random process we call independant selection. The Swiss system is far from random since it contains if/then statments based on a priori data, being the players ratings. :owned:

Cheers :D

ursogr8
10-04-2005, 08:55 AM
Is it just me, or am I getting warped by too much time on the bb,......but I have to confess I am enjoying kegless _ jase almost as much as the thread-game annotations of Kaitlin versus Barry Cox. ;)


starter

Garvinator
10-04-2005, 10:08 AM
I don't think the Dutch System is as effective as pairing rules I was taught, but I'd prefer one or the other, not a bit of both [what I'd love is a computer pairing program that places more importance on top v bottom].
I believe you can get what you are looking for jase. In swiss perfect, just tick the little box in the pairings tab that says ignore colours. I believe you then get the top half v bottom half pairings that you are after. This also then means you have to manually check colour allocation.

Garvinator
10-04-2005, 10:10 AM
but consistancy is more important than perfection.
isnt perfection consistent? there is no point having consistency if the consistency is at least partly wrong :doh:

jase
10-04-2005, 12:17 PM
So you don't want to take advantage of the technology?

So you do want to take advantage of the technology?
I'm confused
I was responding to the topic - one's preference for prioritising colour allocation or top v bottom in the Swiss pairing system. Your attempt to surmise my position was inaccurate; you've not engaged in the aspect of this topic I was addressing.

My questions to you did refer to the last round, having read the thread you refer to. Perhaps I could have made that clearer.


This is the part where you start patronising a Pure Mathematics and Statistics graduate.
Not at all - I'm simply querying your colour history prior to your current run of 6 last round Blacks - it's possible the 'evening out' you're hoping for has already taken place! Of course it's also possible that you'll get 6 more last round Blacks in a row ...

Good one. You do "what ever the hell [you] want"
I've clearly explained what I "do", Duff. Maybe you could read that bit? Again, whilst I have a preference for a different style of pairing system, I apply SP because it follows a FIDE-approved set of pairing rules and offers consistency.

As for quoting the "Law of Averages" quote - that was largely for Greg, because he's well read and would get it. Although having never used that specific quote before in my life, I was curious to read that I've related it to Greg many times before!

I wasn't delving into the study of statistics [again, it was a philosophical point, with a nod and a wink to Mr. Stoppard], but perhaps you shouldn't either, if you are asserting that there's no such thing as a Law of Averages in Statistics.

Spiny Norman
10-04-2005, 03:21 PM
isnt perfection consistent? there is no point having consistency if the consistency is at least partly wrong :doh:

I know where Duff is coming from. I call it the McDonalds syndrome ... you know its going to make you sick when you eat it ... but wherever you are in the world you know exactly how sick its gonna make you, so you eat it anyway.

arosar
11-04-2005, 07:47 PM
Some interesting controversy about manual intervention. You have to sort of scroll down. I just came upon this when researching about the Saintly Cup!

http://www.auschess.org.au/bulletins/acfb19.htm

AR

jase
11-04-2005, 08:15 PM
Amiel,

The pairing controversy you've sourced from the ACF Bulletin archives is precisely the event I was referring to in post #92 on this thread.

The fallout from that incident was that it became clear to many arbiters that Swiss Perfect was not pairing according to their interpretation of the Dutch pairing criteria.

Ultimately what I learnt was
a. Swiss Perfect does a fantastic job of the Dutch pairing rules, and
b. I don't agree with the philosophy of that system, with its heavy weighting twoards colour preference and alternation.

Nice ressearch mate :clap:

Oepty
13-04-2005, 06:00 PM
Amiel,

The pairing controversy you've sourced from the ACF Bulletin archives is precisely the event I was referring to in post #92 on this thread.

The fallout from that incident was that it became clear to many arbiters that Swiss Perfect was not pairing according to their interpretation of the Dutch pairing criteria.

Ultimately what I learnt was
a. Swiss Perfect does a fantastic job of the Dutch pairing rules, and
b. I don't agree with the philosophy of that system, with its heavy weighting twoards colour preference and alternation.

Nice ressearch mate :clap:

Jason. I been following this thread with interest, but also have been very busy so I have not posted for a few days.
What you seem to be saying is that Swiss Perfect follows the Dutch Pairing rules but because you think that the Dutch pairing rules are inferior to the rules of the Lim Pairing system that you initially learnt you intervene into the pairings of Swiss Perfect. This seems to mean you change them to what you think are better, but not necessarily more correct pairings. Is this right? I am sorry if this is wrong.
I know this is difficult but if what I have said above is true then is this also the case with Australia other International Arbiters Cathy Rogers, Charles Zworestine, Gary Bekkar etc?

Scott

arosar
13-04-2005, 08:31 PM
Bill -

The NSWCA Tournament Guidelines will need to be updated because it still refers to GMB Swiss and PROTOS.

AR

Thunderspirit
13-04-2005, 10:31 PM
I've always paired for opponents, rather than colours. 99% of the time colours balance so players get +1 or -1 with their colours which should happen.

As long as organisers don't run 6 round weekenders, which give 4 and 2 or 2 and 4 often than it is not a problem. It all the events I been associated with over the years I've only seen 1 situation where a player got 5 Whites and 2 Blacks in a 7 round event...

pax
14-04-2005, 09:26 AM
In all the events I been associated with over the years I've only seen 1 situation where a player got 5 Whites and 2 Blacks in a 7 round event...

Surely that is a breach of the "absolute" colour rules (just as bad as 3 whites in a row). Still, I guess the player wasn't complaining.

Garvinator
14-04-2005, 09:55 AM
Surely that is a breach of the "absolute" colour rules (just as bad as 3 whites in a row). Still, I guess the player wasn't complaining.
you would think so, but according to the pairings rules on swiss perfect, it is legal.

