Stockfish time management is BADLY broken

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

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Jouni
Posts: 2121
Joined: Wed Mar 08, 2006 7:15 pm

Stockfish time management is BADLY broken

Post by Jouni » Thu Oct 12, 2017 7:34 am

I tried to play 22.9.2017 version with 60 moves in minute level. I must stop, when 15% of games were lost by time against Houdini.
Example game:

[pgn]
[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "Houdini 6 x64-pext"]
[Black "Stockfish 220917 64 BMI2"]
[Result "1-0"]
[SetUp "1"]
[FEN "rn1qkbnr/pp1bpppp/3p4/1Bp5/4P3/5N2/PPPP1PPP/RNBQK2R w KQkq - 0 1"]
[PlyCount "119"]

1. c4 Nf6 2. Nc3 Bxb5 3. cxb5 a6 4. bxa6 Nxa6 5. d4 e6 6. Be3 Be7 7. O-O O-O 8.
a3 Ng4 9. Qb3 Nxe3 10. fxe3 Rb8 11. Qa4 Qb6 12. Rf2 Qc6 13. Qxc6 bxc6 14. Rc2
d5 15. g3 Rfc8 16. Kg2 c4 17. Rf1 Rb6 18. Rff2 Nc7 19. Na4 Ra6 20. Nc3 Nb5 21.
Nxb5 cxb5 22. Nd2 Rd8 23. Nb1 dxe4 24. Nc3 Rb6 25. Nxe4 f5 26. Nc3 Kf7 27. h3
g6 28. g4 Bg5 29. Kf3 Ke7 30. Rce2 Rf8 31. Kg3 Bf6 32. Rf1 Ra8 33. Kf4 Ra5 34.
Rd1 b4 35. axb4 Rxb4 36. gxf5 gxf5 37. e4 Bxd4 38. Nd5+ exd5 39. exd5+ Kd6 40.
Rxd4 Rxd5 41. Rxd5+ Kxd5 42. Kxf5 Kd4 43. Kg5 Kd3 44. Rg2 Kd4 45. Re2 c3 46.
bxc3+ Kxc3 47. Re7 Rb5+ 48. Kh6 Kd4 49. h4 Rf5 50. h5 Kd5 51. Rxh7 Ke6 52. Rg7
Kf6 53. Rg1 Kf7 54. Ra1 Kg8 55. Kg6 Rf7 56. h6 Rg7+ 57. Kf5 Rf7+ 58. Kg6 Rg7+
59. Kf5 Rb7 60. Kg6 1-0

[/pgn]

This game was draw already 10 moves before time control (I have syzygy installed and cached to RAM) :o.
Jouni

Guenther
Posts: 3280
Joined: Wed Oct 01, 2008 4:33 am
Location: Regensburg, Germany
Full name: Guenther Simon
Contact:

Re: Stockfish time management is BADLY broken

Post by Guenther » Thu Oct 12, 2017 8:05 am

Jouni wrote:I tried to play 22.9.2017 version with 60 moves in minute level. I must stop, when 15% of games were lost by time against Houdini.
Example game:

[pgn]
[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "Houdini 6 x64-pext"]
[Black "Stockfish 220917 64 BMI2"]
[Result "1-0"]
[SetUp "1"]
[FEN "rn1qkbnr/pp1bpppp/3p4/1Bp5/4P3/5N2/PPPP1PPP/RNBQK2R w KQkq - 0 1"]
[PlyCount "119"]

1. c4 Nf6 2. Nc3 Bxb5 3. cxb5 a6 4. bxa6 Nxa6 5. d4 e6 6. Be3 Be7 7. O-O O-O 8.
a3 Ng4 9. Qb3 Nxe3 10. fxe3 Rb8 11. Qa4 Qb6 12. Rf2 Qc6 13. Qxc6 bxc6 14. Rc2
d5 15. g3 Rfc8 16. Kg2 c4 17. Rf1 Rb6 18. Rff2 Nc7 19. Na4 Ra6 20. Nc3 Nb5 21.
Nxb5 cxb5 22. Nd2 Rd8 23. Nb1 dxe4 24. Nc3 Rb6 25. Nxe4 f5 26. Nc3 Kf7 27. h3
g6 28. g4 Bg5 29. Kf3 Ke7 30. Rce2 Rf8 31. Kg3 Bf6 32. Rf1 Ra8 33. Kf4 Ra5 34.
Rd1 b4 35. axb4 Rxb4 36. gxf5 gxf5 37. e4 Bxd4 38. Nd5+ exd5 39. exd5+ Kd6 40.
Rxd4 Rxd5 41. Rxd5+ Kxd5 42. Kxf5 Kd4 43. Kg5 Kd3 44. Rg2 Kd4 45. Re2 c3 46.
bxc3+ Kxc3 47. Re7 Rb5+ 48. Kh6 Kd4 49. h4 Rf5 50. h5 Kd5 51. Rxh7 Ke6 52. Rg7
Kf6 53. Rg1 Kf7 54. Ra1 Kg8 55. Kg6 Rf7 56. h6 Rg7+ 57. Kf5 Rf7+ 58. Kg6 Rg7+
59. Kf5 Rb7 60. Kg6 1-0

[/pgn]

This game was draw already 10 moves before time control (I have syzygy installed and cached to RAM) :o.
Do you have the real pgn with score/time infos?
What GUI was used? What Syzygy tables are installed (6men or 5men etc). What hash etc.
Where does the compile com from? How much threads on what machine etc.
Much too much info is missing for useful detection/help.
https://rwbc-chess.de
https://rwbc-chess.de/chronology.htm
--------------------------------------------------
The troll explosion at talkchess:
https://docs.google.com/spreadsheets/d/ ... KSptBx9AUs

Volker Pittlik
Posts: 599
Joined: Wed Mar 08, 2006 8:10 pm
Location: Murten / Morat, Switzerland

Re: Stockfish time management is BADLY broken

Post by Volker Pittlik » Thu Oct 12, 2017 8:59 am

Jouni wrote:I tried to play 22.9.2017 version with 60 moves in minute level. I must stop, when 15% of games were lost by time against Houdini...
Strange. SF just finished 5330 games here at 40 moves per second without any error. My version is from 19.09.2017. If there is something broken in Stockfish then certainly between this two versions.

