PDA

View Full Version : Absences skew the draw



Desmond
10-11-2007, 12:25 AM
http://www.crcchess.com/Results/index.html

In rd 6 (penultimate round) we had a bit of an unusual pairing due to absences in the field. Basically, I just wanted to know if there can be no valid draw made because too many players have met before, what is the procedure? Do players get paired against someone they have played already? Do players who cannot be paired receive a full-point bye (even if this is more than 1 player)? Or some other option?

Basil
10-11-2007, 12:43 AM
Nice question!

Garvinator
10-11-2007, 12:49 AM
Hello Brian,

Before I answer, was this a pairing done manually or by computer? If computer, what pairing program was used. This is so I know roughly what pairing laws were used.

Desmond
10-11-2007, 12:51 AM
Hello Brian,

Before I answer, was this a pairing done manually or by computer? If computer, what pairing program was used. This is so I know roughly what pairing laws were used.
I'm afraid I don't know the answer for sure, but I would assume Swiss Sys was used. No manual adjustments afaik.

Basil
10-11-2007, 12:51 AM
I'd like to see your answer of what you believe is correct before he answers that! :P

Garvinator
10-11-2007, 12:52 AM
Ok, looking at the pairings and before Bill answers ;)

I think the pairings should have been:

Cashman/Hunter
Thomas/Bargo
McClymont/Truscott
Weller/Stokes.

These were worked out quickly, trying to pair just like in a tournament, so I might have missed something, but they seem more sensible than what was paired.

Desmond
10-11-2007, 01:17 AM
Thanks for that Garvin, but what about my original questions?

Garvinator
10-11-2007, 02:18 AM
Thanks for that Garvin, but what about my original questions?
What about your original questions;)


In rd 6 (penultimate round) Looks like final round to me.


Basically, I just wanted to know if there can be no valid draw made because too many players have met before, what is the procedure? Do players get paired against someone they have played already? Do players who cannot be paired receive a full-point bye (even if this is more than 1 player)? Or some other option?
I don't believe there is a strict procedure in this case and so it would be a matter of making new pairings that affect the tournament the least.

So if there were no legal pairings in the linked round, I would have paired the two lowest scoring players together.

Kevin Bonham
10-11-2007, 10:40 AM
Basically, I just wanted to know if there can be no valid draw made because too many players have met before, what is the procedure? Do players get paired against someone they have played already? Do players who cannot be paired receive a full-point bye (even if this is more than 1 player)? Or some other option?

There actually is no specified procedure in the pairing rules because it is assumed the tournament will be designed such that this will never happen.

B1(a) Two players shall not meet more than once.

and

B1(b) A player who has received a point without playing, either through a bye or due to an opponent not appearing in time, shall not receive a bye.

have no specified override, so it is up to the arbiter's discretion.

Garvin's suggestion of shuffling the damage to the lowest possible board is a good one and is consistent with the principles of the system.

I had a rapid junior muckaround tournament with about ten players once where some had entered late and some had to leave early, and SP told me that it could not pair round five. I suspected that there was a valid legal pairing but I did not have time to look for it so I simply declared the tournament finished as of round four and suggested everyone have a practice game instead of the final round.

I have just noticed something interesting regarding SP and player absences. Not sure how other programs deal with this. If a player is known to be absent in advance for a round and given zero, there is nothing in the rules I can see that prevents this player subsequently being given a bye. But SP can only code an absence as a bye (whatever its value) and therefore will not permit that player to receive any further byes.

Example - it is known that player Z will miss round one and player Z is therefore not paired for that round and scores zero. Every other game is drawn. In round 2, player Z should now receive the bye, but SP won't allow it. Indeed I have just tested this and found that SP will never give Z the bye, not even eight rounds later.

A hack around this is to not include Z in the list of players until round 2 starts, but in the case of an absence after round 1 this is not possible. An alternative would be to enter a fictional opponent for Z for that round only.

Denis_Jessop
10-11-2007, 02:36 PM
I have just noticed something interesting regarding SP and player absences. Not sure how other programs deal with this. If a player is known to be absent in advance for a round and given zero, there is nothing in the rules I can see that prevents this player subsequently being given a bye. But SP can only code an absence as a bye (whatever its value) and therefore will not permit that player to receive any further byes.

