News about Rebel or Chess System Tal ?

CCRL testing - https://tlcv.net/16070/result-table

Results from Rebel 16 can be followed via my spreadsheet: https://docs.google.com/spreadsheets/d/11vXZe4Dm7nIXWIYD9EoZguR4xOuRWUQHpzSfQW2HaMo/edit?pli=1#gid=1460182227

You need to scroll down..

Ipmanchess wrote:

Results from Rebel 16 can be followed via my spreadsheet: https://docs.google.com/spreadsheets/d/11vXZe4Dm7nIXWIYD9EoZguR4xOuRWUQHpzSfQW2HaMo/edit?pli=1#gid=1460182227

You need to scroll down..

Looking good....

Admin wrote:

Looking good....

Time management of Rebel 16 looking bad...

After 3400 games, Rebel 16 used 178 secs in average in my testings (3min+1sec). Other engines in my ratinglist use (in average) 233 sec per game !!
So Rebel 16 uses only (around) 75% of the time, other engines use.

I used the GTS-tool by Deuterium-author. I made a small batch tool for starting it:
https://drive.google.com/file/d/1tPPqkRkfbilLUbscxUE8nvinhehbVddH/view?usp=sharing

(The tool needs games with "GameDuration"-TAG)

Average game duration of my ratinglist: 7min 46sec = 233 secs per color/engine
Average game duration of Rebel 16 gauntlet games: 6min 50sec = 205sec per color/engine, so 28 secs less. But, if the other engines does not play faster (why should they?), Rebel has 56 sec less (28 *2), because Rebel 16 is the only reason for the smaller amount of time...

Of course, 25% more time (with a fix in the time-management), would give Rebel 16 not much Elo (around 5 Elo, perhaps), but it is definitly a waste of gametime/Elo. And should get checked (and fixed) in the forthcoming CSTal release. Using the tool, I linked here, this can be done very easy. If cutechess(-cli) is used, each game has a GameDuration-TAG...

Have same findings.. plays too fast ,each 1000games it finish +/-30min. faster then other engines!
TC: 10s+1s and Rebel 16 ends with +30sec left ..so the 10sec. don't go down ,or plays the moves faster then 1s increment it gets.

For sure it do well ..but in this case Rebel could win easy some extra Elo by tuning TM.

Will see how it handle 5min.games that will start now on other system..

Thanks for those. The time control algorithms are 1997, and I don’t think we ever tested game in N, let alone plus increment. IIRC the move control for game in N is time left divided by 50 (or something). I’ll try decreasing the divisor by the ratio suggested in your post.

Admin wrote:

The next (and last) Rebel will be the next step in the cooperation between Chris and me. We have been working very hard to replace Toga and port the code base of Chess System Tal instead. The advantages of this move are, 1) clean compiles, 2) a stronger search and last but not least we now work together on creating stronger nets and I can do what I like most (and was my only goal to return to computer chess) getting the most out of NNUE programming.

On the latter, the last Rebel will come with the same version 10 net architecture as Rebel 15.1 but much improved, peppered with special HCE ProDeo evaluation stuff.

In the meantime Chris has created another even more brilliant net architecture, version 20 [!], currently already 30-35 elo stronger than net version 10, with no end in sight.

So this will be the new reality, 1+1 = 3, team work.

I like to present the first beta result.

[code]No. Name                 Win Draw Loss Unf.  Score Games       %
----------------------------------------------------------------
 1 BETA-ONE           +1525 =4180 -1295 *0 3615.0  

No idea how much elo that is.

But net version 20 beats them all.

Thank you Ed and Chris! You are 2 stars in chess engines history and I'm so happy that you are back!

