Three-Move Draw Scoring

Sort:
MD1806
HonSec wrote:
MD1806 wrote:
magipi wrote:
MD1806 wrote:
magipi wrote:
MD1806 wrote:
magipi wrote:
MD1806 wrote:

Yes, what's wrong with proposing a rule change based on my experiences?

That it doesn't make any sense?

This is your last draw:

https://www.chess.com/game/live/123054679284?username=md1806&move=60

You were up a pawn, but your position was actually slightly worse, because your bad king position. Your opponent gave perpetual check thinking it was a good result for him. Now you propose that you should have lost the game after that perp? Seriously? In which world would that make any sense? In which world would that fix any problem?

Yes, it would be hypocritical of me not to apply my own rule to myself. I should've lost 8 points.

"In which world would that make any sense? In which world would that fix any problem?"

Did you read my original post? It would solve the problem of time-wasting draws, and it would incentivize playing to win rather than playing to suck out of a tied/advantageous position.

Your own game is a counterexample to both points. That game was not a time-wasting draw, it was a hard fought one. And the guy who gave perpetual check (your opponent) would be incentivized to do so even more, as he would have won.

Bold of you to assume how I feel about my own chess games. Every game I play that ends in a draw is a waste of time. The ones that do so when I'm down material are less so, but still. If this site actually allowed you to regularly play against people rated hundreds of points above you, like others do (cough one that starts with Li), I would resign games rather than play move draws. But I'm stuck challenging only people within a narrow band, so rating points are precious.

There are still several key problems with your idea, of which the most pertinent are:

1) you haven't proven that draws are a problem, beyond the fact that you yourself regard them as a waste of time. A subsidiary problem here is that you haven't proven that draws by repetition in particular are a problem at the level you're talking about.

2) you haven't addressed the point that being up or down material doesn't correlate reliably with whether your position was worse, better, or equal. You are trying to apply a moral standard to the player who is up material, on the basis that they 'should' have pressed harder to win, ignoring the fact that the player who is up material might never have had an advantage on the board, and indeed might be doing very well to escape with a draw.

3) you haven't addressed the problem that your solution creates, namely that there would be a serious incentive to play for a draw by repetition if you can sacrifice material to do so. Given that draws by repetition are pretty rare to begin with, I think there's a reasonable chance you might actually make them more common if this rule change was instigated.

You're proposing a fairly major rule change. You need to justify the need for it by identifying a problem, and you need to show how your change is a solution. Thus far you've done neither.

1. What would prove to you that draws are a problem?

2. Nobody is applying a "moral standard;" where did I mention morality?

3. Your logic here doesn't make sense. People who sacrifice material and don't gain from it are already trying to play for a draw instead of a loss (if the sacrifice works, they end up either winning or regaining the sacrificed material). Are repetition draws rare? Do you have data to back up that assertion?

HonSec
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
magipi wrote:
MD1806 wrote:
magipi wrote:
MD1806 wrote:
magipi wrote:
MD1806 wrote:

Yes, what's wrong with proposing a rule change based on my experiences?

That it doesn't make any sense?

This is your last draw:

https://www.chess.com/game/live/123054679284?username=md1806&move=60

You were up a pawn, but your position was actually slightly worse, because your bad king position. Your opponent gave perpetual check thinking it was a good result for him. Now you propose that you should have lost the game after that perp? Seriously? In which world would that make any sense? In which world would that fix any problem?

Yes, it would be hypocritical of me not to apply my own rule to myself. I should've lost 8 points.

"In which world would that make any sense? In which world would that fix any problem?"

Did you read my original post? It would solve the problem of time-wasting draws, and it would incentivize playing to win rather than playing to suck out of a tied/advantageous position.

Your own game is a counterexample to both points. That game was not a time-wasting draw, it was a hard fought one. And the guy who gave perpetual check (your opponent) would be incentivized to do so even more, as he would have won.

 

You're proposing a fairly major rule change. You need to justify the need for it by identifying a problem, and you need to show how your change is a solution. Thus far you've done neither.

1. What would prove to you that draws are a problem?

2. Nobody is applying a "moral standard;" where did I mention morality?

3. Your logic here doesn't make sense. People who sacrifice material and don't gain from it are already trying to play for a draw instead of a loss (if the sacrifice works, they end up either winning or regaining the sacrificed material). Are repetition draws rare? Do you have data to back up that assertion?

