SO WHO'S GOING TO SHEFFIELD? I AM!

[Roger de Coverly]

Bouncing. In Bermuda I played White v 1400 and won. Black against 2500 and lost. White v 1600 and won. Black against 2400 and lost. White agaist 1800 and won. I was 2200. I think that was unsatisfactory, 3/5 and playing nobody in my reasonabled range. Surely it can't happen to the bottom strength and rated players?

That's a problem with having all the players in the same section and a wide rating range though. In Dutch events, they usually apply rating restrictions so you end up with an Open A and an Open B etc. In the UK, there are usually parallel events which reduce the entry amongst lower rated players in the main event.

When acceleration ends, seeded pairings take over. With players outside of their natural score bands, rating mismatches become a feature again. There's almost a theory, that if the tournament has a double round day, that you make it the rounds immediately after the accelerating finishes, so as to get short games between mismatched players.

In this year's British, there's not much to choose between the third and fourth quartiles, a hundred points on many boards. So there will be players from the second half, possibly even the forth quartile in the top echelons for several rounds.

I happened to look up one of the old pairing rules for the British Championship. At one time they would order the players by grade and seed the top half. So the number one seed would be on the top board. They would then select their opponents from the second half by lot.

[Alex Holowczak]

Alex, have you not read my book?

No. Make it available in a web format, and I'd give it a read.

33 v 1,3,5,7. 31 v 63, 61, 59, 57 with the same colours. Ludicrous. Unlikely, but it has happened similarly to players in the British. Testing a system to its limits often shows whether it works.

There are scenarios in which the FIDE-rating system, and ECF-grading system, don't work if you test them to their limits. I think it's just the nature of the beast of having a tournament where each player plays a selective number of opponents, rather than an APA.

Bouncing. In Bermuda I played White v 1400 and won. Black against 2500 and lost. White v 1600 and won. Black against 2400 and lost. White agaist 1800 and won. I was 2200. I think that was unsatisfactory, 3/5 and playing nobody in my reasonabled range. Surely it can't happen to the bottom strength and rated players?

I agree with the post already that this was caused by having a wide rating range of players in the competition, rather than a fault with the Swiss system.

The Olympiad does not use Acceleration anymore; my advice was ignored and it was a disaster. The Burstein system is used, relying on Bucholz. That is, I think, more complex than Acceleration. So is Dubov.

Why? I thought it worked perfectly well this time around. I don't know what any of the latter pairing systems are. I assumed they used a system of sorting by matchpoints, and sorting within each group on gamepoints. Then pair each matchpoint group within the newly established seeds, where seeds are determined by gamepoints descending. That may or may not be what you gave a name to above. The only pairing disaster at the Olympiad was the refusal to pair Israel against someone who'd play them.

[Simon Ansell]

Bouncing. In Bermuda I played White v 1400 and won. Black against 2500 and lost. White v 1600 and won. Black against 2400 and lost. White agaist 1800 and won. I was 2200. I think that was unsatisfactory, 3/5 and playing nobody in my reasonabled range. Surely it can't happen to the bottom strength and rated players?

I agree with the post already that this was caused by having a wide rating range of players in the competition, rather than a fault with the Swiss system.

Exactly. It's what happens when you have 1800s playing in the same section as 2700s...

[Adam Raoof]

Alex, have you not read my book?

No. Make it available in a web format, and I'd give it a read.

http://englishchessorganisers.blogspot.com/

though not all of it is available.

[Stewart Reuben]

If you are at the bottom of the top half of the British and it was a seeded Swiss, in the first round you would meet the lowest rated opponent. If you drew, then again you would meet a very low rated opponent. Doesn't seem much fun to me.

In Gibraltar the event is open. Range has been 2700 to 700. Bounces few and far between. I wonder why? It wouldn't happen with Dubov either.

It was 2008 that a flawed accelerated system was used in Dresden in the Olympiad. In 2010 they used a sensible, but complex system.
Stewart Reuben

[Simon Ansell]

The one time I played in Gibraltar I only played one player within 130 points of my own rating. That was in round 2 (vs Gary Quillan), after drawing my first game against a GM. After that I bounced...

[Roger de Coverly]

If you are at the bottom of the top half of the British and it was a seeded Swiss, in the first round you would meet the lowest rated opponent. If you drew, then again you would meet a very low rated opponent. Doesn't seem much fun to me.

Would it really happen that way? If you are midway in the ranking list, wouldn't your second round be equally likely to be a top half player also held to a draw. So you switch from being bottom of the top half to top of the bottom half. There may be some players below 2000 in the British, I don't think there will be any really weak players though.