Testing It right now on my TCEC 23 Active chess Gauntlet with all other bigs! http://t.ly/iToU-

Ed and chris, our two old fellows, are the best that could happen to us.

here the first results:

    Motor                                Punkte      Dr   St   Be   Dr   Re   Dr   Et   Ka   Dr   Ru   Lc   Re   Lc   Re   Ur   De   Wa   Ko   Bo   Wa   Ko   Ch   Bo   Eq   De   De   Ro   CS   Ro   In   Co   Na   De   CT    S-B
01: Dragon-2.5-64bit-avx2 [t4, mcts]     102,0/132 ···· ===1 0=== =0=0 ==== 11== 1==0 =11= ==0= 1=== ==0= =111 1==1 ==1= 1=11 101= 1111 111= 10=1 111= 1==1 1111 1111 1111 1111 1111 0111 11=1 111= 1111 1111 1111 1111 1111  5978,7
02: Stockfish_15_x64_avx2 [t4]           101,0/132 ===0 ···· 1110 ===0 ==01 =1== =111 ==1= 11== =1=1 11=1 0=1= ==01 =1=1 =0=1 ===0 ===1 1111 1=11 =110 =111 1110 11=1 1111 1111 1111 0111 1101 1111 1111 1111 1111 1111 1111  5970,0
03: Berserk-9-x64-avx2-pext [t4]         100,5/132 1=== 0001 ···· ===1 0=== =1=0 110= ===1 ==== =11= =1== ==1= ==== 1=11 =1== =111 1111 11== 1111 1011 10== =111 111= 0111 =111 11=1 1111 1111 1111 =111 1111 1111 1111 1111  5902,5
04: Dragon-3.1-64bit-avx2 [t4]           99,5/132  =1=1 ===1 ===0 ···· ==== ==0= =1=1 ==== 1==0 ===1 ==0= 1111 101= ==== 1101 ==11 1=10 1==1 1=== 11=1 11== 1111 =11= =111 1111 1111 1111 1111 1101 1111 1111 1=11 1111 1111  5851,7
05: Rebel 16 [t4, rebel.bin]             98,0/132  ==== ==10 1=== ==== ···· ===1 ==1= ==1= ===0 ==1= ==1= ==== ==== ==== 111= 1=11 1==1 ==== =1=1 1=1= ==== 1111 1111 1=11 =11= 1=11 1111 1111 11=1 1111 1111 1111 1111 1111  5709,0
06: Dragon-3.1-64bit-avx2 [t4, mcts]     98,0/132  00== =0== =0=1 ==1= ===0 ···· ==10 ==1= ==== ==== ==01 1==1 ==== 1==1 10=1 =011 1111 ===1 11=1 1=11 11== 11=1 1111 =111 1111 1==1 =111 1111 1111 1111 1111 1111 1111 1111  5614,7
07: Ethereal-13.75-pext-avx2 [t4]        94,5/132  0==1 =000 001= =0=0 ==0= ==01 ···· ==== ==== 0=0= 11== =1== =11= ==== 01=1 ==== 10== 11=1 111= 111= 1=1= 1111 11== 1111 1111 1111 1111 1111 =111 1111 1111 1111 1111 1111  5263,5
08: Kayra 1.5 Attack-avx2 [t4]           93,5/132  =00= ==0= ===0 ==== ==0= ==0= ==== ···· =1== =1== 1=0= 00== 110= 111= 11== ==== =011 11== 1=11 1111 111= 1111 1101 1111 1110 1110 1111 1101 0111 1111 1111 1111 1111 1111  5331,2
09: Dragon-2.5-64bit-avx2 [t4]           93,0/132  ==1= 00== ==== 0==1 ===1 ==== ==== =0== ···· 0=11 11== ==== ==1= 11=1 1=1= =1=0 1=10 1=11 11=1 1110 ===0 0100 1110 0111 1111 1111 1111 1111 1110 1111 1111 1111 1111 1101  5432,7
10: RubiChess20220813_64-avx2 [t4, nnue] 88,5/132  0=== =0=0 =00= ===0 ==0= ==== 1=1= =0== 1=00 ···· 1==1 1=== =111 ==== ===1 ==1= =0=1 010= =1== =101 =11= ===1 11=1 1111 01=1 1111 =111 1111 1111 0111 =111 1111 1111 1111  4994,0