1. I would say that they need to occur an unacceptably large proportion of the time. Your mileage will vary on what is deemed acceptable, but at the level you're talking about the vast majority of games are decisive. Looking at our stats both you and I, for example, who are in the range you're talking about, draw significantly less than 10% of the time in blitz (7% for you, just less than 6% for me). I consider that draw rate entirely acceptable. Regarding draws by repetition specifically, my rate is just shy of 2.5%. Yours is a little higher, at 5%, but that's still only one such draw every 20 games. Assuming we are representative I see absolutely no need to change the rules here.

2) you didn't mention it by name, but that's what you're doing. Your position is that the player up material should lose in a draw by repetition because they should be pushing harder for a win - which is to say, you're arguing that they should be punished because they haven't behaved correctly. That's a moral standard. It just happens to be an arbitrary one.

3) 'People who sacrifice material and don't gain from it are already trying to play for a draw instead of a loss' - that's precisely my point. Your rule (potentially) rewards these people with a win, and therefore acts against its intended purpose. See 1) for the answer to your question about data.

Martin_Stahl
MD1806 wrote:

...Are repetition draws rare? Do you have data to back up that assertion?

Draws of all types at sub-master level are going to likely be less than 10% of games, in most cases.

https://www.chess.com/article/view/interesting-chess-data-time-controls-and-game-results

This is an older article but the draw percentage, of all types, for a given time live control topped out at 4%. Daily is likely going to be higher and probably very long time controls as well (maybe 10-15% max)

My all time numbers are

  • Bullet 5.3% - repetition 60% (3 games)
  • Blitz 4.1% - repetition 44.3%
  • Rapid 3 4% - repetition 36.4% (only 4 games)
  • Daily: 9.7% - repetition 2.9%

I'm sure other members are likely similar, with it being more common, as a percentage of overall draws, in faster time controls, where finding a win is harder in time trouble and a repetition draw is better than a loss on time.

MD1806
Martin_Stahl wrote:
MD1806 wrote:

...Are repetition draws rare? Do you have data to back up that assertion?

Draws of all types at sub-master level are going to likely be less than 10% of games, in most cases.

https://www.chess.com/article/view/interesting-chess-data-time-controls-and-game-results

This is an older article but the draw percentage, of all types, for a given time live control topped out at 4%. Daily is likely going to be higher and probably very long time controls as well (maybe 10-15% max)

My all time numbers are

  • Bullet 5.3% - repetition 60% (3 games)
  • Blitz 4.1% - repetition 44.3%
  • Rapid 3 4% - repetition 36.4% (only 4 games)
  • Daily: 9.7% - repetition 2.9%

I'm sure other members are likely similar, with it being more common, as a percentage of overall draws, in faster time controls, where finding a win is harder in time trouble and a repetition draw is better than a loss on time.

See this is good data. I'd be interested to see how many draws occur before 36 moves.

HonSec
MD1806 wrote:
Martin_Stahl wrote:
MD1806 wrote:

...Are repetition draws rare? Do you have data to back up that assertion?

Draws of all types at sub-master level are going to likely be less than 10% of games, in most cases.

https://www.chess.com/article/view/interesting-chess-data-time-controls-and-game-results

This is an older article but the draw percentage, of all types, for a given time live control topped out at 4%. Daily is likely going to be higher and probably very long time controls as well (maybe 10-15% max)

My all time numbers are

  • Bullet 5.3% - repetition 60% (3 games)
  • Blitz 4.1% - repetition 44.3%
  • Rapid 3 4% - repetition 36.4% (only 4 games)
  • Daily: 9.7% - repetition 2.9%

I'm sure other members are likely similar, with it being more common, as a percentage of overall draws, in faster time controls, where finding a win is harder in time trouble and a repetition draw is better than a loss on time.

See this is good data. I'd be interested to see how many draws occur before 36 moves.

Do you acknowledge that all available data are showing that draws in general, and draws by repetition in particular, are rare?

MD1806
HonSec wrote:
MD1806 wrote:
Martin_Stahl wrote:
MD1806 wrote:

...Are repetition draws rare? Do you have data to back up that assertion?

Draws of all types at sub-master level are going to likely be less than 10% of games, in most cases.

https://www.chess.com/article/view/interesting-chess-data-time-controls-and-game-results

This is an older article but the draw percentage, of all types, for a given time live control topped out at 4%. Daily is likely going to be higher and probably very long time controls as well (maybe 10-15% max)

My all time numbers are

  • Bullet 5.3% - repetition 60% (3 games)
  • Blitz 4.1% - repetition 44.3%
  • Rapid 3 4% - repetition 36.4% (only 4 games)
  • Daily: 9.7% - repetition 2.9%

I'm sure other members are likely similar, with it being more common, as a percentage of overall draws, in faster time controls, where finding a win is harder in time trouble and a repetition draw is better than a loss on time.