We had quite a debate about this kind of situation on the Canberra Chess e-mail list a few years ago when the argument was about whether a half-point bye prevented a player from later getting a full-point bye. It's common practice in Canberra (and, I think elsewhere,) to allow a player a "half-point bye" upon prior notice in which case the player is given half a point and left out of the draw. This debate happened before Swiss Perfect was being used and I argued that such a person could later be allowed a bye as the half-point bye was not really a bye. Nobody else agreed and, when I later began to use SP, I noticed, as you have, that it treats such a case as a bye and won't allow the player to have another. I didn't bother about it any more as the majority of local players who has expressed a view took that one anyway.

Neither the case of a 0-point nor half-point bye (of the kind we are talking about) seems to be taken into account in the Swiss Rules, perhaps because A5 assumes that a player will receive a full point for a bye and B1(b) is addressing that situation when referring to a bye.

It seems fair enough that a player should not receive two free points in a tournament as provided in the Swiss Rules but if a player is left out of the draw and not given a point, there doesn't seem any reason for that player not receiving a bye later, that is, a bye under the pairing principles. I had argued that if a player is left out of the draw and given a half-point bye, that is not a "bye" as referred to in the Swiss Rules and the same argument would apply to a 0-point bye only more so. Curiously, if B1(b) is applied literally to half-point byes, it could allow a player who has had one, but not more, to have a bye later. In fact, in my view, the situation is just not addressed by the Swiss Rules and SP is avoiding the issue by not allowing at least a 0-point "bye" player to be later given a bye.

DJ

Kevin Bonham
10-11-2007, 06:50 PM
The rules don't allow for a player who has had a half-point bye to later get a full-point bye for a simple reason: the rules don't include the concept of a half-point bye in the first place. A player is either known "in advance not to play in a particular round" hence scoring zero (F5) or else receiving a full point bye (A5). Therefore the situation isn't envisaged and is up to the arbiters' discretion.

I'm assuming that those drafting the rules would have envisaged the situation of a player having one or more zero-point absences and then later getting a full-point bye, and would have put something in to stop that situation if bothered by it. (Admittedly this is a risky assumption given how shockingly drafted the rules are generally). On that basis I see no reason why a player who has had a zero-point absence cannot get a full-point bye, although there are many reasons an arbiter might exercise discretion to avoid it.

Denis_Jessop
10-11-2007, 07:55 PM
The rules don't allow for a player who has had a half-point bye to later get a full-point bye for a simple reason: the rules don't include the concept of a half-point bye in the first place. A player is either known "in advance not to play in a particular round" hence scoring zero (F5) or else receiving a full point bye (A5). Therefore the situation isn't envisaged and is up to the arbiters' discretion.

I'm assuming that those drafting the rules would have envisaged the situation of a player having one or more zero-point absences and then later getting a full-point bye, and would have put something in to stop that situation if bothered by it. (Admittedly this is a risky assumption given how shockingly drafted the rules are generally). On that basis I see no reason why a player who has had a zero-point absence cannot get a full-point bye, although there are many reasons an arbiter might exercise discretion to avoid it.

Kevin

I'm not sure that I fully agree because half-point byes are not unknown elsewhere. Moroever, I note that F5 says "They score 0 unless the controller agrees otherwise" and so the drafters of B1(b) should have foreseen the case of an F5 player being awarded something other than 0, for example half a point. See also Art. 11.1 of the Laws that allows for the possibility of a winner on forfeit receiving a score other than 1.

I note also that the rules allowing awards of scores other than 1 or 0 were introduced after B1(b). Both F5 and Art.11.1 were changed between 1997 and 2001 while B1(b) was in its present from in 1997, in the case of F5, the words in italics if F5 quoted above were added.

Thus we have a case of the drafters in 2001 (or thereabouts) not having followed up all the consequentials when they amended the rules - an elementary drafter's error.

