PDA

View Full Version : Swiss Pairing issues at Queenstown



Brian_Jones
03-02-2009, 11:05 AM
There were some delays with the pairings in Round 10 at Queenstown. I understand that the delays were related to the pairing programs not producing correct or acceptable pairings and this was caused by the treatment of floats with half point byes.

Would one of the arbiters care to tell everyone what the problem was and how we avoid these problems in future.

From what I know at the moment I am tempted to recommend that we do not allow half point byes in serious FIDE rated swiss touraments such as Doeberl Cup or Sydney International.

Garvinator
03-02-2009, 01:09 PM
There were some delays with the pairings in Round 10 at Queenstown. I understand that the delays were related to the pairing programs not producing correct or acceptable pairings and this was caused by the treatment of floats with half point byes.
I can not comment specifically regarding Queenstown as I was neither there or have any personal knowledge of the situation of the situation(s) Brian describes.

As a general point though, I will make a couple of comments.

I am not sure how much of a difference half point byes would have made to the final round pairings in terms of the float rules for those players over 50 per cent as floats aren't considered for players over 50 per cent.

What could (most likely would) have made a difference for final round pairings is colours. With it being a ten round tournament, most players would have been on a 5/4 or 4/5 colour split, meaning in the final round almost the whole field would have had strong colour preferences.

The only players who would have had 4/4 would have been those who had taken one half point bye. There would be a couple of players of course who would be on an absolute colour preference if they had 5/3 or 3/5, but this is also ignored for players over 50 per cent in terms of +3/-3 colours.

So, the arbiters would need to meet colour preferences for each player in a score group, with each player having a strong colour preference, instead of having almost the whole field having a 4/4 or 5/5 colour preference. If the tournament was 4/4 or 5/5, it does give the arbiters a bit more 'wiggle room'.

Therefore, if there was a delay, I think it was more likely to be caused by colours, rather than by float status.

Anyways, on a related point, even if the delay was caused by float status, I think the same issues of floats could come about by upfloats and downfloats, unless the claim is that the program is faulty, which I believe is a known issue in swiss perfect.

Denis_Jessop
03-02-2009, 02:34 PM
There were some delays with the pairings in Round 10 at Queenstown. I understand that the delays were related to the pairing programs not producing correct or acceptable pairings and this was caused by the treatment of floats with half point byes.

Would one of the arbiters care to tell everyone what the problem was and how we avoid these problems in future.

From what I know at the moment I am tempted to recommend that we do not allow half point byes in serious FIDE rated swiss touraments such as Doeberl Cup or Sydney International.

Brian

I'd like to hear about this from a general interest viewpoint. If they were using Swiss Perfect, there are known problems with floats in later rounds whether there are half point byes or not.

I was interested in your comment "not producing correct or acceptable pairings ". If the Dutch Swiss Rules were used, there is supposed to be only one correct set of pairings for a round and the pairing program should produce it. I'm not sure, though I have an inkling, what might constitute an "acceptable" pairing as distinct from a "correct" one. We have already discussed that concept in the past where some posters seemed to think that a set of pairings could look right and thus be acceptable even though demonstrably in breach of the Dutch pairing rules.

DJ

Brian_Jones
03-02-2009, 03:49 PM
If they were using Swiss Perfect, there are known problems with floats in later rounds whether there are half point byes or not.
DJ

They used both Swiss Manager and Swissperfect. ;)

Kevin Bonham
03-02-2009, 03:57 PM
There were some delays with the pairings in Round 10 at Queenstown. I understand that the delays were related to the pairing programs not producing correct or acceptable pairings and this was caused by the treatment of floats with half point byes.

Can't see why this would be an issue since irrespective of whether half-point byes are treated as floats or not, the last half-point byes were taken in round seven. The Swiss system only takes notice of floats from the previous two rounds so half-point byes taken up to round seven should have been irrelevant for the whole field by round ten.


From what I know at the moment I am tempted to recommend that we do not allow half point byes in serious FIDE rated swiss touraments such as Doeberl Cup or Sydney International.

It is important to consider how your pairing software treats half-point byes in making such a decision. Normally a bye is treated as a downfloat, the reasons for this being (a) that it is like being paired with an extremely weak player (b) that having gained a point without playing, a player does not deserve a downfloat in the next two rounds.

These considerations do not apply to half-point byes and ideally they should not be treated as floats at all.

