@JimiRecard When we play in Tournaments and are paired with a player, any player who abandons or doesn't move within 20 seconds loses by default. My point once again, (respectively) is that many players will join a game, if they get white, they move, if they get black they simply don't move or leave and aren't penalized. If Lichess wants to start implementing penalizing all players who join a rated game and then either [don't move, abandon, or abort] as a loss, I'm more on board with that (pun intended). I have no problem with people who only want to play white, but I do not want to play them. I leave them to do as they please.
I want to play everyone a fair number of games, especially if I play as black and I lose as black. I don't feel that setting up this function would be difficult. I am enjoying this thread though.
On a side note: I believe Rashid Nezhmedinov is the greatest attacking player of all time, even better than Tal.
I want to play everyone a fair number of games, especially if I play as black and I lose as black. I don't feel that setting up this function would be difficult. I am enjoying this thread though.
On a side note: I believe Rashid Nezhmedinov is the greatest attacking player of all time, even better than Tal.
Lichess does penalize people that abort/intentionally lose a lot of games @juliegirl1999
If this really is a common occurrence — and not just something the OP observed once or twice because her opponents had bad connections 😉 — then might it not explain the disparity in the database, in and of itself? I realize aborted games don’t count as losses. But if a significant number of players really are willing to abort games as Black, maybe they also resign early as Black, or give up more easily in tough positions, etc.?
In other words this might well be psychological, or even purely behavioral. And if that’s the case, it makes the argument circular: Black loses slightly more games on average, simply because some Black players quit their games more readily. If this is indeed the case, then it may be profoundly sad, but is not otherwise profound. Importantly, it in no way proves that moving first is significant enough that it will convert to a winning advantage when the player with Black knows what he’s doing.
And it seems to me this could easily account for White winning slightly more than half the games in the database here: only a small number of Black players exhibit this behavior; but it’s a learned behavior, so people who think that way may be found in all the rating brackets. (“You should play for a win as White, but be content with a draw as Black.” Doesn’t even the Sᵗ Louis Chess Club teach that?)
🤷 Some food for thought.
In other words this might well be psychological, or even purely behavioral. And if that’s the case, it makes the argument circular: Black loses slightly more games on average, simply because some Black players quit their games more readily. If this is indeed the case, then it may be profoundly sad, but is not otherwise profound. Importantly, it in no way proves that moving first is significant enough that it will convert to a winning advantage when the player with Black knows what he’s doing.
And it seems to me this could easily account for White winning slightly more than half the games in the database here: only a small number of Black players exhibit this behavior; but it’s a learned behavior, so people who think that way may be found in all the rating brackets. (“You should play for a win as White, but be content with a draw as Black.” Doesn’t even the Sᵗ Louis Chess Club teach that?)
🤷 Some food for thought.
Ok can we just talk about the title? Chess is sexy enough already.
Honestly I don't see anything wrong with my solution... see http://glicko.net/glicko/glicko2.pdf
E(µ, µj, φj ) is the win expectancy of player P versus player Pj (j ranging from 1 to m) in a rating period with m rounds. So if two 1500 players play, then whichever player has White should have an expectancy like 0.55, and Black 0.45.
E(µ, µj, φj ) is the win expectancy of player P versus player Pj (j ranging from 1 to m) in a rating period with m rounds. So if two 1500 players play, then whichever player has White should have an expectancy like 0.55, and Black 0.45.
@pawnedge the study I reference involves games that predate computerized chess. It's in that Wikipedia link.
I specifically mentioned this database only because № 30 did. I was attempting to tie his and your arguments together; observing that if you are both correct, then your points effectively cancel out. To reiterate: Unless the cause can be shown to be other than psychological, the connection you’re attempting to make is having it both ways: Black players quit playing more often because they’re afraid of losing; Black players lose more often because they’re likelier to quit playing. 🤷 Whoop-de-do.
Of course chess is a very old game, so if the “Black should be content with a draw” school of thought has been around longer than computer chess, this changes nothing. It’s still a behavioral matter — and therefore an effect, rather than a cause. 😉
I am still mulling over Toadofsky’s PDF.
Of course chess is a very old game, so if the “Black should be content with a draw” school of thought has been around longer than computer chess, this changes nothing. It’s still a behavioral matter — and therefore an effect, rather than a cause. 😉
I am still mulling over Toadofsky’s PDF.
Since we can agree on the basic fact (white wins a little more), rest the discussion of the reason for that. It could be behavioral/psychological, a small (but real) advantage for white, or something else.
I could argue that a behavioral/psychological advantage is a real advantage nonetheless. But I believe there's a small objective advantage of having the first move in the standard starting position of chess. My arguments to defend this are still statistical, observing the top dogs playing. DrNykterstein, aka GM Magnus Carsen, has won 2287 games as white on lichess and 2056 as black. Mostly bullet games (supposedly white's advantage is less significant in fast time controls). He's a terrible loser (according to himself) and I don't think he gives up easily because he's playing black, and I don't believe his opponents would do it also (the chance to beat the world champion!).
Another example, RebeccaHarris, aka GM Daniel Naroditsky, has won 2003 games as white and 1875 as black. Super competitive player, almost same points as above.
But sure, let's believe some human opponents are prone to give up easily as black. There's an event in which we can put this to test, an event with the top chess players in the world: TCEC (Top Chess Engine Championship). The superfinal of season 20 ended a couple of days ago, featuring Stockfish vs Leela. The engines played 100 games against each other, with 50 openings predetermined, 2 games for each opening, players taking turns as black and white. The result: Stockfish won 14 games as white and Leela won 8 as white. There was no win as black. All the other games resulted in a draw. It's hard to argue engines have a psychological/behavioral component making them win less as black. This feeds my hypothesis: chess is a draw with perfect play, with a slight advantage for white. Granted, this is not really my hypothesis, I just adopted it from people who have studied chess far more than I will in my life.