Incidentally, I see that Stewart Reuben, in discussing Swiss Pairing Systems, refers to the practice of giving half-point byes in the first and second editions of his "Chess Organiser's Handbook" (at pp48 and 37 respectively) but I cannot find the equivalent passage in the third edition. Nor does Stewart tie his comments in with the operation of the Dutch System Swiss pairing rules.

DJ

Kevin Bonham
10-11-2007, 08:07 PM
Kevin

I agree absolutely - even to the expected foresight of the drafters. But I also believe that half-point byes are not unknown elsewhere and I note that F5 says "They score 0 unless the controller agrees otherwise"

In what version (or source) of the rules does that wording of F5 appear? I haven't seen that wording before and the current wording on the FIDE website under FIDE Swiss Rules is:


Players who withdraw from the tournament will no longer be paired. Players known in advance not to play in a particular round are not paired in that round and score 0.

Bill Gletsos
10-11-2007, 08:23 PM
The wording that Denis quoted is from the 2nd edition of Reuban's Chess Organisers Handbook.

The wording of F5 in the First edition and the Third edition are the same as what is currently on the FIDE website.

Denis_Jessop
10-11-2007, 08:27 PM
In what version (or source) of the rules does that wording of F5 appear? I haven't seen that wording before and the current wording on the FIDE website under FIDE Swiss Rules is:

Sorry Kevin, I have just done a major redraft of my views that now explains things more clearly, I hope. Meanwhile, I missed your response to my earlier version and Bill's post as well:(

DJ

Kevin Bonham
10-11-2007, 08:29 PM
In the second edition of his Handbook Reuben included a number of passages that were merely proposals and had not been formally adopted; some of them never were.

Denis_Jessop
10-11-2007, 08:41 PM
The wording that Denis quoted is from the 2nd edition of Reuban's Chess Organisers Handbook.

The wording of F5 in the First edition and the Third edition are the same as what is currently on the FIDE website.

Hi Bill

This is becoming more confusing as I have just looked at the FIDE Handbook to find that the extra words if F5 have been deleted, as you say. I also know that you have an up-dated version of Reuben's third edition. My copy of 2005 still contains the added words in F5. This is pure FIDE - is there a really 100% authentic version anywhere or is that now in the Handbook just a provisional version until someone has a different idea?:evil: In any case, Art 11.1 is still the same as changed in 2001, I assume.* Between my getting on to their Swiss Rules and trying to get on the their Laws of Chess the web site has refused to operate any more. Perhaps Kirsan is listening in and is embarrassed. Anyway, I can't see why F5 shouldn't be in line with Art. 11.1.

* Fide is back on the air and Art.11.1 is the same as changed in 2001 so there is now some inconsistency between the Laws and the Swiss Rules unless B1(b) of the latter allows one half-point forfeit in which case SP at least is not up with it.

DJ

Bill Gletsos
10-11-2007, 09:24 PM
Hi Bill

This is becoming more confusing as I have just looked at the FIDE Handbook to find that the extra words if F5 have been deleted, as you say. I also know that you have an up-dated version of Reuben's third edition. My copy of 2005 still contains the added words in F5. This is pure FIDE - is there a really 100% authentic version anywhere or is that now in the Handbook just a provisional version until someone has a different idea?:evil: In any case, Art 11.1 is still the same as changed in 2001, I assume.* Between my getting on to their Swiss Rules and trying to get on the their Laws of Chess the web site has refused to operate any more. Perhaps Kirsan is listening in and is embarrassed. Anyway, I can't see why F5 shouldn't be in line with Art. 11.1.

* Fide is back on the air and Art.11.1 is the same as changed in 2001 so there is now some inconsistency between the Laws and the Swiss Rules unless B1(b) of the latter allows one half-point forfeit in which case SP at least is not up with it.

DJActually Denis I was distracted and mucked it up in my above post and got it wrong.

It should have said:

The wording of F5 in the Second edition and the Third edition are the same and differ from what is currently on the FIDE website.

Sorry for the confusion.

Bereaved
10-11-2007, 10:59 PM
Hello, everyone,

Amongst my other clubs, I am a member of a small metropolitan club, in Frankston. That was not in any way to be seen as a criticism of Frankston merely a context setting definition, which has about 20 members, approx, active in any one event.