Ian Rout
03-02-2009, 04:32 PM
It sounds like a case of the tail wagging the dog to not permit half-point byes, if it has been decided that they are otherwise OK, just because the pairing software won't handle it or apparently won't. Has the existence of pairing software led to such serious de-skilling of arbiters? I understood that IAs had to pass an exam including questions on pairing rules.
Normally a bye is treated as a downfloat, the reasons for this being (a) that it is like being paired with an extremely weak player (b) that having gained a point without playing, a player does not deserve a downfloat in the next two rounds.

These considerations do not apply to half-point byes and ideally they should not be treated as floats at all.I would argue the opposite. Half-point byes could be treated the same as either full-point or zero-point byes and I think they have more in common with full-point byes in that the player has advanced their score, just by not as much.

Garvinator
03-02-2009, 04:42 PM
I understood that IAs had to pass an exam including questions on pairing rules.
Nope. A person can get an IA title by being chief or deputy arbiter at four international tournaments, without having to pass an exam or be tested on anything.

So this scenario is possible and indeed does happen from my understanding.

Non title arbiter is deputy arbiter at four international tournaments where playing titles are possible, so IA titles are also possible.

Arbiters use only the pairing programs and do not manually double check pairing program.

Through the four tournaments, there are no incidents and the non title arbiter gets his/her IA title. Supervising arbiter, most likely an IA, signs off on the non titled or FA titled arbiters skill.

The arbiter has not sat a formal exam or has not been questioned on how the pairings from the tournaments were produced.

I think in Australia we are fortunate to have a forum like this, where at least there is discussion of arbiting issues.

I do believe from reading Shaun's blog recently on his pairing question that Fide is looking at introducing a mandatory exam of some sort, but am not sure how the system will all work.

Kevin Bonham
03-02-2009, 06:52 PM
I would argue the opposite. Half-point byes could be treated the same as either full-point or zero-point byes and I think they have more in common with full-point byes in that the player has advanced their score, just by not as much.

I think half-point byes are about equally distinct from either and ideally all three could be treated differently. All other things being equal, in the typical round a player would expect to score 1/2. A full-point bye gives them 1 and an absence gives them 0. So given that a full-point bye is treated as a downfloat, there would be a case for treating a zero-point absence as an upfloat, except that that might encourage deliberate forfeiting.

Bill Gletsos
03-02-2009, 07:08 PM
I think half-point byes are about equally distinct from either and ideally all three could be treated differently. All other things being equal, in the typical round a player would expect to score 1/2. A full-point bye gives them 1 and an absence gives them 0. So given that a full-point bye is treated as a downfloat, there would be a case for treating a zero-point absence as an upfloat, except that that might encourage deliberate forfeiting.My understanding is the following:

1) A 1 point bye is a downfloat as per A5
2) A 0 point bye is not treated as any sort of float.
3) A unplayed win is considered a downfloat. (Added at at the 77th FIDE Congress in Turin in 2006)
4) It was decided at the same FIDE Congress that an unplayed loss is considered to be no float.
5) It was proposed at the same FIDE Congress that a half point bye is not considered as an up or down float.

Spiny Norman
04-02-2009, 06:07 AM
My vote is for PAX and BILL (and any other luminaries present) to team up and produce a new pairing program ... one with the pairing accuracy of a Swiss Master v5.5 (thanks GG!), but with the cost/accessibility/compatibility of a Swiss Perfect. Then we'd all have something that we could standardise on, modify as laws change, and integrate with our ratings processing (e.g. put a function in it that submits the results as a file to the state's ratings manager for verification, downloads up-to-date player ratings off the ACF website in real time when entering player records, etc).

Brian_Jones
04-02-2009, 08:03 AM
It sounds like a case of the tail wagging the dog

But why, in a serious chess tournament, should any player be given a free (unearned) half point without playing a game?

BTW Half point byes are not commonplace in Europe. The practice seems to be growing in popularity in Oceania where we encourage players to come for a holiday.

I suggest we ban half point byes (in the same way we ban short draws) and instead target the opponents rating points! :)

Ian Rout
04-02-2009, 09:18 AM
But why, in a serious chess tournament, should any player be given a free (unearned) half point without playing a game?