My experience of acceleration in the Hastings Masters from a third quartile perspective is that in the first round I get an opponent 200 or 300 points lower. In the second and third rounds I bounce with the acceleration to players 250 to 400 points higher. With one from three, the fourth round is another lower rated player. After that, you tend to get runs of the same type of opponent. So a winning run is rewarded with players 100-250 points higher and a losing or drawing run with players 100-250 points lower.

[Alan Walton]

I would have to agree with Simon each year I played at Gibraltar, unless I got a result against somebody stronger than myself, I was always bouncing as well

[IM Jack Rudd]

If you are at the bottom of the top half of the British and it was a seeded Swiss, in the first round you would meet the lowest rated opponent. If you drew, then again you would meet a very low rated opponent. Doesn't seem much fun to me.

Would it really happen that way? If you are midway in the ranking list, wouldn't your second round be equally likely to be a top half player also held to a draw. So you switch from being bottom of the top half to top of the bottom half.

This can't happen between rounds 1 and 2 unless there are half-point byes; the first-round draws will be equally divided between top-half players and bottom-half players. It's a common scenario between other pairs of rounds, though.

[Alex McFarlane]

Sounds like we need some empirical evidence rather than hearsay. I'll try to do an analysis during the event of 'bouncing'. My gut instinct is that for quarters 1 and 4 it reduces this effect but is less clear for Q2 and Q3.

[Roger de Coverly]

This can't happen between rounds 1 and 2 unless there are half-point byes

I'm not sure I follow that. Suppose 80 players, seeded pairings.

The only draws in round 1 are 1v41 and 80v40. Don't you get second round pairings of 40v1 and 41v80? So 40 has flipped from playing the bottom seed to playing the top seed.

[IM Jack Rudd]

This can't happen between rounds 1 and 2 unless there are half-point byes

I'm not sure I follow that. Suppose 80 players, seeded pairings.

The only draws in round 1 are 1v41 and 80v40. Don't you get second round pairings of 40v1 and 41v80? So 40 has flipped from playing the bottom seed to playing the top seed.

Yes and no. He's flipped from playing the bottom seed to playing the top seed, but he hasn't flipped from top half to bottom half: he's still in the top half of the scoregroup, but has been median-flipped to allow the pairings to work.

[Roger de Coverly]

Yes and no. He's flipped from playing the bottom seed to playing the top seed, but he hasn't flipped from top half to bottom half: he's still in the top half of the scoregroup, but has been median-flipped to allow the pairings to work.

I was looking from the player's viewpoint rather than the arbiter. So for a given score do you get a top half opponent or a bottom half? A first round draw against a player at the foot of the draw can give you a pairing against one of the top seeds for the second round if a number of them draw.

In smaller tournaments with seeded pairings, I know that if I win "from the back of the room" in the first round, I will get one of the top seeds in the second.

[IM Jack Rudd]

Yes and no. He's flipped from playing the bottom seed to playing the top seed, but he hasn't flipped from top half to bottom half: he's still in the top half of the scoregroup, but has been median-flipped to allow the pairings to work.

I was looking from the player's viewpoint rather than the arbiter. So for a given score do you get a top half opponent or a bottom half? A first round draw against a player at the foot of the draw can give you a pairing against one of the top seeds for the second round if a number of them draw.

It can, but there are some pretty specific constraints on that number; if {draws where the seed is white}!=1 and {draws where the seed is black}!=1, I don't think it can happen.

[Richard Bates]

Yes and no. He's flipped from playing the bottom seed to playing the top seed, but he hasn't flipped from top half to bottom half: he's still in the top half of the scoregroup, but has been median-flipped to allow the pairings to work.

I was looking from the player's viewpoint rather than the arbiter. So for a given score do you get a top half opponent or a bottom half? A first round draw against a player at the foot of the draw can give you a pairing against one of the top seeds for the second round if a number of them draw.

It can, but there are some pretty specific constraints on that number; if {draws where the seed is white}!=1 and {draws where the seed is black}!=1, I don't think it can happen.

Surely your argument is assuming equal distribution of colours? Since colours take precedence, any imbalance between top and bottom half will potentially see the player in question "switch".

Anyway it's missing the obvious response to Stewart's scenario (as with the original one). If a "bottom of the top half" player isn't particularly enthralled by the prospect of playing a second weak player, having drawn with the weakest player, then s/he has a fairly easy solution. That is to make sure s/he beats the bottom seed as expected in round1 !