However a policy that I have recently changed is that opposed to re-pairing the round after having arrived at the club, I now have begun to institute forfeits, for persons who fail to notify of non-attendance in a timely fashion. Ie ringing me at the club is unacceptable on the day of play, but instead on the night before the bare minimum.

I would like to hear comments from other persons involved with clubs who have less than 25 members for at least 3 years, about what policy you use regarding absenteeism, ie non ability to attend, or non notification of non attendance. Would be interested what those persons thought about absenteeism, or the forfeiting practice.

Also there is the other aspect of the non attendance, leaving out illegal pairings, but instead instituting very odd looking pairings that some may feel slighted by, ie being re-paired with someone higher rated had another high rated not been away.

Take care and God Bless, Macavity

PS disclaimer a little bit tipsy ( some may read smashed!! lol ) whilst writing this. Many typos died in writing this post. Also much editing required, possibly in vain!!

Rincewind
10-11-2007, 11:24 PM
However a policy that I have recently changed is that opposed to re-pairing the round after having arrived at the club, I now have begun to institute forfeits, for persons who fail to notify of non-attendance in a timely fashion. Ie ringing me at the club is unacceptable on the day of play, but instead on the night before the bare minimum.

I would like to hear comments from other persons involved with clubs who have less than 25 members for at least 3 years, about what policy you use regarding absenteeism, ie non ability to attend, or non notification of non attendance. Would be interested what those persons thought about absenteeism, or the forfeiting practice.

Also there is the other aspect of the non attendance, leaving out illegal pairings, but instead instituting very odd looking pairings that some may feel slighted by, ie being re-paired with someone higher rated had another high rated not been away.

At our club (around the same size) general courtesy is that your opponent should be informed before the night if you intend to be absent. As a general rule notifying the DOP is also a good idea.

The general practice in one round per week swiss tournies is for the draw to be prepared and disseminated 2 days before the club night. Generally, those who don't show on the night forfeit.

There are exceptions though. Personally I have been rung by an opponent on my mobile at the scheduled start time of a game and organised a deferral on the game in question, but I think this should not be allowed to become common practice and the I know that guy was trying to play but was labouring under very heavy work commitments at the time. I think if I had insisted on a forfeit it probably would have been enforced given the late notice.

Regardless of the forfeits, we don't repair the round.

ElevatorEscapee
11-11-2007, 08:55 AM
The Club I play at has similar small numbers (a little more than a dozen regulars) and it tends to have more of a laid back culture when it comes to enforcing the letter of the laws.

However, we are more strict when it comes to enforcing forfeiture as a result of absenteeism, or simply showing up an hour after the scheduled starting time.

Players do have the option, if they know in advance that they can't play on the club night in a particular week, to negotiate with their opponent to play their game during the week (eg at one of their houses, or the public library). The opponent has the option to disagree (ie, if he can only play on the club night) then the first player will foreit. Usually, however, the opponent will be mroe than happy to oblige.

In our more "serious" tournaments (eg the Club Championship), the game must therefore be played, and the tournament director notified of the result prior to the completion of that round so pairings for the next round can be determined. (This will usually mean they have to play before the round starts).

In one instance a few months ago, a player couldn't play at the club on the scheduled night as a result of a babysitter cancellation. His opponent was happy to go around to his house on the night to play their game.

Another instance saw two players, who both knew in advance that they were going to be late on the night, negotiate with the DOP a later starting time for their particular game.

A Club with such small numbers can hardly afford to lose members due to disgruntlement. Generally, if both players are happy, and they are not causing inconvenience to anyone else, then there should be some scope to help things run more smoothly. :)

=====================

In response to the original post, I have come accross the 'no valid pairings' issue on a few occassions, generally when running a lightning tournament held in a single evening with, say, 8 or 9 rounds and only a dozen players. Of course, juniors leaving early (withdrawers), and people showing up late (late entrants) helped to complicate matters. Normally the invalid pairing won't crop up until the tournament is nearly over.

If you think it's frustrating having Swiss Perfect return the error message, it's even worse coming accross it when using pairing cards; you wind up trying all sorts of pairing combinations before finally realising!

