E8400 CHESS ENGINE PERMANENT LADDER

Discussion of computer chess matches and engine tournaments.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
mig2004
Posts: 75
Joined: Tue Oct 01, 2013 10:36 pm

Re: E8400 CHESS ENGINE PERMANENT LADDER

Post by mig2004 » Mon Feb 03, 2020 7:16 am

E8400 TOURNAMENT REPORT (ORDO)

WITH 1-9 GAMES PLAYED

"#" "PLAYER" "RATING" "PLAYED" "W" "D" "L" "D(%)"

"Stockfish 130120 64 T763" 3673.1 4 2 2 0 50
"190604-3c9c945 64mgw91" 3623.8 5 2 3 0 60
"SF 200109-ce88f06 64" 3566.4 4 1 3 0 75
"Stockfish 101219 64 T247" 3527.9 8 3 4 1 50
"SF 191227-9716852 64" 3510.6 6 3 1 2 16.7
"SFT 191130-26eRMek 64" 3486.9 6 0 6 0 100
"SF 191218-77939d6 64" 3486.9 2 0 2 0 100
"SFT 200105-80e282e 64" 3480.1 2 1 0 1 0
"Crystal-NN 011219 64" 3480.1 8 0 6 2 75
"SFT 190817-a12e7ac 64" 3477.7 9 0 8 1 88.9

WITH 10-19 GAMES PLAYED

"PLAYER" "RATING" "PLAYED" "W" "D" "L" "D(%)"
"SF 191210-b648247 64" 3501.1 11 2 8 1 72.7
"SFT 191229-e72453b 64" 3484.2 10 0 10 0 100
"Stockfish 261119 64 T085" 3480.1 12 1 10 1 83.3
"SFT 191123-f436cc2 64" 3476.7 18 2 14 2 77.8
"Crystal-Honey X5i" 3473.4 10 0 10 0 100
"SF 191120-3f41913 64" 3466 18 2 15 1 83.3 33
"SFT 191113-288121e 64" 3457.3 17 3 13 1 76.5 47
"SF 191108-fe35904 64" 3442.2 19 2 15 2 78.9 53
"SF 191031-e8fca71 64" 3438.8 10 0 10 0 100 55
"SF 191116-3468138 64" 3437 14 1 13 0 92.9

WITH 20-29 GAMES PLAYED

"PLAYER" "RATING" "PLAYED" "W" "D" "L" "D(%)
"SFT 191007-620dba1 64" 3493.7 29 3 26 0 89.7
"SF 191118-e0f42aa 64" 3492.3 24 2 21 1 87.5
"SF 191112-a00a336 64" 3471.7 20 2 16 2 80
"Stockfish 221019 64 T366" 3450.2 21 2 17 2 81
"SF 190830-a2162ce 64" 3442.3 28 0 28 0 100 115
"asmFishWCP_2019-10-16_base" 3383.7 25 2 17 6 68 131
"SF 190627-a8de07c 64 LPMP" 3370.2 21 1 20 0 95.2 143
"asmFishW_10_base_f631403" 3364.5 20 2 16 2 80 148
"Deep Synapse RZ8.0 64bit" 3359.7 24 2 20 1 87 259
"Cfish 140119 64 N ALEX76" 3276.9 27 0 26 1 96.3

WITH 30 OR MORE GAMES PLAYED

"PLAYER" "RATING" "PLAYED" "W" "D" "L" "D(%)"
"SF 190830-2e6dfc7 64" 3475.2 54 7 42 5 77.8
"SF 190826-263bbb0 64" 3474.2 62 11 43 8 69.4
"SF 191107-5e82c9d 64" 3458 39 5 31 3 79.5
"SF 191009-5522d3f 64" 3455.4 30 4 23 3 76.7
"SF 190725-d980d7c 64" 3454.1 67 7 55 5 82.1
"SugaR-NN 180919 64" 3451.9 37 3 32 2 86.5
"SFT 191111-a9eec00 64" 3448.6 34 6 21 7 61.8
"SFT 190819-1233390 64" 3446.7 39 6 29 4 74.4
"SF 190801-fcee0ce 64" 3442.3 31 1 29 1 93.5
"SF 190822-a016626 64" 3440.6 54 6 45 3 83.

Pgn and online report: https://www.file-up.org/cgt30wg8d239

mig2004
Posts: 75
Joined: Tue Oct 01, 2013 10:36 pm

Re: E8400 CHESS ENGINE PERMANENT LADDER

Post by mig2004 » Thu Feb 06, 2020 12:01 pm

Here are some of the test versions from stockfish test server, from December 12 up to date.

These are not the abrok versions. These engines were created in an effort to improve stockfish internal functions.

Some of them were never tested and entirely "disappeared" from public eye.. In fact, i tend to think not even the developers who created them got to test them. Probably, they were just superseded by newer versions that promised "better improvements", and we all know how rocky that path has been.

I have tested only a few, so you might have some interesting stuff to experiment with.

the download is about 78 MB.

https://www.file-up.org/tgqdnyro8ewb

mig2004
Posts: 75
Joined: Tue Oct 01, 2013 10:36 pm

Re: E8400 CHESS ENGINE PERMANENT LADDER

Post by mig2004 » Mon Feb 17, 2020 2:04 pm

E8400 CHESS ENGINE TOURNAMENT HIGHLIGHTS (ORDO REPORT)

WITH 1-9 GAMES PLAYED

# PLAYER RATING PLAYED W D L D(%) CFS(%)


1 SFm2k4 200206-794b4a0 64 3551.9 4 1 3 0 75 55
2 SF 200109-ce88f06 64 3539.7 8 2 6 0 75 52
3 190604-3c9c945 64mgw91 3533.8 9 2 7 0 78 76
12 SFT 191217-df32091 64 3467.1 2 0 2 0 100 1
13 SFT 191231-3981b7d 64 3467.1 4 2 0 2 0 50
14 SFT 200113-69b53f9 64 3467.0 4 1 3 0 75 56
18 SFT 200105-fa70746 64 3460.3 2 0 2 0 100 52
22 Stockfish 050220 64 T149 3455.8 4 0 4 0 100 50
23 SFT 191213-B-6fbca25 64 3455.8 4 0 4 0 100 51
24 SF 191218-77939d6 64 3455.0 8 0 8 0 100 52


