40x(2) v Houdini- A New Experiment Has Begun!

[geots]

Yesterday I began a match of different sorts. Engine 40x update v Houdini 2.0c x64. Changed the time control from 40/3 repeating to 10' + 10". And changed the number of games from 50 or 100 to 1000!

I will post updates every 50 games or so, or maybe daily- I shall see where my mood takes me.

The winner here will earn a place in the WCCCFAORI against Strelka 5.5. That is correct- I said Strelka 5.5


gts


PS: Of course I will be running another match or 2 involving other engines as well.

[gleperlier]

[ernest]

I will post updates every 50 games or so, or maybe daily-

Also post the result of each 50-game match: you will see how different the results will be... statistics in action!

But for the result of the 1000-game match, the error will be 4.5 times less (sqrt(1000/50)).
Looking forward to that!

[geots]

I will post updates every 50 games or so, or maybe daily-

Also post the result of each 50-game match: you will see how different the results will be... statistics in action!

But for the result of the 1000-game match, the error will be 4.5 times less (sqrt(1000/50)).
Looking forward to that!

Problem is no matter what you do- some things never change. In a 50-game match I ran Ivanhoe B49jA ag. Fritz 13. Not including the draws, Ivanhoe won 27-2 Ivanhoe would not win every 50 game match by that much, but do you think after 20 50-game matches, that Fritz 13 would have won one of the matches? Of course not.

Or maybe the match where throwing out draws, Houdini won against an opponent by a score of 70-0. How many matches would you have to run before the tide turned and the loser there started winning matches.

Never forget, in this order- you have:

1.Lies
2.Damn lies
3.Statistics