Bullet masters. Pro Deo 1.0 - 2557.

[Igor Korshunov]

Geschrieben von:/Posted by: Igor Korshunov at 02. September 2004 07:36:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Dan Honeycutt at 01 September 2004 21:21:38:

Seems it is very unbalanced engine and clearly is not the best of free available.

You're reaching this kind of conclusion based on bullet games? Get real!!

The best engine must play well at all suitable time controls.
Imho, it is clear.
Do you have another opinion?

Yes, i do. Quite simply, i don't believe bullet is a suitable time control.
Regards
Dave

This is strange.
Do you prefer time control one move per 100 years?
I don't think that this time control is slow enough, because engines will make a lot of mistakes with such a 'bullet' control.
I don't see a big difference between 'game in 1 min' and 'game in 2 hours'.
From my point of view importance of be the best at bullet not smaller than the importance of be the best at slow.
What is the reasons to consider slow is more important?
I heard only one: engines make less mistakes. I don't like such opinion.
But at WBEC engines do so much mistakes, that seems bullet time control used.
At least clear is not slow enough.
From another point: I think the more mistake the more interesting game. Who want to see all games ended in a draw, without any mistakes?
Finally, I think that tournaments with different time controls have the same importance.

Engine search depth vs time is exponential and the probability that the engine will change it's move selection generally follows the slope of that curve - high initially and then decreasing toward zero. (I say generally - there are some positions where the engine flips back and forth between two moves). The engine has to get over the knee and out on the flat part of the curve before you are reasonably sure that it has made up it's mind. Bullet time controls may not allow that.
Volker Pittlik is doing some research into the "fastest" time control that could be considered "slow" time control, ie the amount of time needed so you are 95 % sure the engine would not change it's mind even if given a lot more time. This will be useful data so debates like this thread can be more quantitative rather than qualitative.
Dan H.

If you think so, is this mean that playing games on modern PC is not reasonable?
Because in more than 99% positions ANY engine will don't choose bad moves, if given enough time.
Exceptions only due to some heuristics like null move, because some engines can never find right move.
Follow your logic the most games must be draws, if reasonable time control used.

I CAN'T call such time control reasonable.
For me most reasonable time conrol where engines doing more mistakes. But of course time trouble should not be a decisive factor.
On modern PC it possible play 'game in 1' without getting into time trouble in most games.
For clearness, from my point of view time trouble is danger of lose on time.
If you sure that you can do 50 moves you are not in time trouble.

[Jason Kent]

Geschrieben von:/Posted by: Jason Kent at 02 September 2004 13:26:07:
Als Antwort auf:/In reply to: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Igor Korshunov at 29. August 2004 13:17:

Hi Igor,
I just wanted to let you know I agree that fast games should be included in the testing of chess programs. Various time controls and various opponets bring a more accurate rating. Bullet games may or may not be more luck... I've not seen significant evidence. I do know that with bullet time controls, you get many more games in a short amount of time.

[David Dahlem]

Geschrieben von:/Posted by: David Dahlem at 02 September 2004 15:45:24:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Igor Korshunov at 02. September 2004 07:34:

Seems it is very unbalanced engine and clearly is not the best of free available.

You're reaching this kind of conclusion based on bullet games? Get real!!

The best engine must play well at all suitable time controls.
Imho, it is clear.
Do you have another opinion?

Yes, i do. Quite simply, i don't believe bullet is a suitable time control.
Regards
Dave

This is strange.
Do you prefer time control one move per 100 years?

Of course not, this is an extreme, foolish time control, and game in 1 minute is an extreme, foolish time control.
Dave

I don't see nothing extreme in 'game in 1 min'.
There are a lot of time for engine.
Of course, for human 1 min is highly extreme. But we are talking about engines!
On modern PC 1 min is more than enough to avoid any time trouble.
From another point: 2 hours per move on modern PC is clearly not enough to produce good enough moves.