WITH 10-19 GAMES PLAYED

# PLAYER RATING PLAYED W D L D(%)

4 Stockfish 101219 64 T247 3497.7 12 3 8 1 67
7 SF 191227-9716852 64 3479.2 10 3 5 2 50
8 SF 191210-b648247 64 3475.9 19 3 14 2 74
9 SFT 191229-e72453b 64 3471.6 18 1 16 1 89
17 Crystal-Honey X5i 3460.3 14 0 14 0 100
19 SFT 191113-288121e 64 3459.9 19 3 15 1 79
26 SFT 191123-f436cc2 64 3453.3 18 2 14 2 78
28 SFT 191130-26eRMek 64 3453.1 10 0 10 0 100
43 SFT 190817-a12e7ac 64 3446.4 15 0 14 1 93
52 SF-TCEC-190924-0436f01 64 3438.5 15 1 14 0 93

WITH 20-29 GAMES PLAYED

# PLAYER RATING PLAYED W D L D(%)

6 SF 191118-e0f42aa 64 3482.7 28 3 23 2 82
16 SF 191112-a00a336 64 3462.1 26 2 21 3 81
21 SF 191120-3f41913 64 3455.8 22 2 19 1 86
40 Stockfish 221019 64 T366 3447.8 25 2 21 2 84
46 SF 191108-fe35904 64 3442.3 23 2 19 2 83
48 SF 190830-a2162ce 64 3440.0 28 0 28 0 100
67 Stockfish 261119 64 T085 3430.5 20 1 18 1 90
129 asmFishWCP_2019-10-16_base 3377.3 25 2 17 6 68
138 SF 190627-a8de07c 64 LPMP 3368.7 21 1 20 0 95
148 asmFishW_10_base_f631403 3363.0 24 2 20 2 83

WITH 30 OR MORE GAMES PLAYED

# PLAYER RATING PLAYED W D L D(%)

5 SFT 191007-620dba1 64 3485.1 33 4 28 1 85
10 SF 191009-5522d3f 64 3468.0 34 4 27 3 79
11 SF 190826-263bbb0 64 3467.1 62 11 43 8 69
15 SF 190830-2e6dfc7 64 3463.6 58 7 46 5 79
20 SF 190801-fcee0ce 64 3457.4 35 2 32 1 91
25 SugaR-NN 180919 64 3453.9 41 3 36 2 88
29 SF 191107-5e82c9d 64 3452.7 39 5 31 3 79
37 SF 190725-d980d7c 64 3448.7 71 7 58 6 82
41 SFT 190819-1233390 64 3447.5 43 6 33 4 77
42 SFT 191111-a9eec00 64 3446.9 42 6 29 7 69


online report, (INCLUDES PGN FILE): https://www.file-up.org/vvmqy0gtelke
test engine pack from 20-02-06 to 20-02-11: https://www.file-up.org/hdb30z1eebgj (30.8MB DOWNLOAD)

mig2004
Posts: 75
Joined: Tue Oct 01, 2013 10:36 pm

Re: E8400 CHESS ENGINE PERMANENT LADDER

Post by mig2004 » Wed Feb 19, 2020 3:14 pm

Combo of

http://tests.stockfishchess.org/tests/v ... 8499f63a10
http://tests.stockfishchess.org/tests/v ... 8499f639c6
bench: 4619978

Combo, mstembera authored and mstembera committed on 20-02-12.
1 parent 4e89864 commit 899571f(..)


20-02-12 Fis Combo diff

LLR: 2.94 (-2.94,2.94) {-0.50,1.50}
Total: 57922 W: 11196 L: 10919 D: 35807
Ptnml(0-2): 960, 6681, 13419, 6893, 990

sprt @ 10+0.1 th 1
Combo of http://tests.stockfishchess.org/tests/v ... 8499f63a10 http://tests.stockfishchess.org/tests/v ... 8499f639c6


Commit Combo (Combo of)
Info Combo of http://tests.stockfishchess.org/tests/v ... 8499f63a10 http://tests.stockfishchess.org/tests/v ... 8499f639c6
Submitter Fisherman
TC 10+0.1
SPRT elo0: -0.50 alpha: 0.05 elo1: 1.50 beta: 0.05 (logistic)
LLR 2.94 [-2.94,2.94] (accepted)
Elo 1.31 [-0.59,3.11] (95%)
LOS 91.5%
Games 57922 [w:19.3%, l:18.9%, d:61.8%]
Link /tests/view/5e4467fde70d848499f63d44

Details
id 5e4467fde70d848499f63d44
new_tag Combo (Combo of)
new_signature 4619978
new_options Hash=16
resolved_new 899571fba2d4709f2deccee7777ca2684320000a
base_tag master (Modify singular beta for ttPv positions.)
base_signature 4932981
base_options Hash=16
resolved_base 4e8986483a83296e37433da7b07b64de53613f6f
sprt elo0: -0.50 alpha: 0.05 elo1: 1.50 beta: 0.05 state: accepted (logistic)
num_games 250000
tc 10+0.1
threads 1
book noob_3moves.epd
book_depth 8
( , 8 )
( 8 )
8

https://github.com/mstembera/Stockfish/ ... 684320000a

MD5 HASHES FOR SOURCE FILES:

