Hastings Congress

[IM Jack Rudd]

This system has too many 400 plus mismatches as far as I'm concerned.

That seems mostly to be a feature of the field rather than the pairing system. I suppose a more aggressive acceleration might have damped down the number of such pairings.

[Richard Bates]

This system has too many 400 plus mismatches as far as I'm concerned.

That seems mostly to be a feature of the field rather than the pairing system. I suppose a more aggressive acceleration might have damped down the number of such pairings.

Surely the more aggressive the acceleration then potentially the more likely you are to get such pairings? (although may be misunderstanding how you are defining 'aggressive'). Just in round 1-3 rather than rounds 4-5. The issue with the system in this tournament seems to be that there haven't been enough people in the top half available to competently despatch those doing well in the bottom half. On top of some bizarre anomalies such as one player in the bottom half (rated 2080) who has made it to 2.5 after drawing with an 1800 in round 1 and then beating a couple of 1500s. A situation created by the "Baku bonus" being a full point rather than half a point and therefore there not being many top half players coinciding with his score group.

[NickFaulks]

Let's just cut through the crap and say that the whole thing is misdirected garbage. Only Swiss Pairings Commission know the basis for it and they're not telling.

Which reminds me, we have our very own Councillor on that. Rupert, are you available for questions?

[Ian Thompson]

This system has too many 400 plus mismatches as far as I'm concerned.

I'd like to know whether it makes much difference to the number over the whole tournament when compared to a normal pairing system.

With a normal system you get them in the early rounds; with the Baku system you avoid them in the early rounds, but get them later, when the virtual points reduce, as they have done in round 4 and will do again in round 6.

It would be interesting to compare the number of mismatches in the actual round 4 pairings compared to the mistaken non-accelerated pairings that were briefly published last night. (For me, I've gone from being rated 553 points higher than my non-accelerated opponent to 278 points lower than my accelerated opponent.)

[A.Kluckova]

If you follow the Swiss system, can not happen, that the player with Rp 2416 is 40 ranks lower than player with Rp 1940. Swiss supported with bucholtz as a side criterium follows a Statistics theory.
If you support Acceleration system with the number of wins as a side criterium, results looks like a Theory of relativity...

[Richard Bates]

This system has too many 400 plus mismatches as far as I'm concerned.

I'd like to know whether it makes much difference to the number over the whole tournament when compared to a normal pairing system.

With a normal system you get them in the early rounds; with the Baku system you avoid them in the early rounds, but get them later, when the virtual points reduce, as they have done in round 4 and will do again in round 6.

It would be interesting to compare the number of mismatches in the actual round 4 pairings compared to the mistaken non-accelerated pairings that were briefly published last night. (For me, I've gone from being rated 553 points higher than my non-accelerated opponent to 278 points lower than my accelerated opponent.)

The general difference between accelerated and non-accelerated pairings is the people getting the mismatches. Accelerated systems (designed around assisting norms etc, rather than the old simple "ensure you find a winner motivation" from where they originated in the 70s) generally ensure that people who perform well aren't used to clear away the bottom half of the field, the job being done by those who have dropped points. The consequence is that strong players are more likely to slip up, but it won't significantly affect their chances of doing well overall in the tournament (assuming no tiebreaks based on strength of opponents) if they do. Whereas potential norm seekers, especially GM norm seekers, will know that they won't miss out purely as a result of their field being too weak.

So the pros are mainly all around norms, but big negatives around relative strengths of fields for those amongst the prizes. Possibly Hastings this year is too strong for it actually to make a massive difference on norms (may ultimately be a negative!) - but this would not have been the case in recent years.

I presume that the main attraction of the Baku system, and why it is FIDE endorsed, is that it is very transparent and easy to programme into computer pairing systems. Although does rely on organisers not running parallel pairing systems and accidentally publishing the results...

[Kevin Thurlow]

"So the pros are mainly all around norms, but big negatives around relative strengths of fields for those amongst the prizes. "

A fair assessment. The other thing is you merely postpone the mismatches. Instead of a series of 1-0, 0-1, 1-0... (or 0-1, 1-0, 0-1...) in the first two rounds, and the tournament starts in round 3, you get those games halfway through or later.

[Alex McFarlane]

A few general comments on acceleration not specific to Hastings.
Positives
Strong players meet early on which pleases spectators and sponsors (assuming not every game ends in a quick draw).
In a very large tournament, such as are common abroad, players can score 6.5 and fail to get an IM norm or 7 and fail to get a GM norm because their first two opponents were relatively weak and only one can be 'upgraded'.
There are fewer huge rating mismatches. These are not transferred from round 1 to round 3 unless all of the top quarter lose a game! THe reduction from one bonus point to a half also reduces this.
Negatives
Strong players tend not to like the system as it gives them an extra game which is competitive
Difficult to predict draw

[Kevin Thurlow]

Also, "second quarter" players can be on 0/2 after maybe one bad game. For some reason 7- round Guernsey accelerated about 30 years ago. I lost to Julian Hodgson in Round 1 (which is difficult to complain about), then lost to an "estimated 1800" (hmm), then came the three mismatches, followed by a win over a slightly lower rating, and a win over someone rather good, who had been given tough opposition every game! He was probably quite tired and I wasn't. So I got 5/7 and just about justified my rating. People were congratulating me on this good score and I was trying to tell them that it wasn't good.

It probably works better with longer events. At Rekjavik last year, my round 2 opponent (an IM) found herself on 0/2, but returned to the top half fairly rapidly and at least had several games with good opposition, and duly finished some way ahead of me.

Swisses do always produce some weird results whatever you do.

[Kevin Thurlow]

"Strong players meet early on which pleases spectators and sponsors (assuming not every game ends in a quick draw)."

which I suppose is a variation of the Swiss Gambit, you have a quick draw against another GM and play someone much weaker the next day. In a longer tournament, you are not damaging your chances.

[Alex McFarlane]

All Swisses can produce a bounce where players alternate between high and low rated players.

In some events this can be more like a bungee jump than a yo-yo though.

[Nick Grey]

Nick F - you ought to enjoy it anyway. Maybe you will get an up float to young Patrick. I was impressed by his play at OLYMPIA. Best 6 year old I have played since Luke McShane.

I suppose as bad as any other pairing system in a 9 round swiss.

Any good games to go through today - TV is a bit dull tonight.

[Richard Bates]

Fair to say that another flaw in accelerated pairings can emerge unexpectedly when Bogdan is involved...

[NickFaulks]

I suppose as bad as any other pairing system in a 9 round swiss.

No it isn't. What to you believe is wrong with the system that has been used for so many years?

[NickFaulks]

Meanwhile, guess who's keeping the party waiting. Not grinding out his usual win, sadly.