On-line engine blitz tourney July

[Joost Buijs]

The monthly on-line engine blitz tournament for July will take place on:

Saturday the 11th of July, 2:00 PM EST (Boston time) 20:00 CET (Amsterdam time)

To connect, use the following command:

winboard -zp -ics -icshost nightmare-chess.nl -icshelper timeseal -fcp ENGINE.exe -fd ENGINEFOLDER -autoKibitz -keepAlive 30
(For UCI engines, add -fUCI)

If you don’t have an account, please submit a request to admin@nightmare-chess.nl.

Our tournament manager will be Eendje (Lisebeth @ ICC).

The format will be Swiss system, 4/1 time control, 11 rounds.

Lisebeth prefers to start the tournament on top of the hour, so please log in 15 minutes early. Late entries are not supported.

[Bart Weststrate]

Nice!

[Joost Buijs]

The tournament finished without problems. Belfish won the tournament as usual, Stockfish is still a little bit too strong for 'mortal' engines.

Code: Select all

:Tourney Players: Round 11 of 11 
:
:     Name              Rating Score Perfrm Upset  Results 
:     ----------------- ------ ----- ------ ------ ------- 
:  1 +BelFish           [2948]   8.5  [2940] [   0] +23b =04w =10w =03b +12w +11w =02b +09w =05b +07w +14b 
:  2 +ArasanX           [2872]   7.5  [2871] [  38] +08b +11w =03w =09b =05w =04b =01w =13w =06b +10w +15b 
:  3 +StrongMalt        [2787]   7.0  [2831] [ 165] +17w +07w =02b =01w =11b =05b =06w -04b +18w +13w =09b 
:  4 +Rebel             [2769]   7.0  [2828] [ 219] +19w =01b =09w +14w =06b =02w =11b +03w =07b =05w =13b 
:  5 +Viper             [2726]   7.0  [2811] [ 406] +22w =10b =06w +17b =02b =03w =09b +11w =01w =04b =12w 
:  6 +WaspX             [2818]   6.5  [2669] [  39] =15b +24w =05b =13w =04w =07b =03b =12w =02w =09b +19w 
:  7 +Lunar             [2750]   6.5  [2636] [  90] +20w -03b +18w =12b =09w =06w =14b +08b =04w -01b +22b 
:  8 +WaDuuttie         [2436]   6.5  [2455] [ 172] -02w +21b +19w =10b -14w +17b -13b -07w +22b +24w +20b
:  9 +NightmareX        [2843]   6.0  [2794] [  14] +14b =12w =04b =02w =07b +10w =05w -01b =13b =06w =03w 
: 10 +Goldbar           [2780]   6.0  [2670] [  84] +18w =05w =01b =08w =13b -09b +15w =14w =11b -02b +23w 
: 11 +Samoza            [2753]   6.0  [2622] [  38] +21w -02b +17w +23b =03w -01b =04w -05b =10w =12b +24b 
: 12 +Marvin            [2752]   6.0  [2664] [  78] +16w =09b =13w =07w -01b =15b +17w =06b =14w =11w =05b 
: 13 +Cheng             [2721]   6.0  [2629] [ 252] =24w +15w =12b =06b =10w =14b +08w =02b =09w -03b =04w 
: 14 +clrsrc            [2414]   6.0  [2662] [ 695] -09w +16b +15w -04b +08b =13w =07w =10b =12b +18w -01w 
: 15 +Myrddin           [2409]   6.0  [2533] [ 375] =06w -13b -14b +18w +19b =12w -10b +16b +20w +21b -02w 
: 16 +Gyatso            [2126]   5.5  [2282] [ 940] -12b -14w =20b +24w -18b +19w +21b -15w +23b -22w +17b
: 17 +Arminius          [2396]   5.0  [2334] [ 104] -03b +20w -11b -05w +24b -08w -12b +22w +21w +23b -16w 
: 18 +Orion             [2354]   5.0  [2363] [ 146] -10b +22w -07b -15b +16w +20b +23w =19w -03b -14b =21w 
: 19 +jackychess        [2282]   5.0  [2273] [ 254] -04b +23w -08b +21w -15w -16b +22w =18b +24b =20w -06b 
: 20 +atomFloyd         [2079]   4.5  [2215] [ 638] -07b -17b =16w +22w =21b -18w +24b +23w -15b =19b -08w 
: 21 +Euwe              [2266]   4.0  [2162] [ 278] -11b -08w +22b -19b =20w +23b -16w +24w -17b -15w =18b 
: 22 +Vice              [2003]   3.0  [2142] [ 620] -05b -18b -21w -20b +23w +24w -19b -17b -08w +16b -07w 
: 23 +Izerin            [2500]   1.0  [2003] [   0] -01w -19b +24b -11w -22b -21w -18b -20b -16w -17w -10b 
: 24 +MicroMaxXb        [1648]   0.5  [2034] [ 536] =13b -06b -23w -16b -17w -22b -20w -21b -19w -08b -11w
:
:     Average Rating    2518.0 

See you all next time.