7D67CF16F11C6F29C4779FE40AE86015 benchmark.cpp
73DB3F8776A21155C03CEC6A722B1258 bitbase.cpp
681A742F40E2ED563C1D576F2740F9AB bitboard.cpp
18FCEFE5D7FEA188D927B9199B605B72 bitboard.h
6BECA9DFBEAF83CED3F453BDD748195D endgame.cpp
CD08F0F040ABEA3CA4E5216047BAF3A1 endgame.h
5AE615576F195BE59577F82C1F9D69D2 evaluate.cpp
B224B3AF7B4E4DDA2FA86DCCF7C114E8 evaluate.h
46E0D929B015AB0CD3E99F41253F6AA6 main.cpp
69B8CE1783491AA5B087506B07ABBA99 Makefile
DABC202C23A8BA875018EA1769912C22 material.cpp
D0252E530091B71DC540A9A8EB3ABD0C material.h
B0AF79E1550BA442300FD8E7991D66E9 misc.cpp
DF04A7C999EF95267CBE1D2FC9EDCCA2 misc.h
E5875D882309E5EC4F71A127A26688FE movegen.cpp
A32B107E1578E6A4D09B656DA5FA2B80 movegen.h
50090A20FB4F7D4AB0C51C4AFBFAFC7F movepick.cpp
9464E283EF7E3A3C4A96A2F78CDFCBC7 movepick.h
EBC53C6328A95F8DBBBE4202AAE7444E pawns.cpp
078056898F526BECFF74CEF0D4C3F237 pawns.h
F149636932154D8DE3E019BC61C1A71F position.cpp
B1D59F4FB3BD81F856883DD55C227548 position.h
043A2E7B9E40BCB8E627DB3765CEE28A psqt.cpp
4B59523B9E72CF8C11959E9CAAB79304 search.cpp
7AB329BE89832A2E2FB9C5C1457EB6EE search.h
D58FDF0CE591690FA8E0407A9E398195 thread.cpp
B05A26EC3A5BB0FED54048F1E844F8D4 thread.h
20B8E29AEA0FB1E24502852210411187 thread_win32_osx.h
FFC62C1AB2D0FCEFCFDE07EFAFA39025 timeman.cpp
FF0EFB49F2C67352CD7153335EEB6A9B timeman.h
166855546FBD0CBF34E6B66BC252D2D4 tt.cpp
C32AB4C85FCF6AE9ABAE27C5785367F4 tt.h
A1AF1892266495C925D799742430243C types.h
E3283F1031A878A585D8C094C0612520 uci.cpp
017176566B76B7C7D350C21D80F82EB0 uci.h
2F5458DC35A9AF447F4D6136B5ACA98B ucioption.cpp

6972AE85F563BDAB14DC25C95284EF41

download link: https://www.file-up.org/2phanaiv8qte

mig2004
Posts: 75
Joined: Tue Oct 01, 2013 10:36 pm

Re: E8400 CHESS ENGINE PERMANENT LADDER

Post by mig2004 » Fri Feb 21, 2020 10:48 am

Simplify Futility Move Count, AV 1.0.3444, Submitted on 20-02-15 by protonspring (committed by Joost VandeVondele. Master 10ead8a7.


Simplify Futility Move Count

remove two constants

STC
LLR: 2.95 (-2.94,2.94) {-1.50,0.50}
Total: 62050 W: 11903 L: 11802 D: 38345
Ptnml(0-2): 1002, 7346, 14283, 7320, 1065
http://tests.stockfishchess.org/tests/v ... 8499f63c6d

LTC
LLR: 2.93 (-2.94,2.94) {-1.50,0.50}
Total: 12850 W: 1679 L: 1572 D: 9599
Ptnml(0-2): 82, 1171, 3818, 1249, 96
http://tests.stockfishchess.org/tests/v ... 8499f63cc0

Bench: 4762351

master

@protonspring
@vondele
protonspring authored and vondele committed 5 days ago
1 parent be5a2f0 commit 10ead8a724671132df60ca2e23ddcad7eff7b3e5


Details
id 5e41d73be70d848499f63c6d
new_tag ps_fmc13 (Merge branch 'master' into ps_fmc13)
new_signature 4762351
new_options Hash=16
resolved_new e3143eefd1ec7a51410524a05efe3ad571f326f8
base_tag master (Modify singular beta for ttPv positions.)
base_signature 4932981
base_options Hash=16
resolved_base 4e8986483a83296e37433da7b07b64de53613f6f
sprt elo0: -1.50 alpha: 0.05 elo1: 0.50 beta: 0.05 state: accepted (logistic)
num_games 250000
tc 10+0.1
threads 1
book noob_3moves.epd
book_depth 8
( , 8 )
( 8 )
8

Commit ps_fmc13 (Merge branch 'master' into ps_fmc13)
Info simplify futility move count.
Submitter protonspring
TC 10+0.1
SPRT elo0: -1.50 alpha: 0.05 elo1: 0.50 beta: 0.05 (logistic)
LLR 2.95 [-2.94,2.94] (accepted)
Elo 0.23 [-1.61,1.98] (95%)
LOS 60.2%
Games 62050 [w:19.2%, l:19.0%, d:61.8%]
Link /tests/view/5e41d73be70d848499f63c6d




7D67CF16F11C6F29C4779FE40AE86015 benchmark.cpp
73DB3F8776A21155C03CEC6A722B1258 bitbase.cpp
681A742F40E2ED563C1D576F2740F9AB bitboard.cpp
18FCEFE5D7FEA188D927B9199B605B72 bitboard.h
8C476DF4FDCDE1BAD2B236350E970306 endgame.cpp
CD08F0F040ABEA3CA4E5216047BAF3A1 endgame.h
A6339B3CEE1BA8DDAB4F7BEA31006B9C evaluate.cpp
B224B3AF7B4E4DDA2FA86DCCF7C114E8 evaluate.h
46E0D929B015AB0CD3E99F41253F6AA6 main.cpp
69B8CE1783491AA5B087506B07ABBA99 Makefile
DABC202C23A8BA875018EA1769912C22 material.cpp
D0252E530091B71DC540A9A8EB3ABD0C material.h
B0AF79E1550BA442300FD8E7991D66E9 misc.cpp
DF04A7C999EF95267CBE1D2FC9EDCCA2 misc.h
E5875D882309E5EC4F71A127A26688FE movegen.cpp
A32B107E1578E6A4D09B656DA5FA2B80 movegen.h
50090A20FB4F7D4AB0C51C4AFBFAFC7F movepick.cpp
9464E283EF7E3A3C4A96A2F78CDFCBC7 movepick.h
EBC53C6328A95F8DBBBE4202AAE7444E pawns.cpp
078056898F526BECFF74CEF0D4C3F237 pawns.h
F149636932154D8DE3E019BC61C1A71F position.cpp
B1D59F4FB3BD81F856883DD55C227548 position.h
043A2E7B9E40BCB8E627DB3765CEE28A psqt.cpp
1F6288403C942C538E804CE5176DF97F search.cpp
7AB329BE89832A2E2FB9C5C1457EB6EE search.h
D58FDF0CE591690FA8E0407A9E398195 thread.cpp
B05A26EC3A5BB0FED54048F1E844F8D4 thread.h
20B8E29AEA0FB1E24502852210411187 thread_win32_osx.h
FFC62C1AB2D0FCEFCFDE07EFAFA39025 timeman.cpp
FF0EFB49F2C67352CD7153335EEB6A9B timeman.h
166855546FBD0CBF34E6B66BC252D2D4 tt.cpp
C32AB4C85FCF6AE9ABAE27C5785367F4 tt.h
A1833B5C8E4D285105A4B0357D591456 types.h
E3283F1031A878A585D8C094C0612520 uci.cpp
017176566B76B7C7D350C21D80F82EB0 uci.h
2F5458DC35A9AF447F4D6136B5ACA98B ucioption.cpp

