Probability to have repeated games

[towforce]

Could anyone answer, please, what is the probability to have repeated games in a sample of 1000 games, with a time control 1 minute/game?

This is not the question you asked, but I have a 1980s Tandy/Radio Shack chess computer with a small opening book. Once out of the opening, at a given time control, it always plays the same move in the same position - and I tended to do that as well, so to my annoyance I used to find myself playing the same games against it over and over.

[Milos]

This is not the question you asked, but I have a 1980s Tandy/Radio Shack chess computer with a small opening book. Once out of the opening, at a given time control, it always plays the same move in the same position - and I tended to do that as well, so to my annoyance I used to find myself playing the same games against it over and over.

Dedicated chess computers do not have multitasking and therefore, there is no difference other then the precision of the internal clock. Since it's a real time clock and interrupts are minimal, there should be always exactly the same time per move, which gives always the same moves.

[Steve B]

This is not the question you asked, but I have a 1980s Tandy/Radio Shack chess computer with a small opening book. Once out of the opening, at a given time control, it always plays the same move in the same position - and I tended to do that as well, so to my annoyance I used to find myself playing the same games against it over and over.

Dedicated chess computers do not have multitasking and therefore, there is no difference other then the precision of the internal clock. Since it's a real time clock and interrupts are minimal, there should be always exactly the same time per move, which gives always the same moves.

Many models of dedicated computers have a "random" feature whereby they will play different moves given the same position if the eval of the different moves are about the same
in some models the feature is automatic and in some models the random feature can be turned on/off

Steve

[kingliveson]

1000 Games Tested: Naum 4 vs Naum 4

The test is conducted using Fritz 11 GUI on Windows 7 64-bit. The hardware is an AMD Phenom II 940 overclocked to 3.6 GHz. The test runs continuously--none stop until it reaches 1000 games.

The engine (Naum 4) is 4 CPU 64 bit version with hash size set to 128 MB and pondering/permanent brain is disabled. Time control is 1 minute with no increment.

Naum 4 is renamed to Naum 4 E01 and Naum 4 E02 to accommodate the test. The engines use the same book renamed to 1000E01.ctg (Naum 4 E01) and 1000E02.ctg (Naum 4 E02). The book has only one predefined position with 20 half moves or 10 plies. The final position after exiting the book is:

r2qkb1r/1p1n1pp1/p2p1n2/3Pp2p/8/1N2BP2/PPP3PP/R2QKB1R w KQkq - 0 11

Result:

Games : 1000 (finished)

White Wins : 264 (26.4 %)
Black Wins : 256 (25.6 %)
Draws : 480 (48.0 %)

Final Outcome: No game is repeated.

[Graham Banks]

1000 Games Tested: Naum 4 vs Naum 4

The test is conducted using Fritz 11 GUI on Windows 7 64-bit. The hardware is an AMD Phenom II 940 overclocked to 3.6 GHz. The test runs continuously--none stop until it reaches 1000 games.

The engine (Naum 4) is 4 CPU 64 bit version with hash size set to 128 MB and pondering/permanent brain is disabled. Time control is 1 minute with no increment.

Naum 4 is renamed to Naum 4 E01 and Naum 4 E02 to accommodate the test. The engines use the same book renamed to 1000E01.ctg (Naum 4 E01) and 1000E02.ctg (Naum 4 E02). The book has only one predefined position with 20 half moves or 10 plies. The final position after exiting the book is:

r2qkb1r/1p1n1pp1/p2p1n2/3Pp2p/8/1N2BP2/PPP3PP/R2QKB1R w KQkq - 0 11

Result:

Games : 1000 (finished)

White Wins : 264 (26.4 %)
Black Wins : 256 (25.6 %)
Draws : 480 (48.0 %)

Final Outcome: No game is repeated.

Was learning on?

[kingliveson]

Could anyone answer, please, what is the probability to have repeated games in a sample of 1000 games, with a time control 1 minute/game?
How many games could be repeated in one thousand of games played?

2000 games have now been tested. The first set of 1000 games is between 2 different engines, Naum 4 versus Zappa Mexico II. The second set of 1000 games is between an engine (Naum 4) versus itself. Combining the 2000 games result in no repeated game. Also, There was only a single predefined position tested which speaks volume. Based on the data that I have presented, we can make a reasonable assumption.

For this time control, we know that each game is unique and occurs (1/1000). So the probability of having another identical game is less than [(1/1000)x(1/1000)] = 1E-6. Now if you factor in that the same position is tested 2000 times and not a single repeated game, you would have a probability of less than 2.5E-7.

Franklin

[Spacious_Mind]

This is not the question you asked, but I have a 1980s Tandy/Radio Shack chess computer with a small opening book. Once out of the opening, at a given time control, it always plays the same move in the same position - and I tended to do that as well, so to my annoyance I used to find myself playing the same games against it over and over.

Dedicated chess computers do not have multitasking and therefore, there is no difference other then the precision of the internal clock. Since it's a real time clock and interrupts are minimal, there should be always exactly the same time per move, which gives always the same moves.

Hi Milos

As Steve already pointed out. I am assuming you are just making an assumption because your statement is not really correct unless you talk about some of the very first few dedicated chess computers or very very cheap/weak ones. The Tandy would fall into one of these categories.

There is a case for example with GK2100 clones where the manufacturer repeated the same software into several different housings over the years, but clock speed and everything else remained the same. Even if you let these machines play back the exact same game with exact same settings it is still rare to repeat the exact same game from start to finish. Even with Random switched off. I have tested these several times and I have only managed to repeat the exact same game one time only. More often then not during the sequence of moves a different move does occur at least one time during a game. And here I am talking about two exactly the same dedicated chess computers with me forcing them to play back the exact same moves of a game!

http://www.spacious-mind.com/html/gk_21 ... _test.html

ps... the one time I did manage to repeat the game it was with the Saitek Cougar and Saitek Centurion and funnily enough these two actually have different clock speeds.

regards

Nick

[kingliveson]

Was learning on?

No.