11: Lc0 0.29.0 radeon gpu dx12 [753723]  88,0/132  ==1= 00=0 =0== ==1= ==0= ==10 00== 0=1= 00== 0==0 ···· 00== =1== =11= =1=1 ==== 110= 11== ==11 1=1= =10= ===0 111= 1=11 1111 111= 1111 1111 11=1 1111 1111 1111 1111 1111  4846,0
12: Rebel-15.1a-avx2 [t4, rebel.bin]     87,0/132  =000 1=0= ==0= 0000 ==== 0==0 =0== 11== ==== 0=== 11== ···· ==== ==10 =0== ===0 1=11 =1=0 1110 =110 1101 0=11 1=11 1111 1111 1111 1011 1111 1111 1111 1111 1111 1110 1111  4769,5
13: Lc0 0.28.2 radeon gpu dx12 [752187]  85,0/132  0==0 ==10 ==== 010= ==== ==== =00= 001= ==0= =000 =0== ==== ···· ==== ==1= ==== 11=0 =0== 1=== 1=== =1== 111= =111 =1=1 1110 1111 1=11 1111 1111 1111 1111 1111 1111 1111  4604,0
14: Rebel-15.1-avx2 [t4]                 83,0/132  ==0= =0=0 0=00 ==== ==== 0==0 ==== 000= 00=0 ==== =00= ==01 ==== ···· ==11 =0=1 10=1 =11= ===1 11=1 =10= ==1= 11== 1111 =111 ==11 11=1 =111 1111 11=1 1111 1111 1111 1111  4455,7
15: Uralochka3.38c-avx2 [t4]             80,5/132  0=00 =1=0 =0== 0010 000= 01=0 10=0 00== 0=0= ===0 =0=0 =1== ==0= ==00 ···· 1=== =1== =011 =1== =1== ===1 ==11 1111 1=1= 1111 1=1= =1=1 1111 1111 1111 11=1 1111 1111 1111  4227,5
16: DeepHiarcs15.0 [t4]                  80,0/132  010= ===1 =000 ==00 0=00 =100 ==== ==== =0=1 ==0= ==== ===1 ==== =1=0 0=== ···· =1=1 011= ===1 ===1 ===1 =1=0 ===1 111= =111 ==== ===1 =1=1 1=11 1111 1111 1111 111= 1111  4439,5
17: Wasp600-win-avx [t4]                 79,5/132  0000 ===0 0000 0=01 0==0 0000 01== =100 0=01 =1=0 001= 0=00 00=1 01=0 =0== =0=0 ···· ==== ==== 1=11 11=1 111= ==11 11=1 1111 1111 1111 1111 1111 1111 1111 1111 1111 1111  3978,7
18: Komodo-14.1-64bit [t4, mcts]         77,0/132  000= 0000 00== 0==0 ==== ===0 00=0 00== 0=00 101= 00== =0=1 =1== =00= =100 100= ==== ···· 11== =111 =11= 11== =111 1==1 0111 11== 1=11 101= 1111 1111 1111 1111 1111 1111  3956,7
19: Booot7 [t4]                          68,0/132  01=0 0=00 0000 0=== =0=0 00=0 000= 0=00 00=0 =0== ==00 0001 0=== ===0 =0== ===0 ==== 00== ···· =0=0 =1== 1=00 1=11 1=11 =111 ===1 111= 1111 1111 1111 1=11 1111 1111 1111  3248,2
20: Wasp550-windows-avx [t4]             67,5/132  000= =001 0100 00=0 0=0= 0=00 000= 0000 0001 =010 0=0= =001 0=== 00=0 =0== ===0 0=00 =000 =1=1 ···· 0110 =101 1==1 1111 111= 11== 1110 111= 11=1 1111 1111 1111 1111 1111  3189,5
21: Komodo-14.1-64bit [t4]               67,0/132  0==0 =000 01== 00== ==== 00== 0=0= 000= ===1 =00= =01= 0010 =0== =01= ===0 ===0 00=0 =00= =0== 1001 ···· 1001 101= 10=0 1101 1110 1111 0111 1111 1101 1111 1111 1111 1010  3534,0
22: Chiron-5-64bit [t4]                  61,5/132  0000 0001 =000 0000 0000 00=0 0000 0000 1011 ===0 ===1 1=00 000= ==0= ==00 =0=1 000= 00== 0=11 =010 0110 ···· ==1= =01= 1==1 1==0 =11= ==1= 1111 1=11 1111 1111 1111 1111  2892,5