B253BB8EF75125A0D5BDA2B1EDDE8D07

download link: https://www.mediafire.com/file/tcqd53ef ... 3.zip/file

mig2004
Posts: 75
Joined: Tue Oct 01, 2013 10:36 pm

Re: E8400 CHESS ENGINE PERMANENT LADDER

Post by mig2004 » Fri Feb 21, 2020 11:03 am

P.R. #2553 - Updated KNNKP endgame. AV 1.0.3445. submitted on 20-02-18 by protonspring. Master 0795fa43 ← ps_knnkp4_pr c91665e7.



protonspring commented 10 hours ago •

This is a patch that significantly improves KNNKP endgames. My testing suggests this is better than master by 3:1 (in these specific endings).

Theory: KNNK is a dead draw, however the presence of the additional weakSide pawn opens up some mate opportunities. The idea is to block the pawn (preferably behind the Troitsky line) with one of the knights and press the weakSide king into a corner. If we can stalemate the king, we release the pawn with the knight (to avoid actual stalemate), and use the knight to complete the mate before the pawn promotes. This is also why there is an additional penalty for advancement of the pawn.

Please help verify: You can download my KNNvKPFens.epd and use it as an opening book in cutechess. https://drive.google.com/drive/folders/ ... XMCKLc8gBy

Set book depth to 1, repeat on, random order, disable resigning and draws.

This patch does not regress against master in normal games:

STC
LLR: 2.94 (-2.94,2.94) {-1.50,0.50}
Total: 96616 W: 18459 L: 18424 D: 59733
Ptnml(0-2): 1409, 10812, 23802, 10905, 1380
http://tests.stockfishchess.org/tests/v ... 2b40cd31bc

LTC
LLR: 2.95 (-2.94,2.94) {-1.50,0.50}
Total: 49726 W: 6340 L: 6304 D: 37082
Ptnml(0-2): 239, 4227, 15906, 4241, 250
http://tests.stockfishchess.org/tests/v ... f8c4cc01d0

Bench 4981770
@protonspring
Updated KNNKP endgmae.
c91665e


ps_knnkp4 vs master diff

LLR: 2.94 (-2.94,2.94) {-1.50,0.50}
Total: 96616 W: 18459 L: 18424 D: 59733
Ptnml(0-2): 1409, 10812, 23802, 10905, 1380

Details
id 5e49dfe6f8d1d52b40cd31bc
new_tag ps_knnkp4 (v5)
new_signature 4981770
new_options Hash=16
resolved_new 2e439967a82c7e2483a891dc4359c21e38c4461e
base_tag master (Simplify Futility Move Count)
base_signature 4762351
base_options Hash=16
resolved_base 10ead8a724671132df60ca2e23ddcad7eff7b3e5
sprt elo0: -1.50 alpha: 0.05 elo1: 0.50 beta: 0.05 state: accepted (logistic)
num_games 250000
tc 10+0.1
threads 1
book noob_3moves.epd
book_depth 8
( , 8 )
( 8 )
8


Commit ps_knnkp4 (v5)
Info test updated KNNKP. Low tp
Submitter protonspring
TC 10+0.1
SPRT elo0: -1.50 alpha: 0.05 elo1: 0.50 beta: 0.05 (logistic)
LLR 2.94 [-2.94,2.94] (accepted)
Elo -0.18 [-1.72,1.21] (95%)
LOS 39.9%
Games 96616 [w:19.1%, l:19.1%, d:61.8%]
Link /tests/view/5e49dfe6f8d1d52b40cd31bc


7D67CF16F11C6F29C4779FE40AE86015 benchmark.cpp
73DB3F8776A21155C03CEC6A722B1258 bitbase.cpp
681A742F40E2ED563C1D576F2740F9AB bitboard.cpp
18FCEFE5D7FEA188D927B9199B605B72 bitboard.h
EB17FA7ADDE8534ACD5676AF62852383 endgame.cpp
CD08F0F040ABEA3CA4E5216047BAF3A1 endgame.h
A6339B3CEE1BA8DDAB4F7BEA31006B9C evaluate.cpp
B224B3AF7B4E4DDA2FA86DCCF7C114E8 evaluate.h
46E0D929B015AB0CD3E99F41253F6AA6 main.cpp
69B8CE1783491AA5B087506B07ABBA99 Makefile
DABC202C23A8BA875018EA1769912C22 material.cpp
D0252E530091B71DC540A9A8EB3ABD0C material.h
B0AF79E1550BA442300FD8E7991D66E9 misc.cpp
DF04A7C999EF95267CBE1D2FC9EDCCA2 misc.h
E5875D882309E5EC4F71A127A26688FE movegen.cpp
A32B107E1578E6A4D09B656DA5FA2B80 movegen.h
50090A20FB4F7D4AB0C51C4AFBFAFC7F movepick.cpp
9464E283EF7E3A3C4A96A2F78CDFCBC7 movepick.h
EBC53C6328A95F8DBBBE4202AAE7444E pawns.cpp
078056898F526BECFF74CEF0D4C3F237 pawns.h
F149636932154D8DE3E019BC61C1A71F position.cpp
B1D59F4FB3BD81F856883DD55C227548 position.h
043A2E7B9E40BCB8E627DB3765CEE28A psqt.cpp
1F6288403C942C538E804CE5176DF97F search.cpp
7AB329BE89832A2E2FB9C5C1457EB6EE search.h
D58FDF0CE591690FA8E0407A9E398195 thread.cpp
B05A26EC3A5BB0FED54048F1E844F8D4 thread.h
20B8E29AEA0FB1E24502852210411187 thread_win32_osx.h
FFC62C1AB2D0FCEFCFDE07EFAFA39025 timeman.cpp
FF0EFB49F2C67352CD7153335EEB6A9B timeman.h
166855546FBD0CBF34E6B66BC252D2D4 tt.cpp
C32AB4C85FCF6AE9ABAE27C5785367F4 tt.h
A1833B5C8E4D285105A4B0357D591456 types.h
E3283F1031A878A585D8C094C0612520 uci.cpp
017176566B76B7C7D350C21D80F82EB0 uci.h
2F5458DC35A9AF447F4D6136B5ACA98B ucioption.cpp

