asmFish testresults
Posted: Sun Jul 09, 2017 2:25 pm
asmfish Jun28 - asmfish Dec.07 = +36.6Elo(4000 games,CfA, 0.5s per move)
asmfish Jun28 - asmfish Dec07 = +43.6Elo(4000 games CfA, 0.5s per move,no book)
asmfish Jun28 - asmfish Dec07 = +45.3Elo(1000 games CfA, 1s per move,no book)
asmfish Jun28 - asmfish May22 = +00.3Elo(6000,0.5s per move)
asmfish Jun28 - asmfish May22 = +00.1Elo(2000,1s per move)
asmfish Jun28 - asmfish May22 = +07.8Elo(4000,0.5s per move.no book)
asmfish Jun28 - asmfish May22 = +04.7Elo(2000,1s per move.no book)
asmfish May22 - asmfish Dec07 = +37.3Elo(11000)
total : Jun28 +3.2Elo vs. May22 (23000games , but probably depends on book)
tablets, 1700kn/s in the startposition , Chess for Android, 0.5s per move, built-in-book with 50 openings,
no ponder, no TB, 4 threads, Intel Atom Baytrail Z3735 CPU
[updates here, as long as I can edit this post]
too lazy for an update today
-------------------------------
strangely, I seem to be getting better results for current versions, when I play without book.
Also with some (improved !?!) selfmade version that picks as bestmove the best from all threads
regardless of depth.
I didn't observe such book-things before. Need more games ...
------------------
may22_version2 - may22 = +1.5Elo(22000) CfA-book
may22_version2 - may22 = +4.6Elo(17000) no book
version2 : replace consecutive decimal bytes 124,218
(jl .next_worker3 , in Think.asm) by bytes 144,144 (nop,nop)
in the asmfish executable
--------------------------------------------------------------------------------------------------------------
the 50 openings from Chess for Andoid in >77000 games between
asmfishes and pedantfishes, mostly at 0.5s per move
sorted by drawrate
A: White-performance in 0/00
B: drawrate in 0/00
C: opening
===================================================================
10 recent asmFishes, 100 games round robin, 4500 games in total
0.5s per move CfA, no book , Z3735 tablet, 4 threads
Linux popcnt versions
a-:asmFish, old naming, own compile
p-:pedantFish, old naming, own compile
l-:asmFish, new naming, lantonov executable
asmfish Jun28 - asmfish Dec07 = +43.6Elo(4000 games CfA, 0.5s per move,no book)
asmfish Jun28 - asmfish Dec07 = +45.3Elo(1000 games CfA, 1s per move,no book)
asmfish Jun28 - asmfish May22 = +00.3Elo(6000,0.5s per move)
asmfish Jun28 - asmfish May22 = +00.1Elo(2000,1s per move)
asmfish Jun28 - asmfish May22 = +07.8Elo(4000,0.5s per move.no book)
asmfish Jun28 - asmfish May22 = +04.7Elo(2000,1s per move.no book)
asmfish May22 - asmfish Dec07 = +37.3Elo(11000)
total : Jun28 +3.2Elo vs. May22 (23000games , but probably depends on book)
tablets, 1700kn/s in the startposition , Chess for Android, 0.5s per move, built-in-book with 50 openings,
no ponder, no TB, 4 threads, Intel Atom Baytrail Z3735 CPU
[updates here, as long as I can edit this post]
too lazy for an update today
-------------------------------
strangely, I seem to be getting better results for current versions, when I play without book.
Also with some (improved !?!) selfmade version that picks as bestmove the best from all threads
regardless of depth.
I didn't observe such book-things before. Need more games ...
------------------
may22_version2 - may22 = +1.5Elo(22000) CfA-book
may22_version2 - may22 = +4.6Elo(17000) no book
version2 : replace consecutive decimal bytes 124,218
(jl .next_worker3 , in Think.asm) by bytes 144,144 (nop,nop)
in the asmfish executable
