True or False Chess is a Draw with Best Play from Both Sides

Sort:
SmyslovFan

It can probably be dropped a few more "0s" due to the repetitive nature of the moves considered. But yeah, dropping it an exponent or two does make a difference.

TheOldReb

Tinsley defeated the Chinook beast , I forget the year but a google search should show it . Imagine 40 years of competitive checkers and only losing 7 games !  Two of those losses were to Chinook I believe so he only lost 5 games in 40 years to humans !  Amazing .... and yes  " 3 move restriction " is what ponz is talking about , I think its the most common form of competitive checkers these days but there are still GAYP tournies =  go as you please / no restrictions   but too many of those games are draws , thus 3 move restriction ... 

pfren
SmyslovFan wrote:

Yeah, checkers has been solved. It's a draw.

The official 10*10 checkers has not been solved (yet).

TheGrobe

We don't.

ponz111

The stronger the chess player, the more he knows Black is not winning.

TheGrobe

"Knowing" in this context, is binary, it is not a sliding scale.

The stronger the chess player, the stronger his conviction that Black is not winning.

eoJ1

I haven't read the 99 pages of this thread obviously, so not sure if this has been brought up, but I'll use chess.com Game explorer.

We can firstly assume that both black and white are equally likely to make a blunder, obviously (other than one given by the nature of their colour, which would count against the OP's theory).

There are 698k games in the database starting e4. However, the outcome over these is 38.5% white win, 31.5% draw, 30% black win.

For d4, there are 539k games, 38.9% white win, 34.1% draw, 26.9% black wins.

Overall, 1.237M games.

Combining those two openings, we get:

38.7%, 32.6%, 28.6%

Meaning that white will get a win 6.1% more of the time than they get a draw. I would imagine such a large difference, over that large number of high level games is statistically significant.

SmyslovFan

Still waiting for even one example of a game which was won and no mistake was made. 

Just one.

TheGrobe

How would you know whether it was a mistake or not?  A bit of a circular argument there.

ponz111

eoJ1  yes, that large of a difference is statistically significant. It indicates that it is easier to win with White rather than Black.

But, of course, your opponent has to make a mistake for you to win.

ponz111
TheGrobe wrote:

"Knowing" in this context, is binary, it is not a sliding scale.

The stronger the chess player, the stronger his conviction that Black is not winning.

To me, and to many players,  if one believes the chance that Black is winning is less than 1/100th of 1 percent, that is pragmatically very close to zero.

One could say, we do not know anything with 100 per cent conviction.

There is a chance that you are a figment of my imagination. 

ponz111
TheGrobe wrote:

How would you know whether it was a mistake or not?  A bit of a circular argument there.

A strong enough player, with the help of a strong chess engine would know. Not only that, he could point out the mistake.

eoJ1

As TheGrobe said, the argument is circular.

ponz111 wrote:

eoJ1  yes, that large of a difference is statistically significant. It indicates that it is easier to win with White rather than Black.

But, of course, your opponent has to make a mistake for you to win.

That depends on how we define a mistake. What makes a move become one? If two players follow exactly the most played moves of 100s of thousands of top ranked players before them, the % of wins for white only increases for white. Surely the combined knowledge and trials of this massive amount of top level people demonstrates the advantage of white?

ponz111
eoJ1 wrote:

As TheGrobe said, the argument is circular.

ponz111 wrote:

eoJ1  yes, that large of a difference is statistically significant. It indicates that it is easier to win with White rather than Black.

But, of course, your opponent has to make a mistake for you to win.

That depends on how we define a mistake. What makes a move become one? If two players follow exactly the most played moves of 100s of thousands of top ranked players before them, the % of wins for white only increases for white. Surely the combined knowledge and trials of this massive amount of top level people demonstrates the advantage of white?

AS, I mentioned many times--in this forum we are defining "mistake" or "error" as a move which would change the outcome of the game if your opponent makes the correct moves.

One could define "mistake" as a move which gives less chance of a win but we are not using that definition here.

Frequency of moves played do not define a move as a good move or a bad move. It takes more than that.

I could give a position where a move has been played 1000's of times by White--and on the 4th move--but it loses with the correct responses by Black.

Do not make the assumption that moves played frequently are always the best moves or even the moves without a mistake. 

ponz111

White has an advantage of maybe 1/4 of a pawn to start. This is not enough to win. However having that advantage makes it more likely that White will win.

eoJ1
ponz111 wrote:
TheGrobe wrote:

How would you know whether it was a mistake or not?  A bit of a circular argument there.

A strong enough player, with the help of a strong chess engine would know. Not only that, he could point out the mistake.

A chess engine that's analysed anywhere near all 10^123 possible games? You got a link to one?

eoJ1
ponz111 wrote:
eoJ1 wrote:

As TheGrobe said, the argument is circular.

ponz111 wrote:

eoJ1  yes, that large of a difference is statistically significant. It indicates that it is easier to win with White rather than Black.

But, of course, your opponent has to make a mistake for you to win.

That depends on how we define a mistake. What makes a move become one? If two players follow exactly the most played moves of 100s of thousands of top ranked players before them, the % of wins for white only increases for white. Surely the combined knowledge and trials of this massive amount of top level people demonstrates the advantage of white?

AS, I mentioned many times--in this forum we are defining "mistake" or "error" as a move which would change the outcome of the game if your opponent makes the correct moves.

One could define "mistake" as a move which gives less chance of a win but we are not using that definition here.

Frequency of moves played do not define a move as a good move or a bad move. It takes more than that.

I could give a position where a move has been played 1000's of times by White--and on the 4th move--but it loses with the correct responses by Black.

Do not make the assumption that moves played frequently are always the best moves or even the moves without a mistake. 

The percentage of times that people at master level and above play a move may not automatically make it a good move, but chances are that it is.


If there's a natural advantage, to white, then any opening move for black is a mistake. To quote Wargames, the only way to win [draw] is not to play!

Finally, RE white having an advantage of 1/4 of a pawn - I'd say that's incorrect. Each side has 38 points of material at the start. Over 1.237m games, White wins 6.1% more. 6.1% * 38 makes for a 2.32 point advantage.

SmyslovFan

How's this: for all those thinking they've found a game where there isn't a mistake, post it and see if others can find a mistake in the play. 

It's not circular logic at all. It's what is needed in order to disprove that chess is a draw with best play. If anything, it's definitional. Either chess is a draw and there's no such thing as a perfectly played game that is decisive, or chess is not a draw and there are examples of perfect play (by both sides) leading to a decisive result.

Show us just ONE example of a perfectly played game that ended decisively. 

Just one.

TheGrobe

You cannot know with 100% certainty whether any given move is a mistake or not without exhaustively examining the complete subsequent game tree.

There is an entire class of positions for which this knowlege is in fact certain: all games with seven or less total pieces as per the Lomonosov Endgame Tablebase, and the trival positions (i.e. mates in one or two etc.) with more than seven peices.

It is a very, very small subset of the total number of positions an also a very, very small subset of the set of positions in any given game.

SmyslovFan

"You cannot know with 100% certainty whether any given move is a mistake or not without exhaustively examining the complete subsequent game tree.

There is an entire class of positions for which this knowlege is in fact certain: all games with seven or less total pieces as per the Lomonosov Endgame Tablebase, and the trival positions (i.e. mates in one or two etc.) with more than seven peices.

It is a very, very small subset of the total number of positions an also a very, very small subset of the set of positions in any given game."

 

 

Give us just one example. We can see with our limited skills whether it's a mistake.

Your very notion that we can't tell what a mistake is shows your lack of chess understanding.