3Champs reloaded

[Vinvin]

Code: Select all

3Champs2013 2013.11.28 - 2013.12.03
                                   Score     Houdini 4 Pr Stockfish 24 Komodo 1142.
------------------------------------------------------------------------------------
 1: Houdini 4 Pro x64 x12        12.5 / 24   XXXXXXXXXXXX ===1=11=0==0 =1==0=======  (+4 -3 =17)
 2: Stockfish 241113 64 SSE4.2   12.0 / 24   ===0=00=1==1 XXXXXXXXXXXX 1====00=1=1=  (+5 -5 =14)
 3: Komodo 1142.00 64-bit x12-2  11.5 / 24   =0==1======= 0====11=0=0= XXXXXXXXXXXX  (+3 -4 =17)
------------------------------------------------------------------------------------
36 games: +8 =24 -4

Opening Report with SCID :

Code: Select all

3. Result Trends

3.1 Result lengths and frequencies

                 Score      Game length             Frequency       
                          1-0    =-=    0-1    1-0     =-=     0-1  
 All games       55.5%     81     84     85   22.2%   66.6%   11.1% 

3.2 Shortest wins (White)

  1:  1-0(51) Stockfish 241113 64 SSE4.2 - Houdini 4 Pro x64 x12, ? 2013 [21]
  2:  1-0(66) Komodo 1142.00 64-bit x12-2 - Houdini 4 Pro x64 x12, ? 2013 [10]
  3:  1-0(68) Komodo 1142.00 64-bit x12-2 - Stockfish 241113 64 SSE4.2, ? 2013 [14]
  4:  1-0(76) Stockfish 241113 64 SSE4.2 - Komodo 1142.00 64-bit x12-2, ? 2013 [25]
  5:  1-0(80) Stockfish 241113 64 SSE4.2 - Komodo 1142.00 64-bit x12-2, ? 2013 [13]

3.3 Shortest wins (Black)

  1:  0-1(48) Stockfish 241113 64 SSE4.2 - Houdini 4 Pro x64 x12, ? 2013 [24]
  2:  0-1(80) Stockfish 241113 64 SSE4.2 - Komodo 1142.00 64-bit x12-2, ? 2013 [21]
  3:  0-1(83) Houdini 4 Pro x64 x12 - Stockfish 241113 64 SSE4.2, ? 2013 [25]
  4:  0-1(127) Stockfish 241113 64 SSE4.2 - Houdini 4 Pro x64 x12, ? 2013 [3]

4. Moves and Themes

4.1 Move orders reaching the report position

There was only one move order reaching this position:
  1:   (36)

4.2 Moves from the report position

    Move      Frequency    Score  AvElo Perf AvYear Draw ECO
 1: e4         21: 58.3%   57.1%              2013  57% B00a 
 2: Nf3         7: 19.4%   64.2%              2013  71% A04  
 3: c4          5: 13.8%   50.0%              2013 100% A10  
 4: d4          3:  8.3%   33.3%              2013  67% A40a 
__________________________________________________________________
TOTAL:         36:100.0%   55.5%              2013  67%

4.3 Positional Themes

Frequency of themes in the first 20 moves of each game:
   Same-side castling:             78%    White Isolated Queen Pawn:      14%
   Opposite castling:               8%    Black Isolated Queen Pawn:      14%
   Kingside pawn storm:             8%    White Pawn on 5/6/7th rank:     36%
   Queens exchanged:               39%    Black Pawn on 2/3/4th rank:     25%
   Only one side has Bishop pair:   3%    Open c/d/e file:                50%

4.4 Endgames

Material at the end of each game:
                   P     BN      R   R,BN      Q   Q,BN    Q,R Q,R,BN
 Report games     0%    31%    28%    19%     3%    11%     3%     6%
 All games        0%    31%    28%    19%     3%    11%     3%     6%

5. Theory Table

-------------------------------------------------------------------------------
  +8 =24 -4 (20/36: 56%)
-------------------------------------------------------------------------------
     1        2        3        4        5        6        7        8     
-------------------------------------------------------------------------------
 1  d4       c4[2]    e4       Nf3      d5       Bf4      Be3      Nc3       3:
    d5[1]    dc4      Nc6      Bg4      Ne5      Ng6      Nf6      e5[3]    33%

 2  c4       Nc3      d4[5]    Nf3      Bg5      Rc1      Bd2      e3        5:
    e6[4]    Nf6      Bb4      b6       Bb7      h6       OO       c5[6]    50%

 3  Nf3      g3[8]    c4       Bg2      OO       Qb3      h3       d4        3:
    d5[7]    c6       Bg4      e6       Nf6      Qb6      Bh5      Be7[9]   67%

 4  ...      d4       g3[11]   Bg2      OO       a4       c4       bc3       4:
    Nf6      e6[10]   b5       d5       Be7      b4       bc3      Bb7[12]  63%