My solution to this on the night was to create an extra round and have the players who could not possibly be paired (ie having already played one another) sit out that round. This not being too much of an imposition in a lightning tournament. The next round might see a different couple of other players sit out... and so on, until the final round, when the players sitting out are those that have played all their games already.

This solution is a bit of a pain, and probably not something that could be adapted easily to anything other than a tournament with short time limits.

Spiny Norman
12-11-2007, 06:54 AM
Croydon usually has about 12-16 adults in each senior tournament. We treat absenteeism fairly leniently ... usually just a quiet chat to the person in question, telling them that they need to let someone know in advance where possible. Taking a hard line would be destructive to the club. We do permit games to be played offsite if the 2 players agree in advance and get permission from the tournament organiser, but once the next round has started its "too late".

Basil
12-11-2007, 11:10 AM
I'm in favour of the stricter application of regs in all aspects of chess. Shock eh? ;) I've also enjoyed reading the comments about how the other half live!


Taking a hard line would be destructive to the club.
Would it? One or two people being told during the course of the year that they have forfeited and caused inconvenience. And then what? They get the hump and declare they won't be back!? As Tony Blair would say, "Not Bothered!"


We treat absenteeism fairly leniently ... usually just a quiet chat to the person in question, telling them that they need to let someone know in advance where possible.
Does this work? Ie stop them from repeat offending? Because they are open to reason? If so, then perhaps if the rules of engagement were spelled out to them when they joined the club, they could use the same "open to reason" when the rules were applied against them.


However a policy that I have recently changed is that opposed to re-pairing the round after having arrived at the club, I now have begun to institute forfeits, for persons who fail to notify of non-attendance in a timely fashion. Ie ringing me at the club is unacceptable on the day of play, but instead on the night before the bare minimum.I support you in this, FWIW - which I reckon is approximately zip!

ElevatorEscapee
12-11-2007, 09:02 PM
"One or two people" in a small club can mean the difference between it being financially viable or not. Given the choice between a club with a couple of eccentrics or no club at all, I think I know which way I would vote.

Two people lost from a club of a dozen members represents a 1/6th reduction in membership. Larger clubs might be able to afford the luxury of such a reduction in overall percentage of numbers, in a smaller Club, I would prefer to keep those members showing up if keeping them on meant that the Club didn't die.

Also please consider the situation where a country Club is the only one in the immediate vicinity (ie, a few disgruntled members can't simply 'up stumps' and take their bat and ball and move to another club a few kilometres down the road... we have had members driving in from 40-50km away!)

If it comes down to a choice of 'enforcement of the rules on a persnickety level' or 'not having a Club at all', it becomes a rather easy choice if you want to continue to have a Club at which you can play Chess.

Basil
12-11-2007, 10:40 PM
If it comes down to a choice of 'enforcement of the rules on a persnickety level' or 'not having a Club at all', it becomes a rather easy choice if you want to continue to have a Club at which you can play Chess.
Agreed. With your conclusion. Not with the underlying premise. So yes, some club is better than no club. Deeer :P

But - have you missed the point? Is the club really in danger? In my opinion, this is admin and decision making based on being threatened, viz "if you enforce this rule, I'm not coming back". That is no way to have self-respect for yourself or for the club or for the members that do the right thing.

And what about the silent majority? If it is truly an option that a player will dummy-spit and not return because daddy had the audacity to enforce the rules, is it equally a possibility that a player will not come back because daddy didn't enforce the rules? I have witnessed this reverse of affairs first hand.

Chess can be ... so small and inward-looking sometimes.

---------

I will pay handsomely for 'persnickety'. What a lovely surprise in the middle of a post. I will also note that players driving 5,000 light years to the club might actually be less inclined to fits of hubris. Gaaaaarn - call their bluff and let me know how you go - I reckon you'd be surprised. TIP: Get your "not bothered" face on while talking to them else they double-dare ya!

Capablanca-Fan
13-11-2007, 09:57 AM
At Logan, if you're not there by 15 minutes before the start of the round, then you're just not included in the draw, unless you've given advance warning. For more than one round per night, a latecomer is welcome to play in the subsequent rounds.