0D202EBFB82BC44E20FD73810A76619C

download link: http://www.mediafire.com/file/yrr0umrp1 ... 0.zip/file

mig2004
Posts: 75
Joined: Tue Oct 01, 2013 10:36 pm

Re: E8400 CHESS ENGINE PERMANENT LADDER

Post by mig2004 » Fri Feb 21, 2020 11:44 am

Fix for incorrect VALUE_MATE_IN_MAX_PLY usage.
1.0.3443
by Joost VandeVondele
master be5a2f01




Fix for incorrect VALUE_MATE_IN_MAX_PLY usage.

Fixes #2533, fixes #2543, fixes #2423.

the code that prevents false mate announcements depending on the TT
state (GHI), incorrectly used VALUE_MATE_IN_MAX_PLY. The latter
constant, however, also includes, counterintuitively, the TB win range.

This patch fixes that, by restoring the behavior for TB win scores,
while retaining the false mate correctness, and improving the mate
finding ability. In particular

no alse mates are announced with the poisened hash testcase
```
position fen 8/8/8/3k4/8/8/6K1/7R w - - 0 1
go depth 40
position fen 8/8/8/3k4/8/8/6K1/7R w - - 76 1
go depth 20
ucinewgame
```

mates are found with the testcases reported in #2543
```
position fen 4k3/3pp3/8/8/8/8/2PPP3/4K3 w - - 0 1
setoption name Hash value 1024
go depth 55
ucinewgame
```
and
```
position fen 4k3/4p3/8/8/8/8/3PP3/4K3 w - - 0 1
setoption name Hash value 1024
go depth 45
ucinewgame
```

furthermore, on the mate finding benchmark (ChestUCI_23102018.epd),
performance improves over master, roughly reaching performance with the
false mate protection reverted
```
Analyzing 6566 mate positions for best and found mates:

----------------best ---------------found
nodes master revert fixed master revert fixed
16000000 4233 4236 4235 5200 5201 5199
32000000 4583 4585 4585 5417 5424 5418
64000000 4852 4853 4855 5575 5584 5579
128000000 5071 5068 5066 5710 5720 5716
256000000 5280 5282 5279 5819 5827 5826
512000000 5471 5468 5468 5919 5935 5932
```

On a testcase with TB enabled, progress is made consistently, contrary
to master
```
setoption name SyzygyPath value ../../../syzygy/3-4-5/
setoption name Hash value 2048
position fen 1R6/3k4/8/K2p4/4n3/2P5/8/8 w - - 0 1
go depth 58
ucinewgame
```

The PR (prior to a rewrite for clarity)

passed STC:
LLR: 2.94 (-2.94,2.94) {-1.50,0.50}
Total: 65405 W: 12454 L: 12384 D: 40567
Ptnml(0-2): 920, 7256, 16285, 7286, 944
http://tests.stockfishchess.org/tests/v ... 8499f63d15

passed LTC:
LLR: 2.94 (-2.94,2.94) {-1.50,0.50}
Total: 27096 W: 3477 L: 3413 D: 20206
Ptnml(0-2): 128, 2215, 8776, 2292, 122
http://tests.stockfishchess.org/tests/v ... 8499f63d63

The incorrectly named VALUE_MATE_IN_MAX_PLY and VALUE_MATED_IN_MAX_PLY
were renamed into VALUE_TB_WIN_IN_MAX_PLY and VALUE_TB_LOSS_IN_MAX_PLY,
and correclty defined VALUE_MATE_IN_MAX_PLY and VALUE_MATED_IN_MAX_PLY
were introduced.

One further (corner case) mistake using these constants was fixed (go
mate X), which could lead to a premature return if X > MAX_PLY / 2,
but TB were present.

Thanks to @svivanov72 for one of the reports and help fixing the issue.

closes #2552, Bench: 4932981, master, vondele committed on 20-02-11. 1 parent 4e89864 commit be5a2f0



7D67CF16F11C6F29C4779FE40AE86015 benchmark.cpp
73DB3F8776A21155C03CEC6A722B1258 bitbase.cpp
681A742F40E2ED563C1D576F2740F9AB bitboard.cpp
18FCEFE5D7FEA188D927B9199B605B72 bitboard.h
8C476DF4FDCDE1BAD2B236350E970306 endgame.cpp
CD08F0F040ABEA3CA4E5216047BAF3A1 endgame.h
A6339B3CEE1BA8DDAB4F7BEA31006B9C evaluate.cpp
B224B3AF7B4E4DDA2FA86DCCF7C114E8 evaluate.h
46E0D929B015AB0CD3E99F41253F6AA6 main.cpp
69B8CE1783491AA5B087506B07ABBA99 Makefile
DABC202C23A8BA875018EA1769912C22 material.cpp
D0252E530091B71DC540A9A8EB3ABD0C material.h
B0AF79E1550BA442300FD8E7991D66E9 misc.cpp
DF04A7C999EF95267CBE1D2FC9EDCCA2 misc.h
E5875D882309E5EC4F71A127A26688FE movegen.cpp
A32B107E1578E6A4D09B656DA5FA2B80 movegen.h
50090A20FB4F7D4AB0C51C4AFBFAFC7F movepick.cpp
9464E283EF7E3A3C4A96A2F78CDFCBC7 movepick.h
EBC53C6328A95F8DBBBE4202AAE7444E pawns.cpp
078056898F526BECFF74CEF0D4C3F237 pawns.h
F149636932154D8DE3E019BC61C1A71F position.cpp
B1D59F4FB3BD81F856883DD55C227548 position.h
043A2E7B9E40BCB8E627DB3765CEE28A psqt.cpp
FFD59E8BAB8091E2884476FCE36F9CE1 search.cpp
7AB329BE89832A2E2FB9C5C1457EB6EE search.h
D58FDF0CE591690FA8E0407A9E398195 thread.cpp
B05A26EC3A5BB0FED54048F1E844F8D4 thread.h
20B8E29AEA0FB1E24502852210411187 thread_win32_osx.h
FFC62C1AB2D0FCEFCFDE07EFAFA39025 timeman.cpp
FF0EFB49F2C67352CD7153335EEB6A9B timeman.h
166855546FBD0CBF34E6B66BC252D2D4 tt.cpp
C32AB4C85FCF6AE9ABAE27C5785367F4 tt.h
A1833B5C8E4D285105A4B0357D591456 types.h
E3283F1031A878A585D8C094C0612520 uci.cpp
017176566B76B7C7D350C21D80F82EB0 uci.h
2F5458DC35A9AF447F4D6136B5ACA98B ucioption.cpp

