Author: Dieter Buerssner
Date: 12:31:49 10/03/03
Go up one level in this thread
On October 02, 2003 at 15:49:13, Oliver Roese wrote:
>Ignoring draws, a match over 100 games between 2 programs of equal strength,
>should end with a score of the first programm being in the interval [44:56] in
>about 80.66% of all cases (computed with R, btw).
What do you mean by "computed with R"?
I can confirm your number.
Without ignoring draws, I get for example:
Result of chess matches between equal opponents
White wins 40.0%, black 30.0% and 30.0% draws
A match of 100 games was simulated by 1000000 Monte Carlo tries
result probability <= this > this
50.0 - 50.0 ( 50.0%): 4.793% 4.793% 95.207%
50.5 - 49.5 ( 50.5%): 9.480% 14.273% 85.727%
51.0 - 49.0 ( 51.0%): 9.275% 23.548% 76.452%
51.5 - 48.5 ( 51.5%): 9.008% 32.556% 67.444%
52.0 - 48.0 ( 52.0%): 8.546% 41.102% 58.898%
52.5 - 47.5 ( 52.5%): 8.027% 49.128% 50.872%
53.0 - 47.0 ( 53.0%): 7.440% 56.568% 43.432%
53.5 - 46.5 ( 53.5%): 6.766% 63.334% 36.666%
54.0 - 46.0 ( 54.0%): 6.021% 69.356% 30.644%
54.5 - 45.5 ( 54.5%): 5.375% 74.731% 25.269%
55.0 - 45.0 ( 55.0%): 4.667% 79.398% 20.602%
55.5 - 44.5 ( 55.5%): 3.954% 83.352% 16.648%
56.0 - 44.0 ( 56.0%): 3.406% 86.758% 13.242%
^^^^^^^
56.5 - 43.5 ( 56.5%): 2.848% 89.606% 10.394%
57.0 - 43.0 ( 57.0%): 2.317% 91.923% 8.077%
57.5 - 42.5 ( 57.5%): 1.884% 93.807% 6.194%
58.0 - 42.0 ( 58.0%): 1.517% 95.323% 4.677%
58.5 - 41.5 ( 58.5%): 1.188% 96.511% 3.489%
59.0 - 41.0 ( 59.0%): 0.907% 97.418% 2.582%
59.5 - 40.5 ( 59.5%): 0.712% 98.130% 1.869%
60.0 - 40.0 ( 60.0%): 0.522% 98.653% 1.347%
60.5 - 39.5 ( 60.5%): 0.402% 99.055% 0.945%
61.0 - 39.0 ( 61.0%): 0.295% 99.350% 0.650%
61.5 - 38.5 ( 61.5%): 0.209% 99.559% 0.441%
62.0 - 38.0 ( 62.0%): 0.134% 99.693% 0.307%
62.5 - 37.5 ( 62.5%): 0.099% 99.792% 0.208%
63.0 - 37.0 ( 63.0%): 0.074% 99.865% 0.135%
63.5 - 36.5 ( 63.5%): 0.046% 99.911% 0.089%
64.0 - 36.0 ( 64.0%): 0.032% 99.943% 0.057%
[...]
So almost a significant difference.
Regards,
Dieter
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.