A valid point in itself (which I will discuss below), but I was saying that if it had been decided that half-point byes should otherwise be allowed then changing the policy on the basis of shortcomings in the pairing program seems to be a misprioritisation. If it has been decided that an event is too serious for half-point byes anyway then the issue doesn't arise.

As to why half-point byes are allowed, of course they shouldn't in a round robin where the point is that everybody plays the same field. But in a Swiss people play different fields and we are happy to tot up the points and say that the highest scorer beats the second-highest even if they have no opponents at all in common. (Is there such a thing as a "serious" Swiss?) The fact that the fields are different in size as well as composition doesn't inherently change the validity or otherwise of this process. It just creates an issue of how we factor this into the comparisons and how we do the pairings.

That people gain an unearned half-point is only half the story. They also give away a half-point. Since you need to win most of the time to score a place or even a rating prize in a tournament then taking a half-point bye is more likely to be if anything a disadvantage. In any event players on higher/lower scores get harder/easier games so early byes come out in the wash in the same way as early upsets, but without distorting other players' games.

The exception is in the last couple of rounds when a safe half-point may be more valuable than taking the risk of losing, which is why half-point byes are generally not allowed in the later stages.

Adamski
04-02-2009, 09:24 AM
That people gain an unearned half-point is only half the story. They also give away a half-point. Since you need to win most of the time to score a place or even a rating prize in a tournament then taking a half-point bye is more likely to be if anything a disadvantage.This is true. IM George Xie, e.g., took a half point bye at the NSWCA Aus Day weekender and FM Jesse Sales did not. Jesse won the tournamant and George came second equal.

Garvinator
04-02-2009, 10:12 AM
My vote is for PAX and BILL (and any other luminaries present) to team up and produce a new pairing program ... one with the pairing accuracy of a Swiss Manager, but with the cost/accessibility/compatibility of a Swiss Perfect.
From my experience, it is Swiss Master 5.5 that stands up best to pairing scrutiny.

Denis_Jessop
04-02-2009, 02:52 PM
But why, in a serious chess tournament, should any player be given a free (unearned) half point without playing a game?

BTW Half point byes are not commonplace in Europe. The practice seems to be growing in popularity in Oceania where we encourage players to come for a holiday.

I suggest we ban half point byes (in the same way we ban short draws) and instead target the opponents rating points! :)

This half-point bye business is both odd and interesting. I played chess regularly in Canberra and Melbourne from the 1950s to the early 1980s and half-point byes were completely unknown. When I returned to playing in 1996 in Canberra after a long lay-off, I found that they had become de rigeur. Apparently they were justified by their supporters as encouraging club players to participate in one-round-a-week tournaments where they might have to miss a week or two. They seemed to be generally accepted though whether their rationale was fulfilled was never established to my knowledge. I think that the hpbs at Queenstown were ostensibly for a similar purpose - to allow visitors to sight-see rather than play, though, if so, it's a little odd for an event of such significance.

There is a downside to them that I have encountered locally, namely, their possible tactical use to avoid an unwanted opponent especially in the later rounds of an event. For this reason, some ACT chess clubs restrict them to the early rounds.

I don't see any real justification for them though, if it could be proved that they improved club attendance and membership numbers, I can see a reason for using them at club level though not higher.

DJ

Spiny Norman
04-02-2009, 06:02 PM
From my experience, it is Swiss Master 5.5 that stands up best to pairing scrutiny.
Heh heh ... sorry ... Swiss <insert differentiator here> ... if I can, I will edit and fix.

Garvinator
04-02-2009, 06:47 PM
Heh heh ... sorry ... Swiss <insert differentiator here> ... if I can, I will edit and fix.
Just to be clear, they are two different pairing programs:

1) Swiss Manager
2) Swiss Master

Bill Gletsos
05-02-2009, 04:21 PM
Upon looking at the Swiss Perfect files for Queenstown, I notice that in the last round all pairings on all boards were as determined by Swiss Perfect with the exception of boards 11 & 12. Interestingly no manual overides occured in any of the previous rounds.

The only manual intervention was on boards 11 & 12 in the final round.

The computer wanted to pair Chow V Brown on board 11 and West V Levi on board 12.
The games as played were Chow v Levi on board 11 and West v Brown on board 12.