See this is good data. I'd be interested to see how many draws occur before 36 moves.

Do you acknowledge that all available data are showing that draws in general, and draws by repetition in particular, are rare?

Rare compared to most other outcomes, yes. That doesn't mean they're not a problem. 3-5% is statistically significant.

Martin_Stahl

5 draws out of every 100 games is a problem?

MD1806
Martin_Stahl wrote:

5 draws out of every 100 games is a problem?

What percentage of games ended in locked-up boards before the en passant rule was instituted?

HonSec
MD1806 wrote:
Martin_Stahl wrote:

5 draws out of every 100 games is a problem?

What percentage of games ended in locked-up boards before the en passant rule was instituted?

Since the vast majority of games at the level of play you've specified would have been completely undocumented at the time, there is absolutely no way to answer this question. I'm also not aware that en passant was instituted specifically for this reason - do you have a source for this?

But leaving that aside, what draw percentage would you regard as acceptable, if 5% is too high?

MD1806
HonSec wrote:
MD1806 wrote:
Martin_Stahl wrote:

5 draws out of every 100 games is a problem?

What percentage of games ended in locked-up boards before the en passant rule was instituted?

Since the vast majority of games at the level of play you've specified would have been completely undocumented at the time, there is absolutely no way to answer this question. I'm also not aware that en passant was instituted specifically for this reason - do you have a source for this?

But leaving that aside, what draw percentage would you regard as acceptable, if 5% is too high?

It's fairly self-evident. In a game without the en passant, pawns can move past each other without danger and create an unbroken chain on each side that can't be breached.

Before endgame, and up material? 0%.

MD1806
long_quach wrote:

My solution for so many problems.

Chess as a triathlon.

All you need is 3 games for a championship.

1 game of Western chess.

1 game of Chinese chess.

1 game of Backgammon for a tie-breaker.

Ah, that "western chess" invented in Iran and India

MD1806
long_quach wrote:
MD1806 wrote:

Ah, that "western chess" invented in Iran and India

Chinese chess came from India too.

Chinese chess has more Chaturanga genes than Western chess.

There is no such thing as "western chess," just like there's no such thing as "west." There's just chess.

MD1806
long_quach wrote:
MD1806 wrote:
long_quach wrote:
MD1806 wrote:

Ah, that "western chess" invented in Iran and India

Chinese chess came from India too.

Chinese chess has more Chaturanga genes than Western chess.

There is no such thing as "western chess," just like there's no such thing as "west." There's just chess.

Western chess

Chinese chess.

You need to read the long, long series of books and articles debunking the Huntingtonian reductionism that produces myths like "the west"

HonSec
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
Martin_Stahl wrote:

5 draws out of every 100 games is a problem?

What percentage of games ended in locked-up boards before the en passant rule was instituted?

Since the vast majority of games at the level of play you've specified would have been completely undocumented at the time, there is absolutely no way to answer this question. I'm also not aware that en passant was instituted specifically for this reason - do you have a source for this?

But leaving that aside, what draw percentage would you regard as acceptable, if 5% is too high?

It's fairly self-evident. In a game without the en passant, pawns can move past each other without danger and create an unbroken chain on each side that can't be breached.

Before endgame, and up material? 0%.

You keep coming back to the idea of being up material, even though it has been pointed out, repeatedly, that being up material (particularly if you're including being up a single pawn) doesn't necessarily equate to an over the board advantage. This is a real problem for your rule that you haven't addressed.

A target draw rate of 0% is, I would say, entirely unrealistic for a game that derives a significant amount of its appeal from the fact that it starts in almost perfect balance. A well-fought draw being an option isn't a bug that needs to be fixed; it's a feature. I wonder if there is a cultural aspect here - people from the US seem to have a really low tolerance for draws that I think the rest of the world simply doesn't share. Personally speaking, I find overtime in the NFL (where the game is quite often effectively decided by the result of the coin toss at the start of overtime) more unsatisfactory than just letting the game end as a draw - at least in the regular season where a decisive result isn't a necessity.

Martin_Stahl
MD1806 wrote:
Martin_Stahl wrote:

5 draws out of every 100 games is a problem?

What percentage of games ended in locked-up boards before the en passant rule was instituted?

None as far as I'm aware. En passant was created as part of the change to the initial pawn move being two squares as an option.