CEB3566957D4D4A8518B83529C260BDB

download link: https://userscloud.com/mbomz0et5or1

mig2004
Posts: 75
Joined: Tue Oct 01, 2013 10:36 pm

Re: E8400 CHESS ENGINE PERMANENT LADDER

Post by mig2004 » Sat Feb 22, 2020 1:54 pm

Filename: SF 200221-6123-1893.zip
Size: 2.1 MB


Pull request #2557 - Improve move order near the root
Improve move order near the root
1.0.3449
by Günther Demetz
master 5d30e0dc ← lowPlyHistPR 61233cac
On 2020-02-21

https://github.com/official-stockfish/S ... /pull/2557

download link: https://www.file-up.org/nlowzicxf6mf

pb00068 commented :

Current move histories are known to work well near the leaves, whilst at higher depths they aren't very helpful. If for instance a good pv-move near the root get overtaken by a new best move during iterative deepening and then when played later on it never succeds, it will fall down in the ranking of move order despite it's good potential near the root. Due to LMR it will now take severals ID-cycles until this move will be examined seriously again.
To address this problem this patch introduces a table dedicated for what's happening at plies 0-3.
It's structured like mainHistory with an additional dimension which divides the table in 2 parts:
plies 0 & 1
plies 2 & 3
It get cleared with each new search and is filled during iterative deepening at higher depths when recording successful quiet moves near the root or trasversing nodes which are/were in the prinical variation (ttPv).

Medium TC (20+0.2):
https://tests.stockfishchess.org/tests/ ... 2810d096dc
LLR: 2.94 (-2.94,2.94) {-0.50,1.50}
Total: 100910 W: 16682 L: 16376 D: 67852
Ptnml(0-2): 1177, 10983, 25883, 11181, 1231

Commit lowPlyHistory (Take 4)
Info Take 4: restrict usage to first 4 plies on ttPv, revised bonus-logic. Using increased tc (half throughput) to reach needed depths
Submitter pb00067
TC 20+0.2
SPRT elo0: -0.50 alpha: 0.05 elo1: 1.50 beta: 0.05 (logistic)
LLR 2.94 [-2.94,2.94] (accepted)
Elo 0.75 [-0.73,2.09] (95%)
LOS 85.3%
Games 100910 [w:16.5%, l:16.2%, d:67.2%]
Link /tests/view/5e4d358790a0a02810d096dc

Details
id 5e4d358790a0a02810d096dc
new_tag lowPlyHistory (Take 4)
new_signature 4573156
new_options Hash=16
resolved_new 1f8205a5fe08b37436a08c22094df5a9a0c7f956
base_tag master (Simplify Futility Move Count)
base_signature 4762351
base_options Hash=16
resolved_base 10ead8a724671132df60ca2e23ddcad7eff7b3e5
sprt elo0: -0.50 alpha: 0.05 elo1: 1.50 beta: 0.05 state: accepted (logistic)
num_games 250000
tc 20+0.2
threads 1
book noob_3moves.epd
book_depth 8



LTC:
https://tests.stockfishchess.org/tests/ ... 2810d09714
LLR: 2.95 (-2.94,2.94) {0.25,1.75}
Total: 80444 W: 10495 L: 10095 D: 59854
Ptnml(0-2): 551, 7479, 23803, 7797, 592

To go ahead from here, we could try to:

extend it to more plies
include captures as well
apply penalties too instead to clearing the table w/ each new search
apply bonus for trasversed ttPv-nodes on other history tables too
change bonus weights/logic
simplify

...

P.S.the current version of the patch could be simplified by removing the color dimension and just using ply instead to ply/2 in the new dimension.
True
-----------------