Now prior to this round Brown already had a 9 game IM norm.
Chow had no norm after 9 rounds.
Chows second round opponent had no FIDE rating and his first round opponent only had the lowest FIDE rating of all players he had played of 1934.
As such both of these were uplifted to 2100.
This meant that after 9 rounds the average rating of his opponents was 2317.67. For an IM norm he needed a score after 9 rounds of 6.5 but he only had 6 points.
For a 10 game IM norm it meant that the average rating of his opponents needed to be at least 2302 and his score 7 out of 10 or at least 2341 and an score of 6.5.

If Chow played Brown as the draw dictated then Chow could not get an IM norm even if he won the average rating of his opponents was only 2284.3 as Browns rating was only 1984.
However by manually manipulating the draw and replacing Brown by Levi, the average rating of Chow's opponents was now 2307.8 and hence if Chow beat Levi he would gain a 10 game IM norm.

Chow did in fact beat Levi and thus scored a 10 game IM norm.

Interestingly this manual change also benefited Brown.
After 9 rounds brown had a 9 game IM norm having scored 6 points against an average field of 2340.44.
If Brown had played Chow his average would have dropped to 2333.
By now playing West this took his average to 2341.

A score of 6.5 is enough for a 10 game IM norm against an average field of 2341 whilst against a field rated 2302-2340 the IM norm score is 7 points.

Brown drew with West and thus scored a 10 game norm.

Garvinator
05-02-2009, 04:45 PM
Isn't changing pairings to improve norm chances a massive no-no?

http://www.fide.com/fide/handbook?id=84&view=article


18. Maximising players' opportunities

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. A proposal for varying the normal pairing so as to maximise players' opportunities to fulfil title requirements was not approved in the Graz Congress.


20. Mandatory information to be provided in tournament's reports by arbiters

20.1 While reporting a tournament to FIDE, the Arbiter shall declare which of the official FIDE Swiss Systems was used. (GA '97).

20.2 If another system was used, the Arbiter has to submit the rules of this system for checking by the Swiss Pairing Committee (GA '97)

20.3 Where it can be shown that modifications of the original pairings were made in favour of a player to achieve a norm, a report may be submitted to the Qualification Commissionto initiate disciplinary measures through the Ethics Commission (GA '97)

Of course how much fide cares about any of this is going to actually equal zero.

Kevin Bonham
05-02-2009, 10:19 PM
Sounds like that is something FIDE has in place for cases where a player disadvantaged by a swap to give another player a norm chance kicks up a fuss about it.

Brian_Jones
06-02-2009, 11:19 AM
Upon looking at the Swiss Perfect files for Queenstown, I notice that in the last round all pairings on all boards were as determined by Swiss Perfect with the exception of boards 11 & 12. Interestingly no manual overides occured in any of the previous rounds. .

Yes, but were the SP pairings correct?

What about the Swiss Manager pairings?

Why was the draw not published until after midnight?

Bill Gletsos
06-02-2009, 12:19 PM
Yes, but were the SP pairings correct?I only did a quick check on them but all the ones I checked appeared to be correct.

What about the Swiss Manager pairings?I dont know as I havent checked them yet as I didnt get the Swiss Manager files until after I had checked the SP files.

Why was the draw not published until after midnight?No idea.

Brian_Jones
06-02-2009, 02:59 PM
FA Shaun Press has examined the pairings for the top 5 boards (10 players) going into the final round (round 10) at Queenstown.

He tells me that that the Swiss Manager pairing was correct and the Swissperfect pairing (modified or unmodified) was incorrect.

Once again the human brain has been shown to be susceptible to "plausible" (but wrong) pairings.

David Smerdon should not have been paired with Gawain Jones!

Bill Gletsos
06-02-2009, 03:04 PM
FA Shaun Press has examined the pairings for the top 5 boards (10 players) going into the final round (round 10) at Queenstown.

He tells me that that the Swiss Manager pairing was correct and the Swissperfect pairing (modified or unmodified) was incorrect.He may be right but based on errors he has made in the past I wouldnt put any faith in his ability to do a manual swiss pairing correctly.

David Smerdon should not have been paired with Gawain Jones!On what basis is this claim made.

Garvinator
06-02-2009, 03:25 PM
I am not sure how any of this is really helpful in shining a good light on any good matters at hand.

We already know that swissperfect is an out of date program that has not been updated for ten years, but works ok for general run of the mill tournaments and is probably more consistent than the inconsistency of the opinions of the various arbiters in Australia.