[swisschess]

Nice tournament, Joost! 24 engines in an 11-round Swiss makes for a solid field — BelFish at 8.5/11 is impressive, and that 2800+ performance rating from StrongMalt (2787 → 2831) with 165 upset points is a great story.

The crosstable format is really clean. I particularly like seeing the upset column — it gives a sense of which engines are punching above their rating.

Full disclosure: I built a free web-based Swiss tool at swiss-chess — it handles pairings, Buchholz tie-breaks, and live standings. It's aimed more at OTB club organizers than engine tournaments, but the pairing logic is the same. If anyone here runs Swiss events (human or engine) and wants a quick browser-based option, it's free with no signup required.

Keep these monthly tourneys coming — they're a great community resource.

[Rebel]

Thanks Joost for hosting, great job and "Eendje" the friendly arbiter.

Decided to participate with a somewhat stronger net than Rebel-16.3 created about a year ago but never released plus Chris fix to play with the maximum threads.

3 wins, 8 draws.

But these draws.... in 6 of them Rebel was out-booked by many moves all ending up in drawish positions with nothing to play for.

It's not the first time I noticed opening books play a major role in this tournament.

[Joost Buijs]

3 wins, 8 draws.

But these draws.... in 6 of them Rebel was out-booked by many moves all ending up in drawish positions with nothing to play for.

It's not the first time I noticed opening books play a major role in this tournament.

Books are part of the equation too. Personally I don't see much difference between a book based on selfplay games and a network based on selfplay games.

Of course it's possible to create a bookless tournament if you want to compare engines solely on their evaluation characteristics.

I wonder how many networks are really trained on selfplay games and not on games downloaded from lczero.org or Stockfish binpacks.

One of the reasons that I don't work on my engine anymore is that nowadays all engines are more or less the same, nothing new under the sun. Out of nostalgia I keep the server running, but that's currently all I do with chess.

[Richard Allbert]

The tournament finished without problems. Belfish won the tournament as usual, Stockfish is still a little bit too strong for 'mortal' engines.

Code: Select all

:Tourney Players: Round 11 of 11 
:
:     Name              Rating Score Perfrm Upset  Results 
:     ----------------- ------ ----- ------ ------ ------- 
:  1 +BelFish           [2948]   8.5  [2940] [   0] +23b =04w =10w =03b +12w +11w =02b +09w =05b +07w +14b 
:  2 +ArasanX           [2872]   7.5  [2871] [  38] +08b +11w =03w =09b =05w =04b =01w =13w =06b +10w +15b 
:  3 +StrongMalt        [2787]   7.0  [2831] [ 165] +17w +07w =02b =01w =11b =05b =06w -04b +18w +13w =09b 
:  4 +Rebel             [2769]   7.0  [2828] [ 219] +19w =01b =09w +14w =06b =02w =11b +03w =07b =05w =13b 
:  5 +Viper             [2726]   7.0  [2811] [ 406] +22w =10b =06w +17b =02b =03w =09b +11w =01w =04b =12w 
:  6 +WaspX             [2818]   6.5  [2669] [  39] =15b +24w =05b =13w =04w =07b =03b =12w =02w =09b +19w 
:  7 +Lunar             [2750]   6.5  [2636] [  90] +20w -03b +18w =12b =09w =06w =14b +08b =04w -01b +22b 
:  8 +WaDuuttie         [2436]   6.5  [2455] [ 172] -02w +21b +19w =10b -14w +17b -13b -07w +22b +24w +20b
:  9 +NightmareX        [2843]   6.0  [2794] [  14] +14b =12w =04b =02w =07b +10w =05w -01b =13b =06w =03w 
: 10 +Goldbar           [2780]   6.0  [2670] [  84] +18w =05w =01b =08w =13b -09b +15w =14w =11b -02b +23w 
: 11 +Samoza            [2753]   6.0  [2622] [  38] +21w -02b +17w +23b =03w -01b =04w -05b =10w =12b +24b 
: 12 +Marvin            [2752]   6.0  [2664] [  78] +16w =09b =13w =07w -01b =15b +17w =06b =14w =11w =05b 
: 13 +Cheng             [2721]   6.0  [2629] [ 252] =24w +15w =12b =06b =10w =14b +08w =02b =09w -03b =04w 
: 14 +clrsrc            [2414]   6.0  [2662] [ 695] -09w +16b +15w -04b +08b =13w =07w =10b =12b +18w -01w 
: 15 +Myrddin           [2409]   6.0  [2533] [ 375] =06w -13b -14b +18w +19b =12w -10b +16b +20w +21b -02w 
: 16 +Gyatso            [2126]   5.5  [2282] [ 940] -12b -14w =20b +24w -18b +19w +21b -15w +23b -22w +17b
: 17 +Arminius          [2396]   5.0  [2334] [ 104] -03b +20w -11b -05w +24b -08w -12b +22w +21w +23b -16w 
: 18 +Orion             [2354]   5.0  [2363] [ 146] -10b +22w -07b -15b +16w +20b +23w =19w -03b -14b =21w 
: 19 +jackychess        [2282]   5.0  [2273] [ 254] -04b +23w -08b +21w -15w -16b +22w =18b +24b =20w -06b 
: 20 +atomFloyd         [2079]   4.5  [2215] [ 638] -07b -17b =16w +22w =21b -18w +24b +23w -15b =19b -08w 
: 21 +Euwe              [2266]   4.0  [2162] [ 278] -11b -08w +22b -19b =20w +23b -16w +24w -17b -15w =18b 
: 22 +Vice              [2003]   3.0  [2142] [ 620] -05b -18b -21w -20b +23w +24w -19b -17b -08w +16b -07w 
: 23 +Izerin            [2500]   1.0  [2003] [   0] -01w -19b +24b -11w -22b -21w -18b -20b -16w -17w -10b 
: 24 +MicroMaxXb        [1648]   0.5  [2034] [ 536] =13b -06b -23w -16b -17w -22b -20w -21b -19w -08b -11w
:
:     Average Rating    2518.0 