MD1806
HonSec wrote:
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
Martin_Stahl wrote:

5 draws out of every 100 games is a problem?

What percentage of games ended in locked-up boards before the en passant rule was instituted?

Since the vast majority of games at the level of play you've specified would have been completely undocumented at the time, there is absolutely no way to answer this question. I'm also not aware that en passant was instituted specifically for this reason - do you have a source for this?

But leaving that aside, what draw percentage would you regard as acceptable, if 5% is too high?

It's fairly self-evident. In a game without the en passant, pawns can move past each other without danger and create an unbroken chain on each side that can't be breached.

Before endgame, and up material? 0%.

You keep coming back to the idea of being up material, even though it has been pointed out, repeatedly, that being up material (particularly if you're including being up a single pawn) doesn't necessarily equate to an over the board advantage. This is a real problem for your rule that you haven't addressed.

A target draw rate of 0% is, I would say, entirely unrealistic for a game that derives a significant amount of its appeal from the fact that it starts in almost perfect balance. A well-fought draw being an option isn't a bug that needs to be fixed; it's a feature. I wonder if there is a cultural aspect here - people from the US seem to have a really low tolerance for draws that I think the rest of the world simply doesn't share. Personally speaking, I find overtime in the NFL (where the game is quite often effectively decided by the result of the coin toss at the start of overtime) more unsatisfactory than just letting the game end as a draw - at least in the regular season where a decisive result isn't a necessity.

A game that ends in a move draw before endgame is, by definition, not well fought; it means both players gave up trying to accomplish the point of the game, which is to win.

HonSec
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
Martin_Stahl wrote:

5 draws out of every 100 games is a problem?

What percentage of games ended in locked-up boards before the en passant rule was instituted?

Since the vast majority of games at the level of play you've specified would have been completely undocumented at the time, there is absolutely no way to answer this question. I'm also not aware that en passant was instituted specifically for this reason - do you have a source for this?

But leaving that aside, what draw percentage would you regard as acceptable, if 5% is too high?

A game that ends in a move draw before endgame is, by definition, not well fought; it means both players gave up trying to accomplish the point of the game, which is to win.

You say that as though sub-2000s are out here routinely making the equivalent of Berlin draws. I really don't think that's the case. I would say that the overwhelming majority of threefold repetition draws by players of that calibre are either a defending player finding a perpetual check to snatch a draw from the jaws of defeat, or players shuffling pieces back and forth in a game that is either a dead draw or where any move other than the repetition loses immediately.

Once again, of course, the fundamental problem here is that you haven't provided any meaningful data to support your argument. Martin and I have shown that draws by repetition are rare. Draws by repetition in the middle game are presumably even rarer. And draws by repetition in the middle game where one player has a tangible material advantage are rarer still. So we're talking about a fraction of a fraction of 5%, which hardly seems to need a rule change to me.

And then of course we get to the problems of the proposed rule change itself, which you still haven't addressed.

MD1806
HonSec wrote:
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
Martin_Stahl wrote:

5 draws out of every 100 games is a problem?

What percentage of games ended in locked-up boards before the en passant rule was instituted?

Since the vast majority of games at the level of play you've specified would have been completely undocumented at the time, there is absolutely no way to answer this question. I'm also not aware that en passant was instituted specifically for this reason - do you have a source for this?

But leaving that aside, what draw percentage would you regard as acceptable, if 5% is too high?

A game that ends in a move draw before endgame is, by definition, not well fought; it means both players gave up trying to accomplish the point of the game, which is to win.

You say that as though sub-2000s are out here routinely making the equivalent of Berlin draws. I really don't think that's the case. I would say that the overwhelming majority of threefold repetition draws by players of that calibre are either a defending player finding a perpetual check to snatch a draw from the jaws of defeat, or players shuffling pieces back and forth in a game that is either a dead draw or where any move other than the repetition loses immediately.

Once again, of course, the fundamental problem here is that you haven't provided any meaningful data to support your argument. Martin and I have shown that draws by repetition are rare. Draws by repetition in the middle game are presumably even rarer. And draws by repetition in the middle game where one player has a tangible material advantage are rarer still. So we're talking about a fraction of a fraction of 5%, which hardly seems to need a rule change to me.

And then of course we get to the problems of the proposed rule change itself, which you still haven't addressed.

This has not been my experience. My experience has been that early/mid-game move draws are the result of a lack of imagination on the part of the player with a material advantage who assents to them. Basically, they're unwilling to give up their one obvious line of attack, so they take a draw. It's a bit like the people who can only sacrifice to attack and who can't function when the sacrifice fails.

