Interesting and first(?) 7 piece TB test running

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

Jouni
Posts: 3291
Joined: Wed Mar 08, 2006 8:15 pm

Interesting and first(?) 7 piece TB test running

Post by Jouni »

It has currently 7 piece vs. 6 piece +14,07 and 7 piece vs. no +22,04 after 150k games. 100 kn/move test. Maybe better test will be fixed time like 0,1s!
Jouni
Jouni
Posts: 3291
Joined: Wed Mar 08, 2006 8:15 pm

Re: Interesting and first(?) 7 piece TB test running

Post by Jouni »

Same test with 2 and 8 move book with Cfish gave 5 piece - no TB +9,59. Too much. These figures are some kind of absolute maximum not happening in real games.
Jouni
User avatar
Eelco de Groot
Posts: 4567
Joined: Sun Mar 12, 2006 2:40 am
Full name:   

Re: Interesting and first(?) 7 piece TB test running

Post by Eelco de Groot »

I think that tablebase probes are essentially free, just one node in your test. If I see that correctly? It is still nice from a science viewpoint, to have an upper limit to a problem, and to see that the difference between 6 and 7 is larger than between 5 and 6 so no diminishing returns to the upper limit with a growing number of pieces. You can put your results straight in the ICGA Journal if that is still around :) Very good Jouni.
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
Dann Corbit
Posts: 12541
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Interesting and first(?) 7 piece TB test running

Post by Dann Corbit »

Since the tb files are not distance to mate, there can be lots of probes in a search.
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
syzygy
Posts: 5566
Joined: Tue Feb 28, 2012 11:56 pm

Re: Interesting and first(?) 7 piece TB test running

Post by syzygy »

Dann Corbit wrote: Thu Sep 13, 2018 12:43 am Since the tb files are not distance to mate, there can be lots of probes in a search.
The TB files are not distance to mate, and there can be lots of probes in a search. But the first part does not imply the second part. With distance to mate there would be just as many probes (but they'd take longer because your storage system would have to be much bigger).
Jouni
Posts: 3291
Joined: Wed Mar 08, 2006 8:15 pm

Re: Interesting and first(?) 7 piece TB test running

Post by Jouni »

It was not my test, but naturally :) from Leela forum https://docs.google.com/spreadsheets/d/ ... =358848559.
Jouni