Your logic is crazy. If, as you say, 2 hours per move will not produce good enough moves, how in the world can you justify the game in 1 minute time control? Absolutely unbelievable logic!
Dave

[Dan Honeycutt]

Geschrieben von:/Posted by: Dan Honeycutt at 02 September 2004 17:55:26:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Igor Korshunov at 02. September 2004 07:36:

If you think so, is this mean that playing games on modern PC is not reasonable?
Because in more than 99% positions ANY engine will don't choose bad moves, if given enough time.
Exceptions only due to some heuristics like null move, because some engines can never find right move.
Follow your logic the most games must be draws, if reasonable time control used.

I CAN'T call such time control reasonable.
For me most reasonable time conrol where engines doing more mistakes. But of course time trouble should not be a decisive factor.
On modern PC it possible play 'game in 1' without getting into time trouble in most games.
For clearness, from my point of view time trouble is danger of lose on time.
If you sure that you can do 50 moves you are not in time trouble.

We don't seem to be talking about the same subject. I'm talking about a time control long enough so the engine is likely to select the same move that it would select if given appreciably more time. Has nothing to do with games on a modern PC, most games drawn, or losses on time.
How long is that time control? I don't know but it is of much interest to me for engine development where the goal is to determine if a change is an improvement. I know bullet doesn't tell me anything. I also know 40 moves in 2 hours is much longer than necessary.
Dan H.

[Uri Blass]

Geschrieben von:/Posted by: Uri Blass at 02 September 2004 18:23:39:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Dan Honeycutt at 02 September 2004 17:55:26:

If you think so, is this mean that playing games on modern PC is not reasonable?
Because in more than 99% positions ANY engine will don't choose bad moves, if given enough time.
Exceptions only due to some heuristics like null move, because some engines can never find right move.
Follow your logic the most games must be draws, if reasonable time control used.

I CAN'T call such time control reasonable.
For me most reasonable time conrol where engines doing more mistakes. But of course time trouble should not be a decisive factor.
On modern PC it possible play 'game in 1' without getting into time trouble in most games.
For clearness, from my point of view time trouble is danger of lose on time.
If you sure that you can do 50 moves you are not in time trouble.

We don't seem to be talking about the same subject. I'm talking about a time control long enough so the engine is likely to select the same move that it would select if given appreciably more time. Has nothing to do with games on a modern PC, most games drawn, or losses on time.
How long is that time control? I don't know but it is of much interest to me for engine development where the goal is to determine if a change is an improvement. I know bullet doesn't tell me anything. I also know 40 moves in 2 hours is much longer than necessary.
Dan H.

I guess that if you double the time the engine has more than 5% chance to change it's mind and the time that you choose is not important(I assume that you choose time control that it is possible to finish games).
I will even guess that if you double the time the engine has more than 5% chance to choose a move that it never considered before as best in every reasonable time control that you will try.
Uri

[Dan Honeycutt]

Geschrieben von:/Posted by: Dan Honeycutt at 02 September 2004 18:54:06:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Uri Blass at 02 September 2004 18:23:39:

I guess that if you double the time the engine has more than 5% chance to change it's mind and the time that you choose is not important(I assume that you choose time control that it is possible to finish games).
I will even guess that if you double the time the engine has more than 5% chance to choose a move that it never considered before as best in every reasonable time control that you will try.
Uri

I don't know. I think you'd be exactly right if new tactics surfaced linearly as a function of depth. But I suspect that new tactics vs depth may have an exponential behavior.
Dan H.

[Uri Blass]

Geschrieben von:/Posted by: Uri Blass at 02 September 2004 19:12:40:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Dan Honeycutt at 02 September 2004 18:54:06:

I guess that if you double the time the engine has more than 5% chance to change it's mind and the time that you choose is not important(I assume that you choose time control that it is possible to finish games).
I will even guess that if you double the time the engine has more than 5% chance to choose a move that it never considered before as best in every reasonable time control that you will try.
Uri

I don't know. I think you'd be exactly right if new tactics surfaced linearly as a function of depth. But I suspect that new tactics vs depth may have an exponential behavior.
Dan H.