Making a parallel with a physical sport: it's like having a football (or soccer) team with 11 players against another with 10. The team with 11 players is more likely to win, all else the same. But there's no guarantee, and the team with 10 players can very well win.
Just more food for thought 🤷
I could argue that a behavioral/psychological advantage is a real advantage nonetheless. But I believe there's a small objective advantage of having the first move in the standard starting position of chess. My arguments to defend this are still statistical, observing the top dogs playing. DrNykterstein, aka GM Magnus Carsen, has won 2287 games as white on lichess and 2056 as black. Mostly bullet games (supposedly white's advantage is less significant in fast time controls). He's a terrible loser (according to himself) and I don't think he gives up easily because he's playing black, and I don't believe his opponents would do it also (the chance to beat the world champion!).
Another example, RebeccaHarris, aka GM Daniel Naroditsky, has won 2003 games as white and 1875 as black. Super competitive player, almost same points as above.
But sure, let's believe some human opponents are prone to give up easily as black. There's an event in which we can put this to test, an event with the top chess players in the world: TCEC (Top Chess Engine Championship). The superfinal of season 20 ended a couple of days ago, featuring Stockfish vs Leela. The engines played 100 games against each other, with 50 openings predetermined, 2 games for each opening, players taking turns as black and white. The result: Stockfish won 14 games as white and Leela won 8 as white. There was no win as black. All the other games resulted in a draw. It's hard to argue engines have a psychological/behavioral component making them win less as black. This feeds my hypothesis: chess is a draw with perfect play, with a slight advantage for white. Granted, this is not really my hypothesis, I just adopted it from people who have studied chess far more than I will in my life.
Making a parallel with a physical sport: it's like having a football (or soccer) team with 11 players against another with 10. The team with 11 players is more likely to win, all else the same. But there's no guarantee, and the team with 10 players can very well win.
Just more food for thought 🤷
Almost you convince me, sir. My biggest caveat is with your soccer analogy. I think it would depend on who that 11ᵗʰ player was. For example, if it were Cinderella, then as she would always run away from the ball, I doubt having her on my team would be an advantage, even if she were an extra player. 😉 Whereas I could probably win games with just nine players, provided one of them was Robert Lewandowski, and we were up against the guys from the office next door. 🤷
But let’s bring this analogy back to chess, by speaking of the preparation for a world championship match. If Caruana has one extra coach on his “team” of advisors than Carlsen, does it matter? I think it’s more important how good his coaches are, not even in terms of their rating but rather their knowledge, ideas and insights, as well as simply their ability to communicate. But then, I seem to recall that Fischer beat Spassky without any help. So the argument could even be made that a Serious player “don’t need no stinking coaches.” 😛
Also, I’m not quite sure what to make of any stats from TCEC, considering the various engines are all run on different hardware, and Stockfish always seems to be an older version. (Surely a real contest would use [1] the same CPU for every engine, and [2] the latest version of each engine. 🤔) The TCEC people seem to know what they’re doing, but I don’t.
However, I hadn’t noticed that before about top players’ winning stats. And a cursory glance at other “high-profile profiles” mostly corroborates your data:
(TG = Total Games; WW = White Wins; etc.)
alireza2003
TG: 5,396
WW: 1,582
WL: 991
BW: 1,502
BL: 1,017
howitzer14
TG: 10,000+
WW: 6,087
WL: 1,042
BW: 5,980
BL: 1,159
Zhigalko_Sergei
TG: 10,000+
WW: 10,000+
WL: 5,130
BW: 10,000+
BL: 5,864
german11
TG: 10,000+
WW: 10,000+
WL: 10,000+
BW: 10,000+
BL: 10,000+
While I wish the Advanced Search could get more specific than 10,000, that does seem to be fairly compelling evidence, compared to any I’ve seen so far. I haven’t checked very many profiles yet, but there’s a good chance you are right about this trend you’ve observed. Nice find. 🙂
But let’s bring this analogy back to chess, by speaking of the preparation for a world championship match. If Caruana has one extra coach on his “team” of advisors than Carlsen, does it matter? I think it’s more important how good his coaches are, not even in terms of their rating but rather their knowledge, ideas and insights, as well as simply their ability to communicate. But then, I seem to recall that Fischer beat Spassky without any help. So the argument could even be made that a Serious player “don’t need no stinking coaches.” 😛
Also, I’m not quite sure what to make of any stats from TCEC, considering the various engines are all run on different hardware, and Stockfish always seems to be an older version. (Surely a real contest would use [1] the same CPU for every engine, and [2] the latest version of each engine. 🤔) The TCEC people seem to know what they’re doing, but I don’t.
However, I hadn’t noticed that before about top players’ winning stats. And a cursory glance at other “high-profile profiles” mostly corroborates your data:
(TG = Total Games; WW = White Wins; etc.)
alireza2003
TG: 5,396
WW: 1,582
WL: 991
BW: 1,502
BL: 1,017
howitzer14
TG: 10,000+
WW: 6,087
WL: 1,042
BW: 5,980
BL: 1,159
Zhigalko_Sergei
TG: 10,000+
WW: 10,000+
WL: 5,130
BW: 10,000+
BL: 5,864
german11
TG: 10,000+
WW: 10,000+
WL: 10,000+
BW: 10,000+
BL: 10,000+
While I wish the Advanced Search could get more specific than 10,000, that does seem to be fairly compelling evidence, compared to any I’ve seen so far. I haven’t checked very many profiles yet, but there’s a good chance you are right about this trend you’ve observed. Nice find. 🙂
This topic has been archived and can no longer be replied to.