B2.
a. No player's color difference will become >>+2 or <<-2.
b. No player will receive the same color three times in row.

Note: B2, B5 and B6 do not apply when pairing players with a score of over 50% in the last round.

Bill Gletsos
14-04-2005, 12:15 PM
Be aware that FIDE made an amendment to that last part a while back.

The amendment states The rules B2, B5 and B6 shall not apply in the last round at scoregroups with a score over 50 % if this helps to produce a greater number of pairings in the scoregroup. If the number of pairings would be the same when these rules are used, they shall be used.

Also the following was added in 1999.
Members of the same federation in a scoregroup of over 50 % will be not be paired against each other in the last three rounds of a Swiss tournament in Youth and Junior Championships.

All of the above is listed in section 21 of the FIDE web page http://www.fide.com/official/handbook.asp?level=C0402D

jase
14-04-2005, 01:09 PM
What you seem to be saying is that Swiss Perfect follows the Dutch Pairing rules but because you think that the Dutch pairing rules are inferior to the rules of the Lim Pairing system that you initially learnt you intervene into the pairings of Swiss Perfect. This seems to mean you change them to what you think are better, but not necessarily more correct pairings. Is this right? I am sorry if this is wrong.

Correct. There are quite a few pairing systems; all adhere to the Swiss system, but will produce slightly different outcomes, due to their weighting of matters such as colour and subgroups. Note that I no longer "intervene into the pairings of Swiss Perfect". The Lim Pairings [I am not 100% sure that this is what the system I was taught is called] place greater emphasis on subgroups, when compared to Dutch. Both produce "correct" pairings.


I know this is difficult but if what I have said above is true then is this also the case with Australia other International Arbiters Cathy Rogers, Charles Zworestine, Gary Bekkar etc?

I don't want to speak too much for other arbiters, however essentially I think you are right.

Bill Gletsos
14-04-2005, 02:45 PM
Correct. There are quite a few pairing systems; all adhere to the Swiss system, but will produce slightly different outcomes, due to their weighting of matters such as colour and subgroups. Note that I no longer "intervene into the pairings of Swiss Perfect". The Lim Pairings [I am not 100% sure that this is what the system I was taught is called] place greater emphasis on subgroups, when compared to Dutch. Both produce "correct" pairings.Yes both produce "correct" pairings for their respective systems even though the pairings produced by both systems may well be different. If you follow the Lim rules completely you should get pairings that are correct according to the Lim rules, just like if you follow the Dutch rules you should get pairings that are correct according to the Dutch rules.

The Lim System was designed by Dr. Lim Kok-Ann and has been around prior to 1991. As such if the system you were taught was the FIDE system it is quite likely this was the system you were taught. The Dutch rules have been around since 1992. I believe the LIM pairing rules are at http://www.fide.com/official/handbook.asp?level=C0402
The Dutch are at http://www.fide.com/official/handbook.asp?level=C0401

The Dubov Swiss rules are at http://www.fide.com/official/handbook.asp?level=C0403


I don't want to speak too much for other arbiters, however essentially I think you are right.As FIDE notes on their web site all arbiters following the respective pairing rules should get the same pairings. What this means is that all those using the Lim System should get the same pairings and all those arbiters using the Dutch system should get the same pairings.

It can be reasonably assumed that FIDE will not approve a program that does not generate the one true correct set of pairings. Therefore if a FIDE approved pairing program such as Swiss Master which is pairing based on the Dutch rules is used, it would seem reasonable that the arbiters should not manually interfere with the pairings just because they feel the Dutch rules are inferior in some circumstances.
If thats the case they should use the Lim rules completely from round 1.

Swiss Perfect is not FIDE endorsed and does in fact at times quite clearly contravene the Dutch rules. As such manual intervention on the part of the arbiter would seem a necessity. Unfortunately for the average player it would be difficult to determine if the intervention is due to the program actually contravening the rules or just the arbiters particular interpretation of the rules.

ursogr8
14-04-2005, 03:28 PM
Swiss Perfect is not FIDE endorsed and does in fact at times quite clearly contravene the Dutch rules. As such manual intervention on the part of the arbiter would seem a necessity. Unfortunately for the average player it would be difficult to determine if the intervention is due to the program actually contravening the rules or just the arbiters particular interpretation of the rules.

Bill

I don't want to seem like a pedant (ever), but the phrase I bolded above is crucial to one of the central questions being asked on this thread.

Manual intervention comes at a cost of delays and controversy. The question being asked is what benefit is achieved by manual intervention?
When you say it is a 'necessity' (or your words 'would seem like a necessity') did you have in mind a particular benefit achieved. And I hope you do not just reply 'accuracy' to compliance to Dutch Rules'.

regards
starter

Brian_Jones
14-04-2005, 03:45 PM
Just for the record, the Dutch Pairing rules were around a long time before 1992. For example, Protos was using the Dutch Pairing rules in the 1990 Australian Championship.

BTW. Here is a starter (no pun intended) for 10 (points). Subject: Swiss pairings. "When should rankings be used instead of ratings?"

Bill Gletsos
14-04-2005, 03:57 PM
Bill

I don't want to seem like a pedant (ever), but the phrase I bolded above is crucial to one of the central questions being asked on this thread.

Manual intervention comes at a cost of delays and controversy. The question being asked is what benefit is achieved by manual intervention?
When you say it is a 'necessity' (or your words 'would seem like a necessity') did you have in mind a particular benefit achieved. And I hope you do not just reply 'accuracy' to compliance to Dutch Rules'.Then you hope in vain. ;)
However as far as I am concerned you bolded the wrong part of my quote.
You should have bolded clearly contravene the Dutch rules.
As I said the FIDE site makes it clear there is only one way to do Dutch pairings and that all arbiters should get the same results.
The problem is there is no way to be certain that SP is doing them correctly. If it clearly does them incorrectly the arbiter should intervene. In these cases it wouldnt matter whether you believed ranking or colour to precedence, arbiters of either camp would still say the pairings were incorrect.