src page: https://github.com/pb00068/Stockfish/tr ... 2cab0d53f5


MD5 Hash

7D67CF16F11C6F29C4779FE40AE86015 benchmark.cpp
73DB3F8776A21155C03CEC6A722B1258 bitbase.cpp
681A742F40E2ED563C1D576F2740F9AB bitboard.cpp
18FCEFE5D7FEA188D927B9199B605B72 bitboard.h
4362E8EEA9812D6181FD663851C248FD endgame.cpp
CD08F0F040ABEA3CA4E5216047BAF3A1 endgame.h
A6339B3CEE1BA8DDAB4F7BEA31006B9C evaluate.cpp
B224B3AF7B4E4DDA2FA86DCCF7C114E8 evaluate.h
46E0D929B015AB0CD3E99F41253F6AA6 main.cpp
69B8CE1783491AA5B087506B07ABBA99 Makefile
DABC202C23A8BA875018EA1769912C22 material.cpp
D0252E530091B71DC540A9A8EB3ABD0C material.h
B0AF79E1550BA442300FD8E7991D66E9 misc.cpp
DF04A7C999EF95267CBE1D2FC9EDCCA2 misc.h
E5875D882309E5EC4F71A127A26688FE movegen.cpp
A32B107E1578E6A4D09B656DA5FA2B80 movegen.h
7152F0E221D755FF3E1406CBDA691CCC movepick.cpp
92143D360726E061CE6E5D1C4734668C movepick.h
EBC53C6328A95F8DBBBE4202AAE7444E pawns.cpp
078056898F526BECFF74CEF0D4C3F237 pawns.h
F149636932154D8DE3E019BC61C1A71F position.cpp
B1D59F4FB3BD81F856883DD55C227548 position.h
043A2E7B9E40BCB8E627DB3765CEE28A psqt.cpp
8AE8CEBA309CE42DF052EE034D5182A8 search.cpp
7AB329BE89832A2E2FB9C5C1457EB6EE search.h
D9327278A5685121102AFDEAAE399190 thread.cpp
831613A444BBA36F2C617A00C78A007D thread.h
20B8E29AEA0FB1E24502852210411187 thread_win32_osx.h
FFC62C1AB2D0FCEFCFDE07EFAFA39025 timeman.cpp
FF0EFB49F2C67352CD7153335EEB6A9B timeman.h
166855546FBD0CBF34E6B66BC252D2D4 tt.cpp
C32AB4C85FCF6AE9ABAE27C5785367F4 tt.h
A1833B5C8E4D285105A4B0357D591456 types.h
E3283F1031A878A585D8C094C0612520 uci.cpp
017176566B76B7C7D350C21D80F82EB0 uci.h
2F5458DC35A9AF447F4D6136B5ACA98B ucioption.cpp

MD5 DIGEST

1893A706D93F630E515520EAF78DD709

mig2004
Posts: 75
Joined: Tue Oct 01, 2013 10:36 pm

Re: E8400 CHESS ENGINE PERMANENT LADDER

Post by mig2004 » Sun Feb 23, 2020 4:04 pm

https://github.com/xoto10/stockfish-xot ... 8a5748e7b0


compiled name: SFT 200209-e7e6501-7F4D5

Adjust draw_value() by net pawn count. Bench 4877773. xoto10 committed on 200203. 1 parent 0f37da0 commit

e7e6501.

download link: https://www.file-up.org/0dppbah8yftc


20-02-08 xot drawpawns1^ diff

LLR: 2.95 (-2.94,2.94) {-0.50,1.50}
Total: 13205 W: 2128 L: 1960 D: 9117
Ptnml(0-2): 127, 1325, 3538, 1447, 151

sprt @ 25+0.25 th 1
Prio -1 scaling test: Adjust draw_value() by net pawn count.


Commit drawpawns1^ (Adjust draw_value() by net pawn count.)
Info Prio -1 scaling test: Adjust draw_value() by net pawn count.
Submitter xoto
TC 25+0.25
SPRT elo0: -0.50 alpha: 0.05 elo1: 1.50 beta: 0.05 (logistic)
LLR 2.95 [-2.94,2.94] (accepted)
Elo 4.15 [0.84,7.43] (95%)
LOS 99.3%
Games 13205 [w:16.1%, l:14.8%, d:69.0%]
Link /tests/view/5e3e069fe70d848499f63b50

Details
id 5e3e069fe70d848499f63b50
new_tag drawpawns1^ (Adjust draw_value() by net pawn count.)
new_signature 4877773
new_options Hash=40
resolved_new e7e6501821a0095ee4ea9c91a658ab8a5748e7b0
base_tag 0f37da0 (Simplify away king infiltration.)
base_signature 5047825
base_options Hash=40
resolved_base 0f37da0e34e02efaaca907878d1a3e0d916f447c
sprt elo0: -0.50 alpha: 0.05 elo1: 1.50 beta: 0.05 state: accepted (logistic)
num_games 250000
tc 25+0.25
threads 1
book noob_3moves.epd
book_depth 8
( , 8 )
( 8 )
8




MD5 HASH

7D67CF16F11C6F29C4779FE40AE86015 benchmark.cpp
45C152D9BFA269A67D0DBC97FDA93260 bitbase.cpp
681A742F40E2ED563C1D576F2740F9AB bitboard.cpp
18FCEFE5D7FEA188D927B9199B605B72 bitboard.h
6BECA9DFBEAF83CED3F453BDD748195D endgame.cpp
D4F8F7081BCDDDCD66F8B9939FAC76BF endgame.h
393A2E2F7D5E417B987659CBBABE6A25 evaluate.cpp
B224B3AF7B4E4DDA2FA86DCCF7C114E8 evaluate.h
580245A523BA93DEAB8043BEB13FF0CD main.cpp
69B8CE1783491AA5B087506B07ABBA99 Makefile
DABC202C23A8BA875018EA1769912C22 material.cpp
D0252E530091B71DC540A9A8EB3ABD0C material.h
4B2369157C91294B26427404F516760B misc.cpp
166F24480A4CC376DAD6F904E2713296 misc.h
D9A4FA1C5D288E17005692F4D9D8DB3B movegen.cpp
A32B107E1578E6A4D09B656DA5FA2B80 movegen.h
50090A20FB4F7D4AB0C51C4AFBFAFC7F movepick.cpp
9464E283EF7E3A3C4A96A2F78CDFCBC7 movepick.h
EBC53C6328A95F8DBBBE4202AAE7444E pawns.cpp
078056898F526BECFF74CEF0D4C3F237 pawns.h
ABC9D43D72AD6B999F5CDE52328885F9 position.cpp
B1D59F4FB3BD81F856883DD55C227548 position.h
043A2E7B9E40BCB8E627DB3765CEE28A psqt.cpp
AA7AD7C4101A6F216D6C2CC706480052 search.cpp
7AB329BE89832A2E2FB9C5C1457EB6EE search.h
38CA159EF42BAF468FF2ED723490DBFD thread.cpp
448FDB3555C69DC705B11550C15DCB0F thread.h
20B8E29AEA0FB1E24502852210411187 thread_win32_osx.h
FFC62C1AB2D0FCEFCFDE07EFAFA39025 timeman.cpp
FF0EFB49F2C67352CD7153335EEB6A9B timeman.h
5E06BB7068501D4730D119325B61F6E1 tt.cpp
C32AB4C85FCF6AE9ABAE27C5785367F4 tt.h
A1AF1892266495C925D799742430243C types.h
E3283F1031A878A585D8C094C0612520 uci.cpp
017176566B76B7C7D350C21D80F82EB0 uci.h
2F5458DC35A9AF447F4D6136B5ACA98B ucioption.cpp