--------------------------------------------------------------------------------------------------------------
the 50 openings from Chess for Andoid in >77000 games between
asmfishes and pedantfishes, mostly at 0.5s per move
sorted by drawrate
A: White-performance in 0/00
B: drawrate in 0/00
C: opening
Code: Select all
A B C
------------------------------------------------------
353 382,1.e4 e5,2.f4 exf4,3.Nf3 g5,4.Bc4 g4
721 432,1.d4 c5,2.d5 e5,3.e4 d6
691 501,1.d4 d5,2.c4 e5,3.dxe5 d4,4.Nf3 Nc6
724 509,1.e4 d5,2.exd5 Qxd5,3.Nc3 Qa5,4.d4 e5
650 548,1.e4 d6,2.d4 Nf6,3.Nc3 g6,4.Bg5 h6
618 552,1.d4 Nf6,2.c4 g6,3.Nc3 Bg7,4.e4 d6
585 560,1.d4 Nf6,2.c4 e6,3.Nf3 c5,4.d5 b5
640 600,1.d4 e6,2.c4 Bb4+,3.Nc3 c6,4.Qb3 Qe7
651 606,1.d4 d5,2.c4 Nf6,3.cxd5 Nxd5,4.Nf3 Bf5
406 609,1.d4 d5,2.Nc3 Nf6,3.f3 Bf5,4.Bg5 Nbd7
553 638,1.d4 Nf6,2.c4 g6,3.f3 d5,4.cxd5 Nxd5
563 652,1.d4 f5,2.Nf3 Nf6,3.g3 g6
437 662,1.e4 e5,2.f4 exf4,3.Nf3 Be7,4.Bc4 Nf6
595 663,1.e4 b6,2.d4 Bb7
578 668,1.Nf3 Nf6,2.b3 g6,3.Bb2 Bg7,4.e3 d6
507 669,1.e4 e5,2.Nf3 Nc6,3.Bc4 Nf6,4.Ng5 d5
554 674,1.e4 c5,2.Nf3 Nc6,3.Bb5 g6,4.O-O Bg7
428 674,1.e4 e5,2.d4 exd4,3.Nf3 Bc5,4.Nxd4 Nf6
582 688,1.d4 d5,2.c4 e6,3.Nc3 c5,4.cxd5 exd5
526 689,1.e4 e5,2.Nf3 Nc6,3.Bb5 a6,4.Bxc6 dxc6
586 690,1.d4 d5,2.c4 e6,3.Nc3 Nf6,4.Bg5 Nbd7
561 695,1.d4 Nf6,2.c4 g6,3.Nc3 d5,4.cxd5 Nxd5
439 695,1.d4 Nf6,2.f4 g6,3.Nf3 Bg7,4.e3 O-O
540 701,1.e4 e6,2.d4 d5,3.Nc3 Bb4,4.e5 c5
562 704,1.e4 c5,2.Nf3 Nc6,3.d4 cxd4,4.Nxd4 g6
590 708,1.e4 e5,2.Nf3 d6,3.d4 Nf6,4.Nc3 Nbd7
526 721,1.d4 d5,2.c4 e6,3.Nf3 Nf6,4.Bg5 Bb4+
567 722,1.e4 c5,2.Nc3 Nc6
464 731,1.c4 e5,2.Nc3 Nf6,3.Nf3 Nc6,4.d4 exd4
540 740,1.e4 e5,2.Nf3 Nc6,3.Bc4 Bc5,4.O-O Nf6
535 750,1.d4 Nf6,2.c4 e6,3.Nf3 Bb4+,4.Nc3 c5
537 752,1.d4 d5,2.c4 c6,3.Nc3 dxc4,4.e3 b5
533 752,1.e4 e5,2.Nc3 Nf6,3.Nf3 Nc6,4.Bb5 Bb4
576 753,1.e4 c6,2.c4 d5,3.exd5 cxd5,4.cxd5 Nf6
567 753,1.e4 c6,2.d4 d5,3.exd5 cxd5,4.c4 Nf6
587 757,1.d4 d5,2.c4 dxc4,3.Nf3 Nf6,4.e3 e6
482 759,1.d4 d5,2.Nf3 Nf6,3.e3 Bf5,4.Bd3 e6
532 759,1.e4 e6,2.d4 d5,3.exd5 exd5,4.Nf3 Bd6
544 760,1.d4 d5,2.c4 c6,3.Nf3 Nf6,4.cxd5 cxd5
518 761,1.d4 Nf6,2.Nf3 g6,3.Bf4 Bg7,4.h3 d6
522 767,1.d4 d5,2.e4 e6
525 768,1.d4 Nf6,2.c4 e6,3.Nf3 Bb4+,4.Bd2 Bxd2+
480 788,1.e4 e5,2.Nf3 Nc6,3.c3 Nf6
531 794,1.e4 e5,2.Nf3 Nf6,3.d4 Nxe4,4.Bd3 d5
519 798,1.c4 c5,2.Nf3 Nf6,3.d4 cxd4,4.Nxd4 e6
531 806,1.e4 Nf6,2.e5 Nd5,3.c4 Nb6,4.c5 Nd5
532 807,1.e4 c5,2.Ne2 Nf6,3.Nbc3 d5,4.exd5 Nxd5
532 807,1.e4 e6,2.d4 d5,3.Nd2 c5,4.exd5 exd5
552 811,1.d4 Nf6,2.c4 e6,3.g3 Bb4+,4.Nd2 c5
512 879,1.e4 e5,2.Nf3 Nc6,3.Nc3 Nf6,4.d4 exd4
===================================================================
10 recent asmFishes, 100 games round robin, 4500 games in total
0.5s per move CfA, no book , Z3735 tablet, 4 threads
Linux popcnt versions
a-:asmFish, old naming, own compile
p-:pedantFish, old naming, own compile
l-:asmFish, new naming, lantonov executable
Code: Select all
name % bytes , description
--------------------------------------------------------------------------------
t-Jun28,52.17,113986,best thread, regardless of depth, displays best moves for each thread with eval.
i-Jun28,51.28,113634,best thread, regardless of depth, only 2 bytes different from a-Jun28
l-Jun28,50.89,113634, lantonov executable
a-Jun28,50.28,113634,same as l-Jun28
p-Jul04,49.72,113634,
a-Jul04,49.56,113986,
l-Jul04,49.28,113634,lantonov executable
t-Jul04,49.06,113986,best thread, regardless of depth, displays best moves for each thread with eval.
p-Jun28,48.94,113634,
c-may22,48.83,113626, l-may22 with contempt=10, 1 byte difference