Volker

Guenther
Posts: 3280
Joined: Wed Oct 01, 2008 4:33 am
Location: Regensburg, Germany
Full name: Guenther Simon
Contact:

Re: Stockfish time management is BADLY broken

Post by Guenther » Thu Oct 12, 2017 9:03 am

Volker Pittlik wrote:
Jouni wrote:I tried to play 22.9.2017 version with 60 moves in minute level. I must stop, when 15% of games were lost by time against Houdini...
Strange. SF just finished 5330 games here at 40 moves per second without any error. My version is from 19.09.2017. If there is something broken in Stockfish then certainly between this two versions.

Volker
BTW 60...Rg7+ would have been a three time rep, if the responsible GUI had received that move in time.

Guenther
https://rwbc-chess.de
https://rwbc-chess.de/chronology.htm
--------------------------------------------------
The troll explosion at talkchess:
https://docs.google.com/spreadsheets/d/ ... KSptBx9AUs

Modern Times
Posts: 2554
Joined: Thu Jun 07, 2012 9:02 pm

Re: Stockfish time management is BADLY broken

Post by Modern Times » Thu Oct 12, 2017 11:41 am

I get time losses from Stockfish 180917 on a repeating time control. Ridiculous that basic time management should be broken.
.

Opinions expressed here are my own, and not necessarily those of the CCRL Group.

gotogo
Posts: 91
Joined: Tue Jan 26, 2010 4:03 am

Re: Stockfish time management is BADLY broken

Post by gotogo » Thu Oct 12, 2017 1:01 pm

Looks like Stockfish is doing everything it can to stop progress! One would imagine Stockfish is giving Komodo a chance to catch up hence tcec delay!

:shock: :cry: :oops:

User avatar
Eelco de Groot
Posts: 4245
Joined: Sun Mar 12, 2006 1:40 am
Location: Groningen

Re: Stockfish time management is BADLY broken

Post by Eelco de Groot » Thu Oct 12, 2017 1:18 pm

Hello Ray,

I know from experience that it is almost impossible to trigger time losses in 2 minute per player per game with Shredder. Is this another GUI? Dr. X had some trouble with Avast but for the CCRL I think we can exclude that. If it is not something broken with the interface code, I mean UCI support, which at the moment is not changed that I know of but Vondele is going to introduce some new things, (using exceptions as that is more standard C++ way of handling special situations) then I can only think that it could be the repeating time control. 60 milliseconds move overhead is a reserve for the whole game (meant for Fishtest) There is no other reserve. 60 ms is extremely small amount and Windows more critical than Linux. If you have repeating timeconrols; I'm not sure but possibly there is only one failsafe move overhead. If that is used in one timecontrol, Stockfish may go over the next. I don't know but I'm sure it has not been tested in Fishtest.

Regards, Eelco
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

Jouni
Posts: 2121
Joined: Wed Mar 08, 2006 7:15 pm

Re: Stockfish time management is BADLY broken

Post by Jouni » Thu Oct 12, 2017 1:43 pm

Ooops i made a mistake by setting Move Overhead=30, when default is now 100 :oops: . But SF loses some games also with 100. It's weird, when SF uses clearly less time as average than SF8 or Houdini! Some figures.

SFdev tpm=1045.1
SF8 tpm=1100.0
H6 tpm=1091.6

Note, that this traditional time control is also important not only stupid increment :) .
Jouni

syzygy
Posts: 4573
Joined: Tue Feb 28, 2012 10:56 pm

Re: Stockfish time management is BADLY broken

Post by syzygy » Thu Oct 12, 2017 3:06 pm

Jouni wrote:I tried to play 22.9.2017 version with 60 moves in minute level. I must stop, when 15% of games were lost by time against Houdini.
Do more of these time losses occur AT move 60 (like in the game you posted)?

If yes, then either there is a clear bug in SF's time allocation or the GUI you are using does not stick to the meaning of the various UCI timing parameters.

Please tell us what GUI you are using.

Guenther
Posts: 3280
Joined: Wed Oct 01, 2008 4:33 am
Location: Regensburg, Germany
Full name: Guenther Simon
Contact:

Re: Stockfish time management is BADLY broken

Post by Guenther » Thu Oct 12, 2017 3:40 pm

syzygy wrote:
Jouni wrote:I tried to play 22.9.2017 version with 60 moves in minute level. I must stop, when 15% of games were lost by time against Houdini.
Do more of these time losses occur AT move 60 (like in the game you posted)?

If yes, then either there is a clear bug in SF's time allocation or the GUI you are using does not stick to the meaning of the various UCI timing parameters.

Please tell us what GUI you are using.
Well I asked this already and requested the necessary infos...
https://rwbc-chess.de
https://rwbc-chess.de/chronology.htm
--------------------------------------------------
The troll explosion at talkchess:
https://docs.google.com/spreadsheets/d/ ... KSptBx9AUs

Post Reply