23: Booot6 [t4]                          52,5/132  0000 00=0 000= =00= 0000 0000 00== 0010 0001 00=0 000= 0=00 =000 00== 0000 ===0 ==00 =000 0=00 0==0 010= ==0= ···· 011= =1=1 =1=1 01== 1111 11== 1111 1111 1111 1111 =111  2201,7
24: Equinox 3.30-x64 [t4]                47,0/132  0000 0000 1000 =000 0=00 =000 0000 0000 1000 0000 0=00 0000 =0=0 0000 0=0= 000= 00=0 0==0 0=00 0000 01=1 =10= 100= ···· 10== 01=0 ==11 1=1= 0111 1111 1111 1111 1111 1111  1825,2
25: Deep Rybka 4.1 x64 [t4]              42,5/132  0000 0000 =000 0000 =00= 0000 0000 0001 0000 10=0 0000 0000 0001 =000 0000 =000 0000 1000 =000 000= 0010 0==0 =0=0 01== ···· =1=1 =111 =001 1111 1=11 0111 1111 1111 1101  1608,0
26: DeepShredder13UCIx64 [t4, buffer6]   42,0/132  0000 0000 00=0 0000 0=00 0==0 0000 0001 0000 0000 000= 0000 0000 ==00 0=0= ==== 0000 00== ===0 00== 0001 0==1 =0=0 10=1 =0=0 ···· =011 1==1 1010 1110 1101 1011 1101 1111  1773,2
27: Rodent-iv-x64 [t4]                   40,0/132  1000 1000 0000 0000 0000 =000 0000 0000 0000 =000 0000 0100 0=00 00=0 =0=0 ===0 0000 0=00 000= 0001 0000 =00= 10== ==00 =000 =100 ···· 0=10 =111 1111 1111 1111 =111 1111  1503,0
28: CSTal3 5'                            38,0/132  00=0 0010 0000 0000 0000 0000 0000 0010 0000 0000 0000 0000 0000 =000 0000 =0=0 0000 010= 0000 000= 1000 ==0= 0000 0=0= =110 0==0 1=01 ···· 1111 1111 0111 1111 =111 1111  1308,5
29: Rodent-iv-x64 [t4, tal]              25,5/132  000= 0000 0000 0010 00=0 0000 =000 1000 0001 0000 00=0 0000 0000 0000 0000 0=00 0000 0000 0000 00=0 0000 0000 00== 1000 0000 0101 =000 0000 ···· =001 1011 1111 1=10 1111  934,50
30: Inanis 1.1.1 win64 [t4]              18,5/132  0000 0000 =000 0000 0000 0000 0000 0000 0000 1000 0000 0000 0000 00=0 0000 0000 0000 0000 0000 0000 0010 0=00 0000 0000 0=00 0001 0000 0000 =110 ···· =10= 10=0 =111 1111  522,50
31: Colossus2022a [t4]                   15,5/132  0000 0000 0000 0000 0000 0000 0000 0000 0000 =000 0000 0000 0000 0000 00=0 0000 0000 0000 0=00 0000 0000 0000 0000 0000 1000 0010 0000 1000 0100 =01= ···· 10== 1111 1100  382,50
32: Nawito-22.07-winx64                  11,5/132  0000 0000 0000 0=00 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0100 0000 0000 0000 01=1 01== ···· ==10 =111  219,50
33: Delfi                                9,5/132   0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0001 0000 0000 0000 000= 0000 0000 0000 0000 0000 0000 0000 0000 0000 0010 =000 =000 0=01 =000 0000 ==01 ···· 1010  296,50
34: CT800_V1.44_x64                      9,0/132   0000 0000 0000 0000 0000 0000 0000 0000 0010 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0101 0000 =000 0000 0010 0000 0000 0000 0000 0000 0011 =000 0101 ····  351,50

