Author: Dann Corbit
Date: 23:46:15 02/01/05
Go up one level in this thread
On February 02, 2005 at 02:36:41, Uri Blass wrote: >On February 02, 2005 at 02:09:12, Tony Werten wrote: > >>On February 01, 2005 at 19:03:35, Uri Blass wrote: >> >>>On February 01, 2005 at 16:58:27, Uri Blass wrote: >>> >>>>I find in Hyatt sites some information like >>>> >>>>ftp://ftp.cis.uab.edu/pub/hyatt/TB/tbs/kbppkn.tbs >>>> >>>>Unfortunately I am interested in statistics about 3 and 4 piece tablebases. >>>> >>>>Where can I find similiar statistics about 3 and 4 piece tablebases? >>>>I plan to write a program to calculate them directly from the tablebases but I >>>>do not trust my calculation to be correct. >>> >>> >>>Here is some statistics about kpk(white with the pawn). >>>I hope no errors. >>> >>> >>>pawn to move >>> >>>49468 draws >>>80 mates in 1 >>>226 mates in 2 >>>502 mates in 3 >>>876 mates in 4 >>>2036 mates in 5 >>>3874 mates in 6 >>>7784 mates in 7 >>>15572 mates in 8 >>>23392 mates in 9 >>>26872 mates in 10 >>>27182 mates in 11 >>>23010 mates in 12 >>>15240 mates in 13 >>>8054 mates in 14 >>>2146 mates in 15 >>>2324 mates in 16 >>>2164 mates in 17 >>>1742 mates in 18 >>>1316 mates in 19 >>>1116 mates in 20 >>>1212 mates in 21 >>>1124 mates in 22 >>>686 mates in 23 >>>288 mates in 24 >>>128 mates in 25 >>>38 mates in 26 >>>14 mates in 27 >>>6 mates in 28 >>> >>> >>>black to move >>> >>>101156 draws >>>0 mates >>>18 mate in 1 >>>46 mate in 2 >>>128 mate in 3 >>>306 mate in 4 >>>664 mate in 5 >>>1654 mate in 6 >>>4328 mate in 7 >>>8692 mate in 8 >>>15712 mate in 9 >>>19026 mate in 10 >>>25028 mate in 11 >>>17378 mate in 12 >>>10180 mate in 13 >>>5520 mate in 14 >>>2826 mate in 15 >>>2952 mate in 16 >>>1916 mate in 17 >>>1422 mate in 18 >>>1194 mate in 19 >>>872 mate in 20 >>>1162 mate in 21 >>>860 mate in 22 >>>584 mate in 23 >>>218 mate in 24 >>>62 mate in 25 >>>28 mate in 26 >>>8 mate in 27 >>>4 mate in 28 >> >>wtm: Draws: 19184 >>wtm: Mate in 28: 3 >>wtm: Mate in 27: 7 >>wtm: Mate in 26: 19 >>wtm: Mate in 25: 64 >>wtm: Mate in 24: 144 >>wtm: Mate in 23: 343 >>wtm: Mate in 22: 562 >>wtm: Mate in 21: 606 >>wtm: Mate in 20: 558 >>wtm: Mate in 19: 658 >>wtm: Mate in 18: 871 >>wtm: Mate in 17: 1066 >>wtm: Mate in 16: 1154 >>wtm: Mate in 15: 1065 >>wtm: Mate in 14: 3829 >>wtm: Mate in 13: 6719 >>wtm: Mate in 12: 8178 >>wtm: Mate in 11: 8601 >>wtm: Mate in 10: 8395 >>wtm: Mate in 9: 7541 >>wtm: Mate in 8: 5647 >>wtm: Mate in 7: 3121 >>wtm: Mate in 6: 1636 >>wtm: Mate in 5: 915 >>wtm: Mate in 4: 422 >>wtm: Mate in 3: 219 >>wtm: Mate in 2: 97 >>wtm: Mate in 1: 40 >>btm: Lost in 1: 9 >>btm: Lost in 2: 23 >>btm: Lost in 3: 64 >>btm: Lost in 4: 153 >>btm: Lost in 5: 332 >>btm: Lost in 6: 812 >>btm: Lost in 7: 2089 >>btm: Lost in 8: 4226 >>btm: Lost in 9: 7180 >>btm: Lost in 10: 7857 >>btm: Lost in 11: 7215 >>btm: Lost in 12: 5843 >>btm: Lost in 13: 4185 >>btm: Lost in 14: 2501 >>btm: Lost in 15: 1026 >>btm: Lost in 16: 1194 >>btm: Lost in 17: 902 >>btm: Lost in 18: 711 >>btm: Lost in 19: 597 >>btm: Lost in 20: 436 >>btm: Lost in 21: 565 >>btm: Lost in 22: 430 >>btm: Lost in 23: 292 >>btm: Lost in 24: 109 >>btm: Lost in 25: 31 >>btm: Lost in 26: 14 >>btm: Lost in 27: 4 >>btm: Lost in 28: 2 >>btm: Draws: 35210 >> >>Tony > >Thanks >I see that your number consider symmetric positions as equal. >I verified that my number for mate in 28 and loss in 28 were correct > >I see that I had a bug in my program and it also calculated cases when the pawn >is on the first or the last rank. > >After changing movei not to consider these cases(in games it will get only legal >positions so it does not need to consider errors except not having the >tablebases) that will never happen in the search and consider only pawn in a-d >file I can verify your numbers. Probably, it was just a tbs file listing. Here is an archive with all the tablebase stats files that I have in it: ftp://cap.connx.com/pub/chess-engines/new-approach/TBS.ZIP
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.