It was in these circumstances i said the arbiter should intervene.

Where a problem arises however is if the arbiter just thinks the pairings are wrong based on his interpretation of the rules (which may or may not be correct). In this case the question of whether the pairings should be overridden is more problematic.

What we need is a program with the ease of use and user friendliness of SP but that does the pairings absolutely correctly like the FIDE endorsed programs Swiss Master or Swiss Manager. In that case it would I believe be possible for the ACF for its events and for a State Association for their events to inform there appointed Arbiters that they are not to override the program generated pairings.

Bill Gletsos
14-04-2005, 03:58 PM
Just for the record, the Dutch Pairing rules were around a long time before 1992.I believe you will find they were not FIDE endorsed until 1992.

Oepty
14-04-2005, 05:46 PM
Just for the record, the Dutch Pairing rules were around a long time before 1992. For example, Protos was using the Dutch Pairing rules in the 1990 Australian Championship.

BTW. Here is a starter (no pun intended) for 10 (points). Subject: Swiss pairings. "When should rankings be used instead of ratings?"

A players score is of greater importance than their ratings when ranking a players. So if you are ordering a hetrogenous group of players then the players with the highest score most be ordered ahead of players of a higher rating. From memory this is covered in A2 of the Dutch Pairings rules.

Scott

jase
14-04-2005, 05:51 PM
Swiss Perfect is not FIDE endorsed and does in fact at times quite clearly contravene the Dutch rules.

Can you provide evidence of this fact please Bill?


there is no way to be certain that SP is doing them correctly.

Yes there is. Arbiters can run pairings manually and compare their pairings with SP.

You assert that Swiss Perfect contravenes the Dutch pairing rules; an hour later you can't be sure if the program is right or not. This contradiction suggests a little less than fact; please entertain us with some examples.

Perhaps you mean to suggest that as FIDE has not endorsed Swiss Perfect, it must not be pairing correctly? The potential flaw in this logic is that endorsements by FIDE are frequently linked with the appropriate financial arrangements, not strictly bound to the quality of a product.

Denis_Jessop
14-04-2005, 05:57 PM
A players score is of greater importance than their ratings when ranking a players. So if you are ordering a hetrogenous group of players then the players with the highest score most be ordered ahead of players of a higher rating. From memory this is covered in A2 of the Dutch Pairings rules.

Scott

This is an interesting one. A2 provides that players are "ranked" in a certain order but E4 re colour allocation provides that the colour preference of the "higher ranked player" is to be granted (if you get to step E4) and "ranked" here apparently has a meaning different from that in A2.

DJ

Oepty
14-04-2005, 05:58 PM
Correct. There are quite a few pairing systems; all adhere to the Swiss system, but will produce slightly different outcomes, due to their weighting of matters such as colour and subgroups. Note that I no longer "intervene into the pairings of Swiss Perfect". The Lim Pairings [I am not 100% sure that this is what the system I was taught is called] place greater emphasis on subgroups, when compared to Dutch. Both produce "correct" pairings.


Well there are only 4 FIDE endorsed ratings systems at the moment, so unless it one that has been disendorsed I think it would have been the Lim pairing system.
In my view it is good to hear you no longer intervene in the pairings of Swiss Perfect, sorry I suggested you did.




I don't want to speak too much for other arbiters, however essentially I think you are right.

More generally, I now think, and it a very different view to what I held at the start of this discussion, arbiters should not interfer in the pairings of SP unless they can prove beyond doubt that they contrevene the Dutch pairing rules. If an arbiter or a tournament organiser wishes to use a different swiss system because they sholuld find a computer program that does that pairing system accurately and use that, or do the pairngs manually. They must in my view make it absolutely clear before the start of the tournament, and perhaps even on entry forms what system is going to be used.

Scott

Bill Gletsos
14-04-2005, 06:57 PM
Well there are only 4 FIDE endorsed ratings systems at the moment,I think you mean pairing systems, not rating systems.

Bill Gletsos
14-04-2005, 08:02 PM
Can you provide evidence of this fact please Bill?
I have seen it do it. The case in question is where there are few players and many rounds.
One specific case was a few years back in a world computer championship where they were using SP. It got somthing like rounds 1-5 ok, round 6 was completely stuffed, round 7 was ok and round 8 was stuffed. The arbiters ended up overriding it for those rounds. It was a topic of major discussion at the time in the Computer Chess Club forum.
I duplicated the event and sent the data to the author of SP. He informed me that there was indeed a problem and he could rectify it. Unfortunately a corrected version has never been posted on the SP web site.

Yes there is. Arbiters can run pairings manually and compare their pairings with SP.That is no good as clearly there are arbiters in Australia who interpret the Dutch rules one way and those that interpret them another.

You assert that Swiss Perfect contravenes the Dutch pairing rules;No I asserted there are times when it contravenes them. I gave an actual situation above.


an hour later you can't be sure if the program is right or not. This contradiction suggests a little less than fact; please entertain us with some examples.You misunderstood me. What I'm saying is that at times it will absolutely get the draw wrong and all arbiters irrepsective of how they interpret the Dutch rules would note that it is in error.
However when I said "you cannot be certain if SP has it right or wrong" I meant this from the aspect of two particular arbiters. Arbiter A may believe Sp has it right based on how he interprets the Dutch Rules, whilst Arbiter B thinks its wrong based on how he interpret the Dutch rules.


Perhaps you mean to suggest that as FIDE has not endorsed Swiss Perfect, it must not be pairing correctly?No I didnt mean that at all. Most of the time it seems ok but occasionally it just plain and simple gets the draw wrong.