2244 Partien gespielt / Turnier beendet
Name des Turniers: Ryzen7 5700
Ort/ Land: LAPTOP-JTQQP8M7, Deutschland
Spielstufe: Turnier 40/4
Hardware: AMD Ryzen 7 5700U with Radeon Graphics  mit 4.096 MB Speicher

BUG-FIX - rebel-16a

Fixes time control sudden death + increment

2 matches of each ~200 games with 30+0.5, results speaks for itself

Match-one

Code:

No. Name         Win Draw Loss Unf.  Score Games       %
--------------------------------------------------------
  1 rebel-16-TC  +29 =176  -15   *0  117.0   220   53.2%
  2 Rebel-16     +15 =176  -29   *0  103.0   220   46.8%

Engine        Depth       Time   Games     Moves  Average Forfeit  Book Depth     MIDG   EARLY    ENDG    LATE
rebel-16-TC   21.14    3:02:42     220     13601    0.81     0     2113  9.60    19.15 | 19.34 | 21.51 | 26.12
Rebel-16      20.65    2:25:52     220     13629    0.64     0     2113  9.60    18.48 | 18.93 | 21.14 | 25.89

Match-two

Code:

No. Name         Win Draw Loss Unf.  Score Games       %
--------------------------------------------------------
  1 rebel-16-TC  +34 =150  -17   *0  109.0   201   54.2%
  2 Rebel-16     +17 =150  -34   *0   92.0   201   45.8%
  
Engine        Depth       Time   Games     Moves  Average Forfeit  Book Depth     MIDG   EARLY    ENDG    LATE
rebel-16-TC   21.86    2:55:36     201     13341    0.79     0     1941  9.66    19.49 | 19.89 | 21.19 | 27.11
Rebel-16      21.26    2:21:17     201     13349    0.64     0     1950  9.70    18.58 | 19.06 | 20.69 | 26.94

http://rebel13.nl/dl/rebel-16a.7z

Mclane wrote:

here the first results:



2244 Partien gespielt / Turnier beendet
Name des Turniers: Ryzen7 5700
Ort/ Land: LAPTOP-JTQQP8M7, Deutschland
Spielstufe: Turnier 40/4
Hardware: AMD Ryzen 7 5700U with Radeon Graphics  mit 4.096 MB Speicher

Thorsten is really a BACK TO THE FUTURE ... How many nice souvenirs of 25 years ago, when I was a Ed's Rebel Italian reseller Rebel was so wonderful (also the gui)

Even my first results against Stockfish & derivatives are good

I have played 1400 games with Rebel 16 so far.
I use Cutechess (GUI), 8 games are played simultaneously.

In 3 games (against different opponents) it happened that Rebel 16.0 stopped calculating already on search depth 1,
mate values and normal evaluations were displayed alternately.
I then had to restart the GUI to continue the match.

Have other testers had the same experience ?

I had no freezes, but something is definitly wrong with Rebel. Look at move 109 by Rebel 16 (black)...
Drawn position, but a thinking depth of 1 after 1.2 seconds and a bogus evaluation: Rh6+ {-M124/1 1.2s}
(This happens again and again in my testrun) (I use cutechess-cli)

(And Rebel 16a moves still too fast. Now, 7min 5secs per game in my testings (better than Rebel 16 (6min 50sec))
Normal average value in my ratinglist is 7min 45secs

So,  I aborted the Rebel testruns. Makes no sense right now. Engine is buggy. And time management is still bad.

[Event "CPU-engine testing"]
[Site "?"]
[Date "2022.11.30"]
[Round "6"]
[White "Berserk 10 avx2"]
[Black "Rebel 16 avx2"]
[Result "1/2-1/2"]
[ECO "C24"]
[Opening "Bishop's Opening"]
[PlyCount "277"]
[Termination "adjudication"]
[TimeControl "180+1"]
[Variation "Urusov gambit"]