This problem was investigated in the past and people found that even at big depth there is often a change in the best move
Here are results
http://supertech.lcs.mit.edu/~heinz/dt/node54.html
You can see that even at depth 14 for Crafty there were change in the best move in 15% of the cases and in 40% of the cases the change was to a move that was not considered before as best.
I know from experience in correspondence games that cases when a program changes it's mind to a new move after many hours is not rare.
Uri

[Dan Honeycutt]

Geschrieben von:/Posted by: Dan Honeycutt at 02 September 2004 19:23:27:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Uri Blass at 02 September 2004 19:12:40:

I guess that if you double the time the engine has more than 5% chance to change it's mind and the time that you choose is not important(I assume that you choose time control that it is possible to finish games).
I will even guess that if you double the time the engine has more than 5% chance to choose a move that it never considered before as best in every reasonable time control that you will try.
Uri

I don't know. I think you'd be exactly right if new tactics surfaced linearly as a function of depth. But I suspect that new tactics vs depth may have an exponential behavior.
Dan H.

This problem was investigated in the past and people found that even at big depth there is often a change in the best move
Here are results
http://supertech.lcs.mit.edu/~heinz/dt/node54.html
You can see that even at depth 14 for Crafty there were change in the best move in 15% of the cases and in 40% of the cases the change was to a move that was not considered before as best.
I know from experience in correspondence games that cases when a program changes it's mind to a new move after many hours is not rare.
Uri

I didn't say it would ever stop changing it's mind. Taking the point in the search where your 5% is correct: I double the time with a 5% chance of a change. I double the time again but now I get a 3% chance of a change. Another doubling with a 2% chance, etc.
The link you posted suggests exactly that sort of behavior.
Dan H.

[Uri Blass]

Geschrieben von:/Posted by: Uri Blass at 02 September 2004 20:19:28:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Dan Honeycutt at 02 September 2004 19:23:27:

I guess that if you double the time the engine has more than 5% chance to change it's mind and the time that you choose is not important(I assume that you choose time control that it is possible to finish games).
I will even guess that if you double the time the engine has more than 5% chance to choose a move that it never considered before as best in every reasonable time control that you will try.
Uri

I don't know. I think you'd be exactly right if new tactics surfaced linearly as a function of depth. But I suspect that new tactics vs depth may have an exponential behavior.
Dan H.

This problem was investigated in the past and people found that even at big depth there is often a change in the best move
Here are results
http://supertech.lcs.mit.edu/~heinz/dt/node54.html
You can see that even at depth 14 for Crafty there were change in the best move in 15% of the cases and in 40% of the cases the change was to a move that was not considered before as best.
I know from experience in correspondence games that cases when a program changes it's mind to a new move after many hours is not rare.
Uri

I didn't say it would ever stop changing it's mind. Taking the point in the search where your 5% is correct: I double the time with a 5% chance of a change. I double the time again but now I get a 3% chance of a change. Another doubling with a 2% chance, etc.
The link you posted suggests exactly that sort of behavior.
Dan H.

The link suggests that probability of more than 95% not to change it's mind when you double the thinking time will not happen in a reasonable time.
It is not going 5% 3% 2% and clearly go down more slowly based on the link.
I see that I was wrong in suggesting that in more than 5% of the cases it s going to suggest a new move that it did not suggest earlier.
Uri

[Igor Korshunov]

Geschrieben von:/Posted by: Igor Korshunov at 02. September 2004 23:38:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: David Dahlem at 02 September 2004 15:45:24:

Seems it is very unbalanced engine and clearly is not the best of free available.

You're reaching this kind of conclusion based on bullet games? Get real!!

The best engine must play well at all suitable time controls.
Imho, it is clear.
Do you have another opinion?

Yes, i do. Quite simply, i don't believe bullet is a suitable time control.
Regards
Dave

This is strange.
Do you prefer time control one move per 100 years?

Of course not, this is an extreme, foolish time control, and game in 1 minute is an extreme, foolish time control.
Dave

I don't see nothing extreme in 'game in 1 min'.
There are a lot of time for engine.
Of course, for human 1 min is highly extreme. But we are talking about engines!
On modern PC 1 min is more than enough to avoid any time trouble.
From another point: 2 hours per move on modern PC is clearly not enough to produce good enough moves.