The potential flaw in this logic is that endorsements by FIDE are frequently linked with the appropriate financial arrangements, not strictly bound to the quality of a product.Perhaps you could actually follow your own question to me at the top of your post and provide some sort of factual evidence to back this claim up specifically in regards to pairing programs.

Spiny Norman
14-04-2005, 08:24 PM
I have seen it do it. The case in question is where there are few players and many rounds.

We recently ran a small tournament ... started with only 10 participants, and 2 withdrew after the first round, leaving only 8 players ... five (5) rounds. The final round pairings looked a bit odd to me (pairing a strong player on 3/4 against someone on 0/4):

http://www.croydonchess.com/results.htm

but since I don't know what I'm supposed to be looking at I just shrugged my shoulders. Given the small numbers mentioned above, could anyone be bothered to look at the pairings round by round and give an opinion on whether SP got them right or not?

jase
15-04-2005, 12:37 AM
Frosty,

This Croydon tournament is terrific!! I've had a lot of fun with that last round of yours.

I paired this tournament manually and arrived at precisely the same pairings as the actual tournament, with one final round exception: on board 1 I paired Waller [3] with Whitham [2], rather than SP's pairing of S.Frost [2] -Waller [3]. The reason for the difference is that I was observing the Dutch rules' insistence that players not be upfloated more than once within two rounds [S.Frost had upfloated the previous round]. I cannot ignore the float rules because S.Frost is not on a scoregroup of "more than 50%".

I did consider applying a subclause of C.10, which states: If no alternative pairing for this player exists then drop criterion B6 first and then B5 for upfloats and restart at C2.

This would have produced identical last round pairings to SP; however C.10 relates to homogenous groups, whereas the bracket in question is heterogeneous.

I'm going to give some more thought to why SP permitted S.Frost to upfloat twice consecutively - perhaps I've overlooked something.

There's no problem with LeRoy [3] playing Wales [0] - in what became an 8 player 5 round event, that was simply a forced pairing.

Duff McKagan
15-04-2005, 12:42 AM
This is the link for the SP pairing of a computer tourament
http://chessprogramming.org/cccsearch/ccc.php?find_thread=125768

I like the comment "Anyway, better use a software (even if it is not "perfect" :-) :-) than letting Mr Van den Herik do the pairings by hand!"

All the problems only appear to be occurring with very small fields and a large number of rounds, in final stages of the tournament. As far as I can tell, SP pairing of a field greater than 30 with an odd number of rounds should be pretty darn good. So, what is all the fuss about SP getting it wrong, Bill? Surely, the field size of the Doeberl and NSW and Vic opens etc is large enough to trust the program.

I found out that kegless was very unhappy that the last round in the City of Sydney last weekend had yet again been fiddled with. I suspect that arbiters are re-pairing simply because they have the power to do so :eek:

At the risk of thread drift and being told I am trolling I will ask if the reason we are staying with SP, even when Bill Gletsos Man of Logic, faults it is that the ACF ratings officer wont write the script to import tournament results from other better pairing programs is because he doesnt know how to do it.

Garvinator
15-04-2005, 01:35 AM
Would using the uscf option for how to pair players in sp be of any help here instead of using fide style. I believe the sp uscf option has a bigger rating 'bias' than the fide style. Can anyone clarify this?

Spiny Norman
15-04-2005, 10:37 AM
This Croydon tournament is terrific!! I've had a lot of fun with that last round of yours.

I paired this tournament manually and arrived at precisely the same pairings as the actual tournament, with one final round exception: on board 1 I paired Waller [3] with Whitham [2], rather than SP's pairing of S.Frost [2] -Waller [3]. The reason for the difference is that I was observing the Dutch rules' insistence that players not be upfloated more than once within two rounds [S.Frost had upfloated the previous round]. I cannot ignore the float rules because S.Frost is not on a scoregroup of "more than 50%".

jase, can you (or someone) point me to some information that explains what an upfloat is? (pardon my ignorance, but this is all new to me!). if i were guessing i would think that 'upfloat' refers to playing an opponent in a higher scoregroup than one's own. is that right?

secondly, who would you have paired me with in the final round as a matter of interest?

Spiny Norman
15-04-2005, 10:44 AM
just an afterthought .... maybe SP is expressing a preference for pairing rated players against each other? .... maybe its taking into account rating performance in some way (as I had performed a bit above my rating i think, at least before the final round was over, as i'd drawn against two moderately stronger players, beaten an unrated player who ended up on 2pts, and lost the first round to someone ~350pts higher than moi)?

Garvinator
15-04-2005, 10:51 AM
(pardon my ignorance, but this is all new to me!). if i were guessing i would think that 'upfloat' refers to playing an opponent in a higher scoregroup than one's own. is that right?
correct and you cant float in the same direction ie either up or down, two rounds in a row.

Bill Gletsos
15-04-2005, 11:04 AM
This is the link for the SP pairing of a computer tourament
http://chessprogramming.org/cccsearch/ccc.php?find_thread=125768

I like the comment "Anyway, better use a software (even if it is not "perfect" :-) :-) than letting Mr Van den Herik do the pairings by hand!"

All the problems only appear to be occurring with very small fields and a large number of rounds, in final stages of the tournament. As far as I can tell, SP pairing of a field greater than 30 with an odd number of rounds should be pretty darn good. So, what is all the fuss about SP getting it wrong, Bill? Surely, the field size of the Doeberl and NSW and Vic opens etc is large enough to trust the program.Its not only in small fields. As I understand it SP does not under some circumstances handle the floater correctly according to the Dutch rules. I beleieve KB can provide more details as he has experienced it.


I found out that kegless was very unhappy that the last round in the City of Sydney last weekend had yet again been fiddled with. I suspect that arbiters are re-pairing simply because they have the power to do so :eek:I certainly dont believe they chnage the draw on a whim just because they can. They intervene because they believe the pairings to be incorrect according to their interpretation of the Dutch rules.