1. e4 {book} e5 {book} 2. Bc4 {book} Nf6 {book} 3. d4 {book} exd4 {book}
4. Nf3 {book} Nxe4 {book} 5. Qxd4 {book} Nf6 {book} 6. Nc3 {-0.64/25 15s}
c6 {+0.47/20 8.0s} 7. Bg5 {-0.50/20 2.2s} d5 {+0.85/21 6.9s}
8. O-O-O {-0.56/20 3.3s} Be6 {+0.47/22 4.0s} 9. Rhe1 {-0.24/24 7.7s}
Be7 {+0.31/20 2.7s} 10. Qh4 {-0.32/23 4.5s} h6 {+0.32/20 7.2s}
11. Bd3 {-0.01/27 13s} Rg8 {0.00/21 7.1s} 12. Rxe6 {+0.01/27 15s}
fxe6 {-0.03/19 3.6s} 13. Bg6+ {+0.03/25 4.8s} Kd7 {-0.56/21 2.5s}
14. Bf4 {-0.01/23 5.6s} Kc8 {-0.28/19 4.6s} 15. Nd4 {+0.01/28 6.8s}
Nfd7 {-0.85/22 4.5s} 16. Qh3 {+0.01/29 4.7s} Bg5 {0.00/20 2.3s}
17. Nxe6 {0.00/28 1.5s} Bxf4+ {0.00/21 3.0s} 18. Nxf4 {-0.01/25 1.6s}
Qg5 {0.00/23 3.9s} 19. g3 {0.00/27 1.5s} Rd8 {0.00/23 2.9s}
20. Rd2 {+0.01/27 4.6s} Qf6 {0.00/21 2.9s} 21. Ncxd5 {0.00/27 7.8s}
cxd5 {-0.82/19 2.2s} 22. Nxd5 {-0.01/26 1.4s} Qf8 {0.00/23 2.2s}
23. Bf5 {-0.02/27 1.6s} Nc6 {0.00/23 4.0s} 24. Bxd7+ {+0.09/24 3.1s}
Rxd7 {0.00/24 1.9s} 25. Nf6 {+0.15/26 4.8s} gxf6 {0.00/25 1.9s}
26. Rxd7 {+0.17/24 1.7s} Kb8 {0.00/26 4.6s} 27. Qe6 {+0.15/24 2.4s}
a5 {+0.37/20 2.0s} 28. Qd6+ {+0.01/27 3.5s} Qxd6 {+0.55/21 2.0s}
29. Rxd6 {+0.02/20 1.3s} Kc7 {+0.40/20 2.0s} 30. Rxf6 {+0.11/22 2.2s}
Rh8 {+0.55/21 2.6s} 31. h4 {+0.01/22 2.3s} b5 {0.00/20 5.4s}
32. h5 {+0.39/20 2.1s} a4 {+0.16/22 1.9s} 33. a3 {+0.29/26 7.8s}
Ne5 {0.00/23 3.2s} 34. Kd2 {+0.35/23 2.5s} Ng4 {0.00/25 4.1s}
35. Rf5 {+0.29/25 2.6s} Kb6 {0.00/22 4.5s} 36. Rf4 {+0.12/22 2.1s}
Rg8 {0.00/23 3.3s} 37. Ke2 {+0.46/24 3.2s} Ne5 {0.00/21 4.3s}
38. b3 {+0.47/24 2.1s} Ra8 {0.00/23 2.2s} 39. Rf6+ {+0.39/24 2.4s}
Kb7 {0.00/23 2.4s} 40. bxa4 {+0.62/21 1.2s} bxa4 {0.00/24 1.5s}
41. Rxh6 {+0.69/23 2.8s} Nc4 {0.00/25 1.5s} 42. Kd3 {+0.47/26 6.7s}
Nxa3 {0.00/20 1.6s} 43. Re6 {+0.44/23 3.3s} Nb5 {0.00/22 1.6s}
44. Re1 {+0.46/22 1.4s} Rd8+ {-0.07/20 4.4s} 45. Kc4 {+0.46/22 1.1s}
Kc6 {0.00/23 1.6s} 46. Kb4 {+0.33/22 2.3s} Nd4 {-0.16/21 1.5s}
47. Rc1 {+0.50/22 2.6s} Rb8+ {-0.31/21 1.5s} 48. Kc4 {+0.46/21 1.3s}
Nb5 {-0.13/22 1.5s} 49. h6 {+0.46/23 3.6s} Nd6+ {-0.23/19 1.7s}
50. Kd3 {+0.38/23 1.3s} Rh8 {-0.15/21 1.5s} 51. Rh1 {+0.36/23 2.2s}