HonSec
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
Martin_Stahl wrote:

5 draws out of every 100 games is a problem?

What percentage of games ended in locked-up boards before the en passant rule was instituted?

Since the vast majority of games at the level of play you've specified would have been completely undocumented at the time, there is absolutely no way to answer this question. I'm also not aware that en passant was instituted specifically for this reason - do you have a source for this?

But leaving that aside, what draw percentage would you regard as acceptable, if 5% is too high?

A game that ends in a move draw before endgame is, by definition, not well fought; it means both players gave up trying to accomplish the point of the game, which is to win.

You say that as though sub-2000s are out here routinely making the equivalent of Berlin draws. I really don't think that's the case. I would say that the overwhelming majority of threefold repetition draws by players of that calibre are either a defending player finding a perpetual check to snatch a draw from the jaws of defeat, or players shuffling pieces back and forth in a game that is either a dead draw or where any move other than the repetition loses immediately.

Once again, of course, the fundamental problem here is that you haven't provided any meaningful data to support your argument. Martin and I have shown that draws by repetition are rare. Draws by repetition in the middle game are presumably even rarer. And draws by repetition in the middle game where one player has a tangible material advantage are rarer still. So we're talking about a fraction of a fraction of 5%, which hardly seems to need a rule change to me.

And then of course we get to the problems of the proposed rule change itself, which you still haven't addressed.

This has not been my experience. My experience has been that early/mid-game move draws are the result of a lack of imagination on the part of the player with a material advantage who assents to them. Basically, they're unwilling to give up their one obvious line of attack, so they take a draw. It's a bit like the people who can only sacrifice to attack and who can't function when the sacrifice fails.

Your experience has also been that such draws happen very rarely. Looking at your blitz stats, you have had only 4 draws by repetition before move 36 where the game ended with a material imbalance. That's just 0.73% of your games. I gave also had 4 such draws, and for me that represents 0.51% of my games. So, assuming we aren't significant outliers (a reasonable assumption I would say, since our numbers are so similar) your rule would change the result of fewer than 1 in 100 games. Is that really worth a rule change of the magnitude you're proposing?

MD1806
HonSec wrote:
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
HonSec wrote:
MD1806 wrote:
Martin_Stahl wrote:

5 draws out of every 100 games is a problem?

What percentage of games ended in locked-up boards before the en passant rule was instituted?

Since the vast majority of games at the level of play you've specified would have been completely undocumented at the time, there is absolutely no way to answer this question. I'm also not aware that en passant was instituted specifically for this reason - do you have a source for this?

But leaving that aside, what draw percentage would you regard as acceptable, if 5% is too high?

A game that ends in a move draw before endgame is, by definition, not well fought; it means both players gave up trying to accomplish the point of the game, which is to win.

You say that as though sub-2000s are out here routinely making the equivalent of Berlin draws. I really don't think that's the case. I would say that the overwhelming majority of threefold repetition draws by players of that calibre are either a defending player finding a perpetual check to snatch a draw from the jaws of defeat, or players shuffling pieces back and forth in a game that is either a dead draw or where any move other than the repetition loses immediately.

Once again, of course, the fundamental problem here is that you haven't provided any meaningful data to support your argument. Martin and I have shown that draws by repetition are rare. Draws by repetition in the middle game are presumably even rarer. And draws by repetition in the middle game where one player has a tangible material advantage are rarer still. So we're talking about a fraction of a fraction of 5%, which hardly seems to need a rule change to me.

And then of course we get to the problems of the proposed rule change itself, which you still haven't addressed.

This has not been my experience. My experience has been that early/mid-game move draws are the result of a lack of imagination on the part of the player with a material advantage who assents to them. Basically, they're unwilling to give up their one obvious line of attack, so they take a draw. It's a bit like the people who can only sacrifice to attack and who can't function when the sacrifice fails.

Your experience has also been that such draws happen very rarely. Looking at your blitz stats, you have had only 4 draws by repetition before move 36 where the game ended with a material imbalance. That's just 0.73% of your games. I gave also had 4 such draws, and for me that represents 0.51% of my games. So, assuming we aren't significant outliers (a reasonable assumption I would say, since our numbers are so similar) your rule would change the result of fewer than 1 in 100 games. Is that really worth a rule change of the magnitude you're proposing?

4 is 4 too many. After all, this site doesn't follow FIDE rules. I've never played anywhere else where my time starts before I make a move as white...