At the risk of thread drift and being told I am trolling I will ask if the reason we are staying with SP, even when Bill Gletsos Man of Logic, faults it is that the ACF ratings officer wont write the script to import tournament results from other better pairing programs is because he doesnt know how to do it.If you believe that then you are being a goose. ;)
If the ACF decides to officially endorse another pairing program then Graham Saint and I will ensure its files can be directly read by the rating system.

Bill Gletsos
15-04-2005, 11:08 AM
Would using the uscf option for how to pair players in sp be of any help here instead of using fide style. I believe the sp uscf option has a bigger rating 'bias' than the fide style. Can anyone clarify this?
We are a member of FIDE and as such run all our events based on the FIDE rules.
We are not a member of the USCF. Why would we want to follow the USCF rules.

Bill Gletsos
15-04-2005, 11:10 AM
just an afterthought .... maybe SP is expressing a preference for pairing rated players against each other? .... maybe its taking into account rating performance in some way (as I had performed a bit above my rating i think, at least before the final round was over, as i'd drawn against two moderately stronger players, beaten an unrated player who ended up on 2pts, and lost the first round to someone ~350pts higher than moi)?I dont believe that is the case at all, because if it were to be doing that then that would be a clear example of it not following the Dutch rules.

pax
15-04-2005, 11:48 AM
At the risk of thread drift and being told I am trolling I will ask if the reason we are staying with SP, even when Bill Gletsos Man of Logic, faults it is that the ACF ratings officer wont write the script to import tournament results from other better pairing programs is because he doesnt know how to do it.

I don't know that any of the other programs are demonstrably better than SP. Just because some French guys says his program is perfect doesn't mean anything.

If any examples of demonstrable breaches of FIDE pairing rules are brought to Robert Rozycki's attention, he will fix them (or more likely point out how you have misunderstood the Dutch rules).

Bill Gletsos
15-04-2005, 12:37 PM
I don't know that any of the other programs are demonstrably better than SP. Just because some French guys says his program is perfect doesn't mean anything.

If any examples of demonstrable breaches of FIDE pairing rules are brought to Robert Rozycki's attention, he will fix them (or more likely point out how you have misunderstood the Dutch rules).The lastest version of SP was created in April 2000. The problem highlighted in the world micro event was in August 2000. Robert did not fix it and as far as I am aware Robert has made it clear that he has no intention of making further versions of SP available.

Spiny Norman
15-04-2005, 12:56 PM
The lastest version of SP was created in April 2000. The problem highlighted in the world micro event was in August 2000. Robert did not fix it and as far as I am aware Robert has made it clear that he has no intention of making further versions of SP available.

Perhaps he could be convinced to make it available 'open source' then?

jase
15-04-2005, 12:58 PM
secondly, who would you have paired me with in the final round as a matter of interest?

I paired you against Jared. Goes a little something like this:

Two players lead on 3/4 - played each other, so I move them down to the next scoregroup, where there's Stones on 2.5. All three have played each other, so all three are moved down to form a 5 player scoregroup with the two players on 2/4. Can't pair you with Waller because of the upfloat rule, but Whitham - Waller is fine. LeRoy - S.Frost is also good, however this leaves Stones floating down, which he did last round. Even dropping this clause, there are no valid pairings for the remaining 4 players.

My reading of C.10 is that this only applies to homogeneous groups, of which this is not, therefore to make valid pairings I need to move these 5 players down to the NEXT score bracket [joining them to J Frost]. Now at this point I'm already aware that at the bottom, there'll be no possible pairings.

C.13 In case of the lowest score bracket: the pairing of the penultimate score bracket is undone. Try to find another pairing in the penultimate score bracket which will allow a pairing in the lowest score bracket. If in the penultimate score bracket p becomes zero (i.e. no pairing can be found which will allow a correct pairing for the lowest score bracket) then the two lowest score brackets are joined into a new lowest score bracket. Because now another score bracket is the penultimate one C13 can be repeated until an acceptable pairing is obtained

Effectively we get one scorebracket of 8 players. Top [Waller, LeRoy, Stones, S.Frost] v Bottom [Whitham, Smith, J.Frost, Wales]. Balancing colour and avoiding repeat pairings, they came out as:

Whitham - Waller
LeRoy - Wales
Smith - Stones
S.Frost - J.Frost

Note: if I were able to apply C.10 regarding downfloats to heterogeneous groups, then I would have arrived at the same pairings as SP.

Spiny Norman
15-04-2005, 01:00 PM
jase, thanks, i hadn't expected such a comprehensive response .... that helps me understand things a whole lot better!

Garvinator
15-04-2005, 02:08 PM
Goes a little something like this: snipped
i wanna see you try and work all that out with the normal situation of 30 players clammering around the arbiters table asking repeatedly where are the next round pairings and who do i play ;) :owned: :whistle:

Denis_Jessop
15-04-2005, 03:22 PM
Perhaps he could be convinced to make it available 'open source' then?

I remember when Graeme Gardiner was ACF President he told the Council that Robert R was not prepared to up-date SP or to allow anyone else to do so.

DJ

jase
15-04-2005, 04:09 PM
i wanna see you try and work all that out with the normal situation of 30 players clammering around the arbiters table asking repeatedly where are the next round pairings and who do i play

I take your point - I've paired the Australian Lightning Championships manually a couple of times, and pumping out accurate draws for 70-80 players every 10 minutes is a scramble. However players loitering around the arbiter's table never bothered me too much; for the curious I sometimes explained what I was doing aloud so they'd learn something, and so they'd conclude that I knew what I was doing.

As a generalisation, lower rated players often took a genuine interest in how pairings were made; whilst those on the top boards couldn't care less how you arrived at the pairings, so long as the outcome was favourable to them.