Your logic is crazy. If, as you say, 2 hours per move will not produce good enough moves, how in the world can you justify the game in 1 minute time control? Absolutely unbelievable logic!
Dave

I am not interested in moves quality, so bullet is most suitable for me.
If you interested in moves quality, why you use so fast time controls?

[Dan Honeycutt]

Geschrieben von:/Posted by: Dan Honeycutt at 03 September 2004 03:22:39:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Uri Blass at 02 September 2004 20:19:28:

I guess that if you double the time the engine has more than 5% chance to change it's mind and the time that you choose is not important(I assume that you choose time control that it is possible to finish games).
I will even guess that if you double the time the engine has more than 5% chance to choose a move that it never considered before as best in every reasonable time control that you will try.
Uri

I don't know. I think you'd be exactly right if new tactics surfaced linearly as a function of depth. But I suspect that new tactics vs depth may have an exponential behavior.
Dan H.

This problem was investigated in the past and people found that even at big depth there is often a change in the best move
Here are results
http://supertech.lcs.mit.edu/~heinz/dt/node54.html
You can see that even at depth 14 for Crafty there were change in the best move in 15% of the cases and in 40% of the cases the change was to a move that was not considered before as best.
I know from experience in correspondence games that cases when a program changes it's mind to a new move after many hours is not rare.
Uri

I didn't say it would ever stop changing it's mind. Taking the point in the search where your 5% is correct: I double the time with a 5% chance of a change. I double the time again but now I get a 3% chance of a change. Another doubling with a 2% chance, etc.
The link you posted suggests exactly that sort of behavior.
Dan H.

The link suggests that probability of more than 95% not to change it's mind when you double the thinking time will not happen in a reasonable time.
It is not going 5% 3% 2% and clearly go down more slowly based on the link.
I see that I was wrong in suggesting that in more than 5% of the cases it s going to suggest a new move that it did not suggest earlier.
Uri

It's not there at 14 ply since 4 or 5 percent of all changes occured at that ply.
OK. 5, 3, 2 were just illustration.
Looking for a time control where the engine has "reasonably well" made up it's mind I'd want to consider any change. But this has a bad drawback if the engine is waffling (going back and forth from ply to ply between two near equal moves). In that case the move selection is a crap shoot unrelated to time control. So if it's waffling I'd want to only consider a change to a fresh move. How you measure that I don't know.
Dan H.

[Albert Silver]

Geschrieben von:/Posted by: Albert Silver at 03 September 2004 19:31:16:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Dan Honeycutt at 01 September 2004 21:21:38:

Seems it is very unbalanced engine and clearly is not the best of free available.

You're reaching this kind of conclusion based on bullet games? Get real!!

The best engine must play well at all suitable time controls.
Imho, it is clear.
Do you have another opinion?

Yes, i do. Quite simply, i don't believe bullet is a suitable time control.
Regards
Dave

This is strange.
Do you prefer time control one move per 100 years?
I don't think that this time control is slow enough, because engines will make a lot of mistakes with such a 'bullet' control.
I don't see a big difference between 'game in 1 min' and 'game in 2 hours'.
From my point of view importance of be the best at bullet not smaller than the importance of be the best at slow.
What is the reasons to consider slow is more important?
I heard only one: engines make less mistakes. I don't like such opinion.
But at WBEC engines do so much mistakes, that seems bullet time control used.
At least clear is not slow enough.
From another point: I think the more mistake the more interesting game. Who want to see all games ended in a draw, without any mistakes?
Finally, I think that tournaments with different time controls have the same importance.

Engine search depth vs time is exponential and the probability that the engine will change it's move selection generally follows the slope of that curve - high initially and then decreasing toward zero. (I say generally - there are some positions where the engine flips back and forth between two moves). The engine has to get over the knee and out on the flat part of the curve before you are reasonably sure that it has made up it's mind. Bullet time controls may not allow that.
Volker Pittlik is doing some research into the "fastest" time control that could be considered "slow" time control, ie the amount of time needed so you are 95 % sure the engine would not change it's mind even if given a lot more time. This will be useful data so debates like this thread can be more quantitative rather than qualitative.
Dan H.

