OTB matches?

[Smythe_Dakota]

@nolan

… A 10-0 result in a match by a players rated 114 points below the opponent would be well outside of expectations and should trigger an alert in validation.

Besides, it wasn’t a single 10-0 result, it was three of them, on three consecutive days.

And I think the rating difference was greater than 114 points, too. The case I’m thinking about was from 1972, which is probably not the same incident Alex mentioned. Were there even rating floors in 1972?

There was another famous case where a player was ceilinged rather than floored, because there was an automatic minimum gain of 2 points for any win, even against an opponent rated 1000 points lower. In this latter case there was a handwritten note on this player’s rating card that said “Do not let this player’s rating exceed 2399 without permission from Ed Edmondson.”

Bill Smythe

[nolan]

The expected performance of a 1286 player against a 1400 player is 0.341585, so in a 10 game match the lower rated player would be expected to have a score of around 3.5 - 6.5.

[Smythe_Dakota]

I was agreeing with you, while pointing out that the actual situation was even more dubious than that, because it was essentially a 30-game match.

Bill Smythe

[relyea]

@nolan

The expected performance of a 1286 player against a 1400 player is 0.341585, so in a 10 game match the lower rated player would be expected to have a score of around 3.5 - 6.5.

But there is a very high probability that a 1400 player isn’t playing at 1400 strength, which is a probability you don’t see with a 1401 player.

Alex Relyea

[nolan]

There aren’t that many 1400 players at their floor compared to higher floors, with 2200 being the largest, because of OLM floors.

[scottrparker]

@relyea

@nolan

The expected performance of a 1286 player against a 1400 player is 0.341585, so in a 10 game match the lower rated player would be expected to have a score of around 3.5 - 6.5.

But there is a very high probability that a 1400 player isn’t playing at 1400 strength, which is a probability you don’t see with a 1401 player.

Alex Relyea

True, but even if the 1400 player is playing below that level the odds of going zero for thirty in this situation are vanishingly small.

[Mulfish]

@scottrparker

@relyea

@nolan

The expected performance of a 1286 player against a 1400 player is 0.341585, so in a 10 game match the lower rated player would be expected to have a score of around 3.5 - 6.5.

But there is a very high probability that a 1400 player isn’t playing at 1400 strength, which is a probability you don’t see with a 1401 player.

Alex Relyea

True, but even if the 1400 player is playing below that level the odds of going zero for thirty in this situation are vanishingly small.

And that’s not even factoring in the probability a player who has been clobbered for two consecutive day would come back for day three.

[scottrparker]

Let’s assume the 1400 player is playing 114 points below his rating, so that he and the 1286 player are playing as equals. The odds that one of them would lose 30 consecutive games to the other are 1 in 1,073,741,824. The Powerball lottery has better odds than that.

[nolan]

Yeah, but the powerball is (in theory) a totally random chance event, a game of chess is not.

Back in my active days there was a player 2 classes above me against whom I had a 10-0 streak. I just had his number, I guess.

[wintdoan]

@scottrparker

Let’s assume the 1400 player is playing 114 points below his rating, so that he and the 1286 player are playing as equals. The odds that one of them would lose 30 consecutive games to the other are 1 in 1,073,741,824. The Powerball lottery has better odds than that.

Actually, it’s not even that good—that doesn’t allow for draws. Even if the probability of a draw in an individual game is just .10, it turns out to be more like 25 times those odds.