Spiny Norman
15-04-2005, 04:29 PM
I remember when Graeme Gardiner was ACF President he told the Council that Robert R was not prepared to up-date SP or to allow anyone else to do so.

Does anyone have Robert's contact details? FWIW I may as well ask him whether his views have subsequently changed ...

arosar
15-04-2005, 04:45 PM
See here: http://www.swissperfect.com/

AR

Oepty
15-04-2005, 06:00 PM
Dennis can you please explain further what you meant in your post. What is the alternative ranking? In discussions I have had with arbiters the score of the players has to be taken into account when determining colour as per E4.

Although Swiss Perfect is not being updated there is a new program called League Watch available which does chess tournaments. Maybe the problem Bill found has been fixed in this although not in Swiss Perfect.

Scott

Bill Gletsos
15-04-2005, 08:40 PM
Does anyone have Robert's contact details? FWIW I may as well ask him whether his views have subsequently changed ...My understanding is that they havent.

Denis_Jessop
16-04-2005, 08:13 PM
Dennis can you please explain further what you meant in your post. What is the alternative ranking? In discussions I have had with arbiters the score of the players has to be taken into account when determining colour as per E4.

Although Swiss Perfect is not being updated there is a new program called League Watch available which does chess tournaments. Maybe the problem Bill found has been fixed in this although not in Swiss Perfect.

Scott

Hi Scott

You are right. The ranking concept is the same in both A2 and E4. That is, under E4, if the players have equal scores, the colour preference of the higher rated player is granted but if they have different scores that of the player with the higher score is granted because A2 ranks score above rating.

I don't know about League Watch. I downloaded a copy last year but it expired before I got around to testing it. I am interested because of the problem I had with SP last year which is, I think, a different one from Bill's. I'll do a bit more research on that.

DJ

Garvinator
16-04-2005, 08:22 PM
I have downloaded League Watch and am playing around with it. If anyone wants me to try something specific, feel free to send me some information.

Jason, how do i access the croydon tournament to see if League Watch gets the same pairings you do.

I just wish I could import sp files into League Watch. ;)

Bill Gletsos
16-04-2005, 10:03 PM
I have downloaded League Watch and am playing around with it. If anyone wants me to try something specific, feel free to send me some information.

Jason, how do i access the croydon tournament to see if League Watch gets the same pairings you do.

I just wish I could import sp files into League Watch. ;)As Frosty mentioned earlier the Croydon tournament is at http://www.croydonchess.com/results.htm

Btw with Leaguewatch I am yet to find a way to show the players id number in any view/list.

Garvinator
17-04-2005, 12:17 AM
Ok by accident and chance i think i might have found an example of a massive differential between pairings using sp and pairings done using sp with the ignore colours option ticked. The difference is dramatic and I believe sp has ignored the score part of the criteria for pairing, never mind ratings.

The first four rounds of the tournament:

Round 1

No Name Loc Total Result Name Loc Total

1 Robinson, Andrew J (1) 1596 [0] 1:0 Gray, Garvin (10) 1188 [0]
2 Hiller, Peter (11) 1159 [0] -:+ Kirkman, Phillip (2) 1552 [0]
3 Howes, T (3) 1527 [0] 1:0 Chadwick, Marty (12) 1048 [0]
4 Wilson, Stuart (13) 882 [0] 0:1 Quaresmini, Harvey A (4) 1525 [0]
5 Adams, Ray (5) 1468 [0] 1:0 Hiller, James (14) 619 [0]
6 De Vere, Cameron (15) [0] -:+ Boross, Gabriel (6) 1418 [0]
7 Jenkins, Deborah (7) 1360 [0] 1:0 Martin, Shane (16) [0]
8 Taylor, Mark (17) [0] 0:1 Willcock, Brian (9) 1246 [0]

Round 2


No Name Loc Total Result Name Loc Total

1 Boross, Gabriel (6) 1418 [1] 1:0 Robinson, Andrew J (1) 1596 [1]
2 Kirkman, Phillip (2) 1552 [1] 1:0 Adams, Ray (5) 1468 [1]
3 Willcock, Brian (9) 1246 [1] 0:1 Howes, T (3) 1527 [1]
4 Quaresmini, Harvey A (4) 1525 [1] 0:1 Jenkins, Deborah (7) 1360 [1]
5 Hiller, James (14) 619 [0] -:+ Waters, Mick (8) 1290 [0]
6 Gray, Garvin (10) 1188 [0] 1:0 Wilson, Stuart (13) 882 [0]
7 Martin, Shane (16) [0] +:- Hiller, Peter (11) 1159 [0]
8 Chadwick, Marty (12) 1048 [0] 0:1 Taylor, Mark (17) [0]

Round 3


No Name Loc Total Result Name Loc Total

1 Howes, T (3) 1527 [2] .5:.5 Kirkman, Phillip (2) 1552 [2]
2 Jenkins, Deborah (7) 1360 [2] 1:0 Boross, Gabriel (6) 1418 [2]
3 Robinson, Andrew J (1) 1596 [1] .5:.5 Willcock, Brian (9) 1246 [1]
4 Martin, Shane (16) [1] 0:1 Waters, Mick (8) 1290 [1]
5 Adams, Ray (5) 1468 [1] 1:0 Quaresmini, Harvey A (4) 1525 [1]
6 Taylor, Mark (17) [1] 0:1 Gray, Garvin (10) 1188 [1]
7 Wilson, Stuart (13) 882 [0] 0:1 Chadwick, Marty (12) 1048 [0]
8 Hiller, James (14) 619 [0] .5:0 BYE

Round 4


No Name Loc Total Result Name Loc Total

1 Kirkman, Phillip (2) 1552 [2.5] 1:0 Jenkins, Deborah (7) 1360 [3]
2 Waters, Mick (8) 1290 [2] 0:1 Howes, T (3) 1527 [2.5]
3 Gray, Garvin (10) 1188 [2] 1:0 Adams, Ray (5) 1468 [2]
4 Chadwick, Marty (12) 1048 [1] 0:1 Robinson, Andrew J (1) 1596 [1.5]
5 Willcock, Brian (9) 1246 [1.5] 1:0 Martin, Shane (16) [1]
6 Quaresmini, Harvey A (4) 1525 [1] 1:0 Taylor, Mark (17) [1]
7 Hiller, James (14) 619 [.5] 0:1 Wilson, Stuart (13) 882 [0]
8 Boross, Gabriel (6) 1418 [2] .5:0 BYE

Ok there are the results from rounds 1-4.