A similar study has already been done and published by Ernst Heinz on diminishing returns. This pretty much answers the question as it deals with the curve at which point greater depth (hence longer time controls) yields less and less gains. You can find it by searching the CCC archives.
Albert

[Albert Silver]

Geschrieben von:/Posted by: Albert Silver at 03 September 2004 19:42:57:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Igor Korshunov at 02. September 2004 07:36:

Seems it is very unbalanced engine and clearly is not the best of free available.

You're reaching this kind of conclusion based on bullet games? Get real!!

The best engine must play well at all suitable time controls.
Imho, it is clear.
Do you have another opinion?

Yes, i do. Quite simply, i don't believe bullet is a suitable time control.
Regards
Dave

This is strange.
Do you prefer time control one move per 100 years?
I don't think that this time control is slow enough, because engines will make a lot of mistakes with such a 'bullet' control.
I don't see a big difference between 'game in 1 min' and 'game in 2 hours'.
From my point of view importance of be the best at bullet not smaller than the importance of be the best at slow.
What is the reasons to consider slow is more important?
I heard only one: engines make less mistakes. I don't like such opinion.
But at WBEC engines do so much mistakes, that seems bullet time control used.
At least clear is not slow enough.
From another point: I think the more mistake the more interesting game. Who want to see all games ended in a draw, without any mistakes?
Finally, I think that tournaments with different time controls have the same importance.

Engine search depth vs time is exponential and the probability that the engine will change it's move selection generally follows the slope of that curve - high initially and then decreasing toward zero. (I say generally - there are some positions where the engine flips back and forth between two moves). The engine has to get over the knee and out on the flat part of the curve before you are reasonably sure that it has made up it's mind. Bullet time controls may not allow that.
Volker Pittlik is doing some research into the "fastest" time control that could be considered "slow" time control, ie the amount of time needed so you are 95 % sure the engine would not change it's mind even if given a lot more time. This will be useful data so debates like this thread can be more quantitative rather than qualitative.
Dan H.

If you think so, is this mean that playing games on modern PC is not reasonable?
Because in more than 99% positions ANY engine will don't choose bad moves, if given enough time.
Exceptions only due to some heuristics like null move, because some engines can never find right move.
Follow your logic the most games must be draws, if reasonable time control used.
I CAN'T call such time control reasonable.
For me most reasonable time conrol where engines doing more mistakes.
But of course time trouble should not be a decisive factor.
On modern PC it possible play 'game in 1' without getting into time trouble in most games.
For clearness, from my point of view time trouble is danger of lose on time.
If you sure that you can do 50 moves you are not in time trouble.

That's highly questionable. Perhaps it depends on what you call bad moves.
But of course that's not the case. Perhaps you have some evidence that shows that there are fewer decisive games in 30 minute games than 2 minute games. I am testing some settings with Pro Deo against Junior 8 in 30 minute games using the Nunn 2 set. 13 games so far and only 2 draws.
Why? What does this prove?
Albert

[David Dahlem]

Geschrieben von:/Posted by: David Dahlem at 03 September 2004 21:18:54:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Igor Korshunov at 02. September 2004 23:38:

Seems it is very unbalanced engine and clearly is not the best of free available.

You're reaching this kind of conclusion based on bullet games? Get real!!

The best engine must play well at all suitable time controls.
Imho, it is clear.
Do you have another opinion?

Yes, i do. Quite simply, i don't believe bullet is a suitable time control.
Regards
Dave

This is strange.
Do you prefer time control one move per 100 years?

Of course not, this is an extreme, foolish time control, and game in 1 minute is an extreme, foolish time control.
Dave

I don't see nothing extreme in 'game in 1 min'.
There are a lot of time for engine.
Of course, for human 1 min is highly extreme. But we are talking about engines!
On modern PC 1 min is more than enough to avoid any time trouble.
From another point: 2 hours per move on modern PC is clearly not enough to produce good enough moves.