Swissperfect also has the advantage of printing and displaying well, so is easy to read and interpret, which is a defect of swiss master and swiss manager.

We can go around and around examining the pairings from this tournament over and over again and arrive once again at the same conclusions. I am just not sure what good it can do in the long run.

Part of the overall issue is that it is arguable what pairing rules work best in Australian/Oceania conditions, where tournaments with players from 2400 to 200 are common with everything in between.

I am certainly interested in partaking in any kind of pairing examination if there is some kind of long term benefit to be gained by it, but if all that is going be gained by it is just that it makes a few people look silly, then I question the whole exercise and purpose.

Brian_Jones
06-02-2009, 03:31 PM
He may be right but based on errors he has made in the past I wouldnt put any faith in his ability to do a manual swiss pairing correctly.

This is a bit of a nasty comment Bill. :evil:

I think I have more faith in someone that has recently passed the Reuben arbiter exam (and is an active FIDE-accredited arbiter) than I do in someone that is a armchair arbiter and hasn't attempted the maths.

Maybe you should apologise Bill. :evil:

Brian_Jones
06-02-2009, 03:38 PM
I am certainly interested in partaking in any kind of pairing examination if there is some kind of long term benefit to be gained by it, but if all that is going be gained by it is just that it makes a few people look silly, then I question the whole exercise and purpose.

For FIDE rated tournaments, we should used FIDE approved programs.

We should not manually change or override the computer-generated pairings
for any reason whatsoever.

We should publish the pairings to the players as soon as possible after the end of the previous round.

The long term benefit is more professionalism.

Bill Gletsos
06-02-2009, 03:40 PM
This is a bit of a nasty comment Bill. :evil:I call it as I see it.

I think I have more faith in someone that has recently passed the Reuben arbiter exam (and is an active FIDE-approved arbiter) than I do in someone that is a armchair arbiter and hasn't attempted the maths.Feel free to put your faith where you like.
As for passing the exam, so what. He actually got the fairly simple swiss pairing question wrong. :doh: :doh: :doh:

Surely you are not suggesting I didnt attempt the maths.

Maybe you should apologise Bill. :evil:Pass.

Denis_Jessop
06-02-2009, 03:56 PM
An interesting mathematical proposition, or so it seems:

1FA>2IAs. :hmm:

DJ

Brian_Jones
07-02-2009, 11:35 AM
Surely you are not suggesting I didnt attempt the maths.


I certainly do suggest that you haven't done the maths correctly in accordance with the rules! :(

Denis_Jessop
07-02-2009, 11:46 AM
I certainly do suggest that you haven't done the maths correctly in accordance with the rules! :(

Stop being so coy, Brian. Just tell the interested spectators what the "Maths" are and how your/Shaun's contention is said to be the right one.

DJ

Bill Gletsos
07-02-2009, 11:51 AM
I certainly do suggest that you haven't done the maths correctly in accordance with the rules! :(No that isnt what you claimed. :liar:

You clearly said "and hasnt attempted the maths."

Having been caught out, you are now changing your tune.

As for who can do the maths right, I could have the maths wrong however I have shown here on numerous occasions that I can in fact do the maths correctly.

Your mate the FA from PNG may be able to do the maths but so far he hasnt shown he has such skills. :hand:

Denis_Jessop
08-02-2009, 04:57 PM
All quiet on the Western front, I see. When will the SM proponents come out with their calculations?

DJ

Igor_Goldenberg
09-02-2009, 10:37 AM
They used both Swiss Manager and Swissperfect. ;)

That's what caused the problem. There was a very long wait for the arbiters to finish the last round draw. When I asked the arbiters about the reason for the delay, they told me that they use two programs that give different pairing for the first board opponent (against David Smerdon).

According to B. Gletsos Swiss Perfect pairing for the last round is correct.
I am not going to dispute the premise. However, it means the second program gave incorrect pairing.
I am not even sure SP was one of the program used.
However, it would be interesting if someone tried to do a last round pairing using different software(s).

Side note:
I witnessed (and was affected by) 4 manual changes to the pairing.
Three of those changes turned out to be incorrect (e.g. the program was right). On the fourth occasion arbiter forgot to flag the last round. When this option was ticked, silicon made draw matched the human intervention.
These experiences taught me to trust computerised pairing.