Below will show what sp spits out as normal pairings.


No Name Loc Total Result Name Loc Total

1 Gray, Garvin (10) 1188 [3] : Kirkman, Phillip (2) 1552 [3.5]
2 Howes, T (3) 1527 [3.5] : Jenkins, Deborah (7) 1360 [3]
3 Boross, Gabriel (6) 1418 [2.5] : Willcock, Brian (9) 1246 [2.5]
4 Robinson, Andrew J (1) 1596 [2.5] : Hiller, James (14) 619 [.5]
5 Martin, Shane (16) [1] : Quaresmini, Harvey A (4) 1525 [2]
6 Adams, Ray (5) 1468 [2] : Chadwick, Marty (12) 1048 [1]
7 Wilson, Stuart (13) 882 [1] : Waters, Mick (8) 1290 [2]
8 Taylor, Mark (17) [1] 1:0 BYE


and now the pairings as soon as you tick the ignore colours box.


No Name Loc Total Result Name Loc Total

1 Gray, Garvin (10) 1188 [3] : Kirkman, Phillip (2) 1552 [3.5]
2 Howes, T (3) 1527 [3.5] : Jenkins, Deborah (7) 1360 [3]
3 Boross, Gabriel (6) 1418 [2.5] : Willcock, Brian (9) 1246 [2.5]
4 Robinson, Andrew J (1) 1596 [2.5] : Quaresmini, Harvey A (4) 1525 [2]
5 Adams, Ray (5) 1468 [2] : Waters, Mick (8) 1290 [2]
6 Martin, Shane (16) [1] : Wilson, Stuart (13) 882 [1]
7 Hiller, James (14) 619 [.5] : Chadwick, Marty (12) 1048 [1]
8 Taylor, Mark (17) [1] 1:0 BYE




I think the revised pairings are more correct. Thoughts of others?

jase
19-04-2005, 02:29 AM
A few remarks about some earlier pairings [and arbitering, possibly] before addressing the final round issues:

In ROUND 3 the top score bracket featured 4 players on 2/2:
1. Kirkman 1552 colour history –W
2. Howes 1527 colour history WB
3. Boross 1418 colour history –W
4. Jenkins 1360 colour history WB

Two desiring White, two Black; 1 v 3 and 2 v 4 doesn’t satisfy all colour preferences, but transposing 3 and 4, we maintain top v bottom and all players get their colour preference:

Jenkins – Kirkman
Howes – Boross.

Curiously the top two players in this score bracket were paired against each other in the actual tournament.

Similar circumstances exist for the next score bracket down, with 8 players on 1 point. I was able to satisfy all colour preferences whilst maintaining top v bottom, whilst the actual pairings produced quite different results.

Two thoughts to explain the discrepancies:
1. Some fiddling by the arbiter concerning round 1 forfeits.
2. SP counting the colours for pairings not actually played.

Lastly, Hiller was only scored with ½ a point for the bye in this round – any reason for not awarding a full point for the bye?

Round 4 has a classic example of colour preference v top v bottom concerning which players on 1 point to float up to the two players on 1.5

I digress; on to the last round you’re asking about …

As a general comment, in my view the pairings that were generated by ignoring colours are more in keeping with the principles of the Swiss System [though not quite the same as saying they are "more correct", a point made earlier on this thread]. The initial/standard SP pairings contrive some significant moving between scoregroups to achieve colour alternations [eg Hiller, in last place on ½/4, floating all the way up to play Robinson on 2.5].

It does take until B.3 in the Dutch pairings to state that players should “ideally” meet opponents on the same score.

Garvin, are you suggesting that for the final round only this “ignore colours” option be implemented? I think there’s plenty of special final round pairing rules in the Dutch system as it stands. Do we need to add more?

As already discussed on this thread, any implementation of the Swiss system contains inherent flaws. Keeping in mind that the objective of any pairing system is to determine the winners, it is significant that both outcomes produced the same pairings on the top three boards.

Garvinator
19-04-2005, 09:28 AM
Garvin, are you suggesting that for the final round only this “ignore colours” option be implemented? I think there’s plenty of special final round pairing rules in the Dutch system as it stands. Do we need to add more?

As already discussed on this thread, any implementation of the Swiss system contains inherent flaws. Keeping in mind that the objective of any pairing system is to determine the winners, it is significant that both outcomes produced the same pairings on the top three boards.
the ignore colours option for final round- i am not suggesting that it be used for the final round only. The round I have questioned is round five out of a seven round tournament. It was just round five I highlighted because of the pairings that were produced using normal sp and then with ignore colours feature ticked.

Also I just thought it was relevant to the discussions on here to give a practical example to ask about :uhoh:

Garvinator
19-04-2005, 09:33 AM
Lastly, Hiller was only scored with ½ a point for the bye in this round – any reason for not awarding a full point for the bye? requested bye I would imagine. Boross took a half point bye in round 4 also.

Rhubarb
19-04-2005, 12:56 PM
I found out that kegless was very unhappy that the last round in the City of Sydney last weekend had yet again been fiddled with.