See you all next time.

Thanks a lot for hosting this!

Vice had one point from pure luck, Gyatso flagged.

I've been working a lot on Vice the last few months - fixing the bugs, adding ponder and quite a few search features / lots of testing, but wow, all the other engines are so strong. From my testing, Vice is about 400 elo better than before, nevertheless, no chance here!!

Even Euwe, one place above Vice, is 2800+ on CCRL.

It's amazing how good people are at writing these things, there are so many really strong programs. Nevertheless, watching the games remains really fascinating,

It's nice to play in the same tournament as Rebel - I can remember printing the "How Rebel Plays Chess" web pages in the early 2000s, and reading it again and again.

Hopefully there will be some more engines < 2400 at some point



Thanks again,

Richard

[Rebel]

3 wins, 8 draws.

But these draws.... in 6 of them Rebel was out-booked by many moves all ending up in drawish positions with nothing to play for.

It's not the first time I noticed opening books play a major role in this tournament.

Books are part of the equation too. Personally I don't see much difference between a book based on selfplay games and a network based on selfplay games.

Of course it's possible to create a bookless tournament if you want to compare engines solely on their evaluation characteristics.

I wonder how many networks are really trained on selfplay games and not on games downloaded from lczero.org or Stockfish binpacks.

One of the reasons that I don't work on my engine anymore is that nowadays all engines are more or less the same, nothing new under the sun. Out of nostalgia I keep the server running, but that's currently all I do with chess.

Is it possible to download the PGN of all games played ?

[mar]

I wonder how many networks are really trained on selfplay games and not on games downloaded from lczero.org or Stockfish binpacks.

can't speak for others, but Cheng and Isa are 100% pure self-play seeded with their own HCEs with zero external influence
(not counting datagen influenced by probing syzygy, of course)

a word of caution: last time I censored (per advice in chat) micromaxXb to avoid the annoying shouts it made each move,
yesterday this caused the Cheng-umax game not starting automatically, delaying the tournament.
then I messed up the manual match without specifying time control so game was aborted and adjourned so the censor adventure
cost me a half-point. I suggest anyone who also censored u-max in the past to uncensor in order to avoid similar problems

[Joost Buijs]

Is it possible to download the PGN of all games played ?

No, I dont think this is possible because many games played are not available for download, and I don't see what this has to do with training a neural-network, you don't need an infinite number of games to train a neural-net. Of course more games are usually better, but above a certain threshold it doesn't get better anymore.

A network trains on sub-pattterns that occur in positions, many positions contain the same sub-patterns, it's important that the positions you train on are labeled correctly, e.g. +1 for a win, 0 for a draw and -1 for a loss. When you use a weak program to produce the games you derive these positions from, the labels will be way off, and you will get a weak network.