Denis_Jessop
09-02-2009, 02:34 PM
As I understand the matter, the top 5 board pairings given by SP and SM were different. The SP pairings were used. Shaun Press/Brian Jones contend that the Swiss Manager pairings were the right ones but have not given any calculations in the application of the Swiss rules to demonstrate why their contention is right. Until they do, we must not only accept the fact that the Swiss Perfect pairings were used but that they were right and the the assertion otherwise is without substance.

One cannot just make a bald assertion that something is not in accordance with the rules without saying why. Not to back up one's assertion's with reasoned argument seriously detracts from one's (or in this case two's :) ) credibility.

DJ

Kevin Bonham
09-02-2009, 02:39 PM
Side note:
I witnessed (and was affected by) 4 manual changes to the pairing.
Three of those changes turned out to be incorrect (e.g. the program was right).

Incorrect overrides are a big (and common) problem. It seems that no matter how bad the pairing program is there will always be experienced DOPs who think it is wrong in a way in which it is not.

My own experience with SP is that it should only be overridden when it produces a draw that is clearly totally wrong (downfloating a player two score groups unnecessarily, for instance). When an SP draw appears to be wrong in a more subtle way, it often turns out to be right.

Brian_Jones
09-02-2009, 02:55 PM
One cannot just make a bald assertion that something is not in accordance with the rules without saying why. DJ

Yes, we can. :hand:

We have taken the discussion private.

There is a tendency for people to be rude on chesschat.

Garvinator
09-02-2009, 03:05 PM
We have taken the discussion private.
Shame. I was hoping to actually see some meaningful information on the situation, which unless I am included in the private conversation, I will not. Which then does not help with education of arbiters.

Bill Gletsos
09-02-2009, 03:12 PM
That's what caused the problem. There was a very long wait for the arbiters to finish the last round draw. When I asked the arbiters about the reason for the delay, they told me that they use two programs that give different pairing for the first board opponent (against David Smerdon).

According to B. Gletsos Swiss Perfect pairing for the last round is correct.
I am not going to dispute the premise. However, it means the second program gave incorrect pairing.Correct Swiss Manager wants to pair Smerdon with Johannessen.

I am not even sure SP was one of the program used.Yes it was. The other program was Swiss Manager.

However, it would be interesting if someone tried to do a last round pairing using different software(s).Oh rest assured some so called armchair arbiter has already done that using Swiss Master. :owned:

Side note:
I witnessed (and was affected by) 4 manual changes to the pairing.
Three of those changes turned out to be incorrect (e.g. the program was right). On the fourth occasion arbiter forgot to flag the last round. When this option was ticked, silicon made draw matched the human intervention.
These experiences taught me to trust computerised pairing.I assume the 4 manual pairings you refer to occurred in other tournaments and not in the recent Queenstown event.

Bill Gletsos
09-02-2009, 03:14 PM
Yes, we can. :hand:

We have taken the discussion private.The discussion was private to start with as you provided nothing to back up your claim that according to FA Shaun Press the arbiters at Queenstown got it incorrect.

Brian_Jones
09-02-2009, 03:15 PM
Shame. I was hoping to actually see some meaningful information on the situation, which unless I am included in the private conversation, I will not. Which then does not help with education of arbiters.

It is not my job to educate arbiters.

It is a pity that this bulletin board cannot discuss things sensibly without personal prejudices getting in the way.

Maybe one day the ACF will show some leadership?

Denis_Jessop
09-02-2009, 03:51 PM
It is not my job to educate arbiters.

It is a pity that this bulletin board cannot discuss things sensibly without personal prejudices getting in the way.

Maybe one day the ACF will show some leadership?

Brian

What on earth is eating you? You are not doing yourself or your cause, whatever it is, any good by being silly like this. Talk of "leadership" without any specifics is as silly as talking about complying with the rules without giving details. Either you engage in an intelligent conversation, which I know you can do, or you say nothing.

DJ

Garvinator
09-02-2009, 06:15 PM
Another solution is that people just post their calculations, or the calculations that they are proposing are better than the others from other sources.

Bill Gletsos
12-02-2009, 10:18 PM
BTW I am advised by a reliable source that FA Shaun Press now acknowledges that the pairing Jones V Smerdon was correct.

Garvinator
13-02-2009, 12:46 AM
Ok, when I go to an online dictionary and look for a definition of a beat-up, will it direct me to this thread?