Duff, since I haven't spoken to you since the Doeberl Cup, how did you get that impression?

I mean, it's interesting that you're so knowledgeable about my supposed discontent with the last round C of S pairings, when I myself, to this day, don't know what the pairings were changed from.

jase
19-04-2005, 06:25 PM
requested bye I would imagine. Boross took a half point bye in round 4 also.

Yeah possibly. Aroused my curiosity because this player was going to get the bye, requested or not.

Garvinator
19-04-2005, 06:53 PM
Yeah possibly. Aroused my curiosity because this player was going to get the bye, requested or not.
I made the comment about the bye, not kegless. :whistle:

Duff McKagan
20-04-2005, 12:43 AM
I mean, it's interesting that you're so knowledgeable about my supposed discontent with the last round C of S pairings, when I myself, to this day, don't know what the pairings were changed from.

:D Perhaps if you knew what had been changed, would have been discontent. But GC, perhaps you will never know what should have happened ;) Anyways, I am sure your run of blacks will run dry soon enough and the whites for the people you're equal with will run out too.

Maybe see you at the NSW Open.

Cheers

Bill Gletsos
20-04-2005, 03:08 AM
Duff, since I haven't spoken to you since the Doeberl Cup, how did you get that impression?

I mean, it's interesting that you're so knowledgeable about my supposed discontent with the last round C of S pairings, when I myself, to this day, don't know what the pairings were changed from.I used the regression test option of SP to show me what the actual pairings were in the COS and what SP's proposed pairings were.
According to that output in no round would your pairing have changed.

In rounds 4, 5, 6 and 8 all the proposed pairings match the actual pairings for all players.
In round 9 SP would have paired Kordahi V Edgar Bautista and Navarro V Capilitan as opposed to the actual pairings of Bautista V Capilitan and Kordahi V Navarro.

Rhubarb
20-04-2005, 01:56 PM
Anyways, I am sure your run of blacks will run dry soon enough and the whites for the people you're equal with will run out too.Not sure I'd know what to do without my favourite bugbear. :wink:


Maybe see you at the NSW Open.Yep, I'm almost a certainty.

btw Duff, aren't you a Hawks fan?


I used the regression test option of SP to show me what the actual pairings were in the COS and what SP's proposed pairings were.
According to that output in no round would your pairing have changed.Okay thanks Bill (not that I actually claimed any of my pairings had changed).

Bill Gletsos
20-04-2005, 06:38 PM
Okay thanks Bill (not that I actually claimed any of my pairings had changed).I know.
I just thought it was worthwhile to mention it.

Sutek
26-04-2005, 04:35 PM
Yes, the last time I copped even an equal deal on colour was a double round robin, last year's NSW Grade Matches.
Would you rather have me continue whingeing, jase? :cool:

Hi Kegless,

If you think it’s bad getting an extra black now and then just imagine how Alexander McDonnell must have felt when he played De Labourdonnais in what is considered to be the first unofficial world championship match of 1834 (25 games).

McDonnell had black in game one, game two, game three and game four!!

Now that would give you something to really complain about! :)

Regards
Sutek

antichrist
26-04-2005, 04:48 PM
Was it a hands behind the back job or similar?

Rincewind
26-04-2005, 05:08 PM
If you think it’s bad getting an extra black now and then just imagine how Alexander McDonnell must have felt when he played De Labourdonnais in what is considered to be the first unofficial world championship match of 1834 (25 games).

McDonnell had black in game one, game two, game three and game four!!

Now that would give you something to really complain about! :)

Some interesting history there. The two men were very close in age, La Bourdonnais roughly 6 months McDonnell's senior. They both died young McDonnell in 1835 (age 37) and La Bourdonnais 5 years later. And they are buried next to one another in Kensal Green All Souls' cemetry, London.

I went to the cemetry's website for a picture of the headstones (and confirm the adjacent burial) but alas neither McDonnell nor La Bourdonnais are famous enough for that. Although I did notice someone else in their famous residents page: William Makepeace Thackeray. Which reminds me, SBS will be screening Kubrick's adaptation of Barry Lyndon tomorrow night from 10pm - 1:05am. A must see period piece.

Sutek
28-04-2005, 08:37 AM
Was it a hands behind the back job or similar?

No

Regards
Sutek

antichrist
28-04-2005, 09:33 AM
No

Regards
Sutek

How was it decided then, whether it was overcast or clear?

Sutek
29-04-2005, 12:07 AM
How was it decided then, whether it was overcast or clear?

Nope, colours were not decided "weather" it was overcast or clear.

Regards
Sutek

antichrist
29-04-2005, 04:45 AM
Nope, colours were not decided "weather" it was overcast or clear.

Regards
Sutek

Aaar I have got it, it was based on the previous day's stock market report plus if the number of deaths in the London Times was an odd or even number and who was the first to pass wind!

edit: sorry about that but I will let it run

Or mabye the previous game's result, like in rugby league who kicks off.

Rincewind
29-04-2005, 07:39 AM
Or mabye the previous game's result, like in rugby league who kicks off.

You might be onto something there AC. From memory the first three games of the 1st match were draws. So maybe they changed colours on a decisive result, kept same colours when there was a draw. Is this right Sutek?

antichrist
29-04-2005, 08:20 AM
Hey Baz,
I just thought up a new method, why not one player keeps the same colour till they come up a win with that colour, that'll learn them

Sutek
29-04-2005, 12:34 PM
You might be onto something there AC. From memory the first three games of the 1st match were draws. So maybe they changed colours on a decisive result, kept same colours when there was a draw. Is this right Sutek?

Yes this is correct.
They also didn't use chess clocks or score sheets in those days.
Moves were recorded by the spectators!

Regards
Sutek