MD5 DIGEST: 7F4D544C44F8F172F3E903D3E297CAD6

mig2004
Posts: 75
Joined: Tue Oct 01, 2013 10:36 pm

Re: E8400 CHESS ENGINE PERMANENT LADDER

Post by mig2004 » Mon Feb 24, 2020 12:14 pm

Improve move order near the root
1.0.3452
a day ago by Günther Demetz (committed by Joost VandeVondele)
master b8c00efa
a day ago in 6 min 31 sec


Filename: SF 200221-b8c0-DD40.zip

DOWNLOAD LINK: https://www.file-up.org/vf5m3y93qshl


Improve move order near the root

Current move histories are known to work well near the leaves, whilst at
higher depths they aren't very helpful. To address this problem this
patch introduces a table dedicated for what's happening at plies 0-3.
It's structured like mainHistory with ply index instead of color.
It get cleared with each new search and is filled during iterative
deepening at higher depths when recording successful quiet moves near
the root or traversing nodes which were in the principal variation
(ttPv).

Medium TC (20+0.2):
https://tests.stockfishchess.org/tests/ ... 2810d096dc
LLR: 2.94 (-2.94,2.94) {-0.50,1.50}
Total: 100910 W: 16682 L: 16376 D: 67852
Ptnml(0-2): 1177, 10983, 25883, 11181, 1231

LTC:
https://tests.stockfishchess.org/tests/ ... 2810d09714
LLR: 2.95 (-2.94,2.94) {0.25,1.75}
Total: 80444 W: 10495 L: 10095 D: 59854
Ptnml(0-2): 551, 7479, 23803, 7797, 592

closes #2557. Bench: 4705960. pb00067 authored and vondele committed ON 20-02-21. 1 parent ab930f8 commit b8c00ef



Commit lowPlyHistory (Take 4)
Info Take 4: restrict usage to first 4 plies on ttPv, revised bonus-logic. Using increased tc (half throughput) to reach needed depths
Submitter pb00067
TC 20+0.2
SPRT elo0: -0.50 alpha: 0.05 elo1: 1.50 beta: 0.05 (logistic)
LLR 2.94 [-2.94,2.94] (accepted)
Elo 0.75 [-0.73,2.09] (95%)
LOS 85.3%
Games 100910 [w:16.5%, l:16.2%, d:67.2%]
Link /tests/view/5e4d358790a0a02810d096dc


Details
id 5e4d358790a0a02810d096dc
new_tag lowPlyHistory (Take 4)
new_signature 4573156
new_options Hash=16
resolved_new 1f8205a5fe08b37436a08c22094df5a9a0c7f956
base_tag master (Simplify Futility Move Count)
base_signature 4762351
base_options Hash=16
resolved_base 10ead8a724671132df60ca2e23ddcad7eff7b3e5
sprt elo0: -0.50 alpha: 0.05 elo1: 1.50 beta: 0.05 state: accepted (logistic)
num_games 250000
tc 20+0.2
threads 1
book noob_3moves.epd
book_depth 8
( , 8 )
( 8 )
8


src code page: https://github.com/official-stockfish/S ... ef7c963319

MD5 HASH:

7D67CF16F11C6F29C4779FE40AE86015 benchmark.cpp
73DB3F8776A21155C03CEC6A722B1258 bitbase.cpp
681A742F40E2ED563C1D576F2740F9AB bitboard.cpp
18FCEFE5D7FEA188D927B9199B605B72 bitboard.h
4362E8EEA9812D6181FD663851C248FD endgame.cpp
CD08F0F040ABEA3CA4E5216047BAF3A1 endgame.h
A6339B3CEE1BA8DDAB4F7BEA31006B9C evaluate.cpp
B224B3AF7B4E4DDA2FA86DCCF7C114E8 evaluate.h
46E0D929B015AB0CD3E99F41253F6AA6 main.cpp
69B8CE1783491AA5B087506B07ABBA99 Makefile
DABC202C23A8BA875018EA1769912C22 material.cpp
D0252E530091B71DC540A9A8EB3ABD0C material.h
B0AF79E1550BA442300FD8E7991D66E9 misc.cpp
DF04A7C999EF95267CBE1D2FC9EDCCA2 misc.h
E5875D882309E5EC4F71A127A26688FE movegen.cpp
A32B107E1578E6A4D09B656DA5FA2B80 movegen.h
DABE42624D4040760E4ABA2C33F6FA64 movepick.cpp
A792C23A8AACB32E1D430D80853B9FC7 movepick.h
EBC53C6328A95F8DBBBE4202AAE7444E pawns.cpp
078056898F526BECFF74CEF0D4C3F237 pawns.h
F149636932154D8DE3E019BC61C1A71F position.cpp
B1D59F4FB3BD81F856883DD55C227548 position.h
043A2E7B9E40BCB8E627DB3765CEE28A psqt.cpp
7C89CEC9AF7E859C6D80F1FA723A43CE search.cpp
7AB329BE89832A2E2FB9C5C1457EB6EE search.h
D9327278A5685121102AFDEAAE399190 thread.cpp
831613A444BBA36F2C617A00C78A007D thread.h
20B8E29AEA0FB1E24502852210411187 thread_win32_osx.h
FFC62C1AB2D0FCEFCFDE07EFAFA39025 timeman.cpp
FF0EFB49F2C67352CD7153335EEB6A9B timeman.h
166855546FBD0CBF34E6B66BC252D2D4 tt.cpp
C32AB4C85FCF6AE9ABAE27C5785367F4 tt.h
A1833B5C8E4D285105A4B0357D591456 types.h
E3283F1031A878A585D8C094C0612520 uci.cpp
017176566B76B7C7D350C21D80F82EB0 uci.h
2F5458DC35A9AF447F4D6136B5ACA98B ucioption.cpp

MD5 DIGEST: DD4088D46EB37DD8CD3A43B26E7A1410

Post Reply