CameronD
13-11-2007, 10:48 AM
At Logan, if you're not there by 15 minutes before the start of the round, then you're just not included in the draw, unless you've given advance warning. For more than one round per night, a latecomer is welcome to play in the subsequent rounds.

I wish they would use this rule in weekend tournaments... even if its a piad up player. I always make sure I'm at the venue 30-60 minuntes before my game starts out of respect.

Trent Parker
20-11-2007, 04:54 PM
At Campbelltown I'm pretty leanient on non attendance. If i'm at home on the monday and the person rings me I re do the draw. If someone rings me while i'm already on my way to the catholic club I'll do one of two things: 1. give the other person the point and the absentee person gets nothing; Or If I have not paired myself in that round and I have not yet played the player who is going to be at the club, I'll pair myself against that person.

At Campbelltown I like to have a "no full point byes" policy. I'll substitute myself in and out to get rid of any byes.

ElevatorEscapee
20-11-2007, 06:12 PM
Agreed. With your conclusion. Not with the underlying premise. So yes, some club is better than no club. Deeer :P

But - have you missed the point? Is the club really in danger? In my opinion, this is admin and decision making based on being threatened, viz "if you enforce this rule, I'm not coming back". That is no way to have self-respect for yourself or for the club or for the members that do the right thing.

And what about the silent majority? If it is truly an option that a player will dummy-spit and not return because daddy had the audacity to enforce the rules, is it equally a possibility that a player will not come back because daddy didn't enforce the rules? I have witnessed this reverse of affairs first hand.

Chess can be ... so small and inward-looking sometimes.

---------
... [UGD snip]

Hmm, I think we are at cross purposes here. The idea is to make the Club more friendly by being flexible to the individual needs of the members.

As to the loss of one member possibly putting a club in financial difficulties, consider a small club with a $1500 a year turnover. Financially speaking, one full time club member who shows up 50 weeks a year and pays his $2 weekly fee contributes $100 a year in revenue to keep the club running, which is fair chunk of the overall revenue.

In my time at my club, no single member has threatened to leave the club because they were upset over an arbiters' decision, a decision by the board at the general meeting, etc.

However, people do leave from time to time for various legitimate reasons of their own, (eg leaving for a new job elsewhere, graduated from uni and wanting to leave the town, becoming despondent over too many losses, etc).

It is all part and parcel of running a small chess club. Fortunately, there always seem to be a handful of new, keen, regular members coming through to replace those who have moved on.

However, if such a small club were to lose four or five members in one hit over a petty dispute (a quick read of some of the disputes on this forum suggests that chess players may be preternaturally susceptable to over-reactions to minor stimuli in this regard... ;) ) or because of an inflexibility in aplying the rules, then you may see a third of that club's revenue disappear overnight! :)

Basil
20-11-2007, 06:25 PM
The idea is to make the Club more friendly by being flexible to the individual needs of the members.
I got that bit. I can't disagree with the goal. I was questioning whether the cited compliancy was in fact counter-productive or at least returning a neutral result while selling out another principle. I see you have now addressed that issue below.


As to the loss of one member possibly putting a club in financial difficulties, consider a small club with a $1500 a year turnover. Financially speaking, one full time club member who shows up 50 weeks a year and pays his $2 weekly fee contributes $100 a year in revenue to keep the club running, which is fair chunk of the overall revenue.
I got that bit. I was questioning
a) would they really really really go?
b) how many would really really go in the face of such a policy.



In my time at my club, no single member has threatened to leave the club because they were upset over an arbiters' decision, a decision by the board at the general meeting, etc.
Aye!


However, if such a small club were to lose four or five members in one hit over a petty dispute (a quick read of some of the disputes on this forum suggests that chess players may be preternaturally susceptable to over-reactions to minor stimuli in this regard... ;) ) or because of an inflexibility in aplying the rules, then you may see a third of that club's revenue disappear overnight! :)
And if this were the case you would be vindicated (in your reasoning). As it happens you don't need to be, as I am just an observer.

I'd like to clarify that my purpose for posting was to challenge the preconception on a cerebral level, not to try and tell you what should be done. You're at the coal-face. You know. I don't.