Your logic is crazy. If, as you say, 2 hours per move will not produce good enough moves, how in the world can you justify the game in 1 minute time control? Absolutely unbelievable logic!
Dave

I am not interested in moves quality, so bullet is most suitable for me.

You're not interested in move quality?? Then why did you open this thread with this comment?
"Seems it is very unbalanced engine and clearly is not the best of free available."
Dave

[Dan Honeycutt]

Geschrieben von:/Posted by: Dan Honeycutt at 04 September 2004 03:24:58:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Albert Silver at 03 September 2004 19:31:16:

A similar study has already been done and published by Ernst Heinz on diminishing returns. This pretty much answers the question as it deals with the curve at which point greater depth (hence longer time controls) yields less and less gains. You can find it by searching the CCC archives.
Albert

Hi Albert:
Is this the same data that Uri posted a link to further down in the thread?
Best
Dan H

[Igor Korshunov]

Geschrieben von:/Posted by: Igor Korshunov at 04. September 2004 05:13:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: David Dahlem at 03 September 2004 21:18:54:

Seems it is very unbalanced engine and clearly is not the best of free available.

You're reaching this kind of conclusion based on bullet games? Get real!!

The best engine must play well at all suitable time controls.
Imho, it is clear.
Do you have another opinion?

Yes, i do. Quite simply, i don't believe bullet is a suitable time control.
Regards
Dave

This is strange.
Do you prefer time control one move per 100 years?

Of course not, this is an extreme, foolish time control, and game in 1 minute is an extreme, foolish time control.
Dave

I don't see nothing extreme in 'game in 1 min'.
There are a lot of time for engine.
Of course, for human 1 min is highly extreme. But we are talking about engines!
On modern PC 1 min is more than enough to avoid any time trouble.
From another point: 2 hours per move on modern PC is clearly not enough to produce good enough moves.

Your logic is crazy. If, as you say, 2 hours per move will not produce good enough moves, how in the world can you justify the game in 1 minute time control? Absolutely unbelievable logic!
Dave

I am not interested in moves quality, so bullet is most suitable for me.

You're not interested in move quality?? Then why did you open this thread with this comment?
"Seems it is very unbalanced engine and clearly is not the best of free available."
Dave

Because, I interested in engines strenght. Relatively to each other, of course

[Igor Korshunov]

Geschrieben von:/Posted by: Igor Korshunov at 04. September 2004 05:22:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Albert Silver at 03 September 2004 19:42:57:

Because in more than 99% positions ANY engine will don't choose bad moves, if given enough time.
Exceptions only due to some heuristics like null move, because some engines can never find right move.
Follow your logic the most games must be draws, if reasonable time control used.
I CAN'T call such time control reasonable.
For me most reasonable time conrol where engines doing more mistakes.

That's highly questionable. Perhaps it depends on what you call bad moves.
But of course that's not the case. Perhaps you have some evidence that shows that there are fewer decisive games in 30 minute games than 2 minute games. I am testing some settings with Pro Deo against Junior 8 in 30 minute games using the Nunn 2 set. 13 games so far and only 2 draws.
Why? What does this prove?

Bad move - this is losing move. It is very hard to find position where engine can't avoid lose, if it have enough time.
Yes. It is practically proven. Slower time control => more percent of draws.
I am talking not about several games, but about thousands.
Because time control which allowed engines make moves without mistakes totally useless.

[David Dahlem]

Geschrieben von:/Posted by: David Dahlem at 04 September 2004 16:37:40:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Igor Korshunov at 04. September 2004 05:13:

Seems it is very unbalanced engine and clearly is not the best of free available.

You're reaching this kind of conclusion based on bullet games? Get real!!

The best engine must play well at all suitable time controls.
Imho, it is clear.
Do you have another opinion?

Yes, i do. Quite simply, i don't believe bullet is a suitable time control.
Regards
Dave

This is strange.
Do you prefer time control one move per 100 years?