 5  e4       d4       Nd2      Ne4      Nf6      c3       Nf3      Be3       6:
    c6[13]   d5       de4      Nf6[14]  gf6      e5[15]   Qe7      Nd7[16]  75%

 6  ...      Nf3      d4[18]   Nd4      Nc6      Qf3      Nd2      Bd3       5:
    e5       Nc6[17]  ed4      Bc5      Qf6      bc6      d6       Qf3[19]  60%

 7  ...      Nf3[20]  d4       Nd4      Nc3      Be2      Qd3      OO        6:
    c5       e6[21]   cd4      Nf6[22]  d6       a6[23]   Nbd7     b6[24]   42%

 8  ...      ...      d4[25]   Nd4      c4[26]   Nc3      Be2      Qd4       4:
    ...      Nc6      cd4      g6       Nf6      d6       Nd4      Bg7[27]  50%

-------------------------------------------------------------------------------

[ouachita]

I've been kicking around one like this:

1. 16 cores
2. ponder, off?
3. TC: 150+30
4. 16 or 32GB hash table
6. Tablebases: Syzygy 3-4-5 TB.
7. Opening test suites: Noomen - six moves?
8. Color reversal for each opening.
9. Engine Settings: Will need some input on this.
10. 100 games each?

[lkaufman]

I've been kicking around one like this:

1. 16 cores
2. ponder, off?
3. TC: 150+30
4. 16 or 32GB hash table
6. Tablebases: Syzygy 3-4-5 TB.
7. Opening test suites: Noomen - six moves?
8. Color reversal for each opening.
9. Engine Settings: Will need some input on this.
10. 100 games each?

Note that ponder. off? is not a choice if you want to use 16 cores per engine, it is a must. You can only use 8 per engine with ponder on, which is why I recommend ponder off. One other point: most of us believe that it is important to run with hyperthreading turned off; otherwise some engines might "steal" resources from others and in any case it just adds an element of the unknown when hyperthreading is on (as it usually will be if you do nothing).

[Hugo]

Very impressive machine you have and nice event - can you also show us a conventional cross table so we can readily see how each engine is doing versus the other?

on the livegame transmission there is a crosstable

http://www.clemens-keck.de/livegames/

regards, C.K.

[Vinvin]

Clemens, did you find why Houdini crashed last night ?

Code: Select all

Qh1 Houdini 4 Pro x64 x12 caused an exception, game stopped. Houdini 4 Pro x64 x12 caused an exception, game stopped. -0.87/33 22 *

[Hugo]

Clemens, did you find why Houdini crashed last night ?

Code: Select all

Qh1 Houdini 4 Pro x64 x12 caused an exception, game stopped. Houdini 4 Pro x64 x12 caused an exception, game stopped. -0.87/33 22 *

Hi Vincent

I dont think that Houdini crashed. When I woke up and looked at the machine were the tournament is running on, I saw that all was gone, engines and gui. So I guess there was an operating system update or something like this.

All in all my intention for this tournament is entertainment. It should not been taken too much serious. Also the setup is far away from perfect.
The 3 participants are all champions no matter who is on top at the end.

best regards, Clemens

[Jouni]

Don't be too modest Clemens, this is like quick TCEC superfinal reprise! When I first looked at crosstable SF was third, but now it's leading by 2,5 points .

[lkaufman]

Hi all computerchess enthusiasts

for the honor to Don Dailey

4 days ago I started a new run of my 3Champs Tournament.
This time the 3 participants are running on a 24core machine, using 12cpu:

Round Robin Tournament 90min +15 sec, 50rounds, 150 games
Ponder = ON
engine = 12 cpu
hashtables = 4 GB
learning = OFF
hyperthreading = OFF
large pages = OFF
tablebases (engine) = OFF
tablebases GUI = ON ( small set 80GB)
opening book = Fritz 14 (5moves), mode "normal", no reversed color games
hardware = Intel E5-2697 v2 @ 3.00 GHz (turboMode)

Engines:

Houdini 4.0 pro, split depth 12
Stockfish 241113, sleeping threads idle ON, min split depth 7
Komodo 1142

livegames : http://www.clemens-keck.de/livegames/

ReplayZone : http://www.clemens-keck.de/ReplayZone/3champs2013.htm

This tour will take a while. I think 4 - 5 weeks.

regards, Clemens Keck

PS.: congratulations to Komodo 1142 for winning the grand TCEC !!

Wow, it's amazingly close now, all three within half a point after so many games! Just like TCEC, it may go down to the wire.

[Lyudmil Tsvetkov]

So far, with almost perfect openings book, we have 15 white wins and 6 black wins from 52 games. So that, if any tornament organizer gets distinctly different ratio of white/black wins, there certainly should be something wrong with the book. I.e., having equality of white/black wins, not to mention black wins being more than white, should be considered unnatural.

[ouachita]

Nice work Vincent - although a statistically small sample, your results are very much like TCEC results. I will eventually run a 150+30 test to see how these results change.