Of course not, this is an extreme, foolish time control, and game in 1 minute is an extreme, foolish time control.
Dave

I don't see nothing extreme in 'game in 1 min'.
There are a lot of time for engine.
Of course, for human 1 min is highly extreme. But we are talking about engines!
On modern PC 1 min is more than enough to avoid any time trouble.
From another point: 2 hours per move on modern PC is clearly not enough to produce good enough moves.

Your logic is crazy. If, as you say, 2 hours per move will not produce good enough moves, how in the world can you justify the game in 1 minute time control? Absolutely unbelievable logic!
Dave

I am not interested in moves quality, so bullet is most suitable for me.

You're not interested in move quality?? Then why did you open this thread with this comment?
"Seems it is very unbalanced engine and clearly is not the best of free available."
Dave

Because, I interested in engines strenght. Relatively to each other, of course

Hi Igor
I don't understand this thinking at all. Isn't engine strength and move quality related? I think it is.
Dave

[Albert Silver]

Geschrieben von:/Posted by: Albert Silver at 04 September 2004 19:48:44:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Igor Korshunov at 04. September 2004 05:22:

Because in more than 99% positions ANY engine will don't choose bad moves, if given enough time.
Exceptions only due to some heuristics like null move, because some engines can never find right move.
Follow your logic the most games must be draws, if reasonable time control used.
I CAN'T call such time control reasonable.
For me most reasonable time conrol where engines doing more mistakes.

That's highly questionable. Perhaps it depends on what you call bad moves.
But of course that's not the case. Perhaps you have some evidence that shows that there are fewer decisive games in 30 minute games than 2 minute games. I am testing some settings with Pro Deo against Junior 8 in 30 minute games using the Nunn 2 set. 13 games so far and only 2 draws.
Why? What does this prove?

Bad move - this is losing move.
It is very hard to find position where engine can't avoid lose, if it have enough time.
Yes. It is practically proven. Slower time control => more percent of draws.
I am talking not about several games, but about thousands.
Because time control which allowed engines make moves without mistakes totally useless.

I think this is a very primitive way of looking at chess with all due respect. I don't think computer chess is decided by tactical blunders by the respective opponents. I think it is much like grandmaster chess, in that a position is slowly built up and improved until a tactical sequence, if possible, appears on the board. For me, a bad move is a move that worsens the position, and not necessarily one that loses the game on the spot.
That is complete nonsense. I noticed that in the SSDF, Shredder 7.04 beat Ruffian 28-14. This was done at a time control of 40 moves in 2 hours, so obviously this wasn't enough time. How much time is necessary for Ruffian to stop losing so badly against Shredder?
If it's practically proven, it should be easy to point to evidence.
First, I find it amazing that anyone could think that engines don't make mistakes. Second, I can't even being to see the point of trying to find a time control where they make the most mistakes. You see, I never play bullet games against engines, so seeing the best bullet player is uninteresting to me, plus at bullet games, the engine will think for 1-2 seconds at most, if that, and I NEVER allow an engine to analyze a move for me for only 1-2 seconds. So if Ruffian can find the best moves in 1 second, and Pro Deo can find the best moves in 15-30 seconds or more, I will always analyze with Pro Deo since I couldn't care less about the results of 1 second of analysis.
Albert

[Albert Silver]

Geschrieben von:/Posted by: Albert Silver at 04 September 2004 21:03:41:
Als Antwort auf:/In reply to: Re: Bullet masters. Pro Deo 1.0 - 2557. geschrieben von:/posted by: Dan Honeycutt at 04 September 2004 03:24:58:

A similar study has already been done and published by Ernst Heinz on diminishing returns. This pretty much answers the question as it deals with the curve at which point greater depth (hence longer time controls) yields less and less gains. You can find it by searching the CCC archives.
Albert

Hi Albert:
Is this the same data that Uri posted a link to further down in the thread?
Best
Dan H

No, it is a different one. Here is the link to the thread from the CCC archives:
http://chessprogramming.org/cccsearch/c ... ead=112634
Albert