a3 {-0.15/21 1.5s} 52. h7 {+0.35/24 8.7s} a2 {0.00/20 1.5s}
53. Ra1 {+0.22/25 2.4s} Rxh7 {0.00/19 1.5s} 54. Rxa2 {+0.20/23 1.2s}
Rh8 {-0.16/20 1.5s} 55. Ra6+ {+0.33/23 3.6s} Kc7 {-0.08/17 2.1s}
56. Ke3 {+0.35/23 4.5s} Re8+ {-0.03/21 1.9s} 57. Kf3 {+0.31/21 0.99s}
Rc8 {0.00/22 2.1s} 58. c4 {+0.33/21 1.4s} Kd7 {0.00/22 1.4s}
59. Ra2 {+0.14/22 6.9s} Rxc4 {0.00/20 2.1s} 60. Re2 {+0.14/22 1.8s}
Kc6 {+0.12/20 1.4s} 61. Rd2 {+0.13/23 3.0s} Ne4 {+0.14/21 3.5s}
62. Rd3 {+0.15/23 1.6s} Ra4 {+0.18/21 2.0s} 63. Ke3 {+0.13/24 1.8s}
Nc5 {+0.14/21 1.4s} 64. Rd2 {+0.13/24 1.5s} Re4+ {+0.09/23 3.0s}
65. Kf3 {+0.13/22 0.24s} Rc4 {+0.07/20 1.3s} 66. Ke3 {+0.12/24 5.0s}
Rb4 {+0.02/22 1.7s} 67. Ra2 {+0.12/21 2.3s} Kd5 {+0.09/22 3.2s}
68. Kf3 {+0.09/23 3.4s} Rb3+ {+0.08/22 2.4s} 69. Kg4 {+0.08/21 1.2s}
Ne4 {0.00/23 2.8s} 70. Kf4 {+0.07/21 0.80s} Nd6 {0.00/20 1.8s}
71. Ra5+ {+0.08/22 2.8s} Ke6 {+0.11/19 1.3s} 72. Ra6 {+0.07/23 6.0s}
Rb2 {0.00/24 2.3s} 73. Ke3 {+0.03/24 5.0s} Ke7 {+0.03/18 1.3s}
74. Ra4 {+0.04/21 0.67s} Kf6 {+0.04/19 3.2s} 75. Ra6 {+0.03/23 2.5s}
Rb3+ {+0.04/22 1.2s} 76. Kf4 {+0.02/23 0.88s} Rb4+ {0.00/24 1.0s}
77. Ke3 {+0.02/21 0.26s} Ke7 {0.00/24 1.0s} 78. Ra8 {+0.02/20 0.78s}
Nc4+ {+0.17/20 2.2s} 79. Kd4 {+0.02/21 0.97s} Nb6+ {+0.14/19 1.3s}
80. Kc3 {0.00/19 0.61s} Nd5+ {+0.12/21 1.6s} 81. Kc2 {+0.17/20 0.62s}
Rb7 {+0.09/18 3.1s} 82. Kd3 {+0.05/21 2.2s} Rb2 {+0.12/20 2.3s}
83. Ke4 {+0.01/19 0.84s} Nf6+ {+0.09/20 2.6s} 84. Kf3 {+0.05/20 0.41s}
Nd7 {0.00/22 1.0s} 85. Ke3 {+0.06/19 0.53s} Ne5 {+0.17/21 2.6s}
86. Ke4 {+0.03/21 1.0s} Re2+ {+0.14/22 2.5s} 87. Kf5 {0.00/23 2.7s}
Nf3 {0.00/21 1.4s} 88. Ra7+ {0.00/25 1.5s} Kd6 {+0.08/19 1.0s}
89. Kf6 {0.00/21 0.22s} Rxf2 {+0.41/21 1.3s} 90. Kf5 {0.00/21 0.51s}
Nd4+ {+0.48/19 2.4s} 91. Kg5 {0.00/22 0.49s} Ne6+ {+0.38/20 0.97s}
92. Kh5 {0.00/20 0.46s} Ke5 {+0.46/21 2.7s} 93. Ra5+ {-0.01/21 0.60s}
Kf6 {+0.49/20 1.2s} 94. Kg4 {-0.01/23 0.64s} Nd8 {+0.56/22 1.9s}
95. Ra3 {-0.01/23 0.97s} Kg7 {+0.57/20 0.93s} 96. Rf3 {-0.03/20 0.94s}
Ra2 {+0.70/21 2.3s} 97. Rb3 {-0.09/19 0.94s} Nf7 {+0.69/20 2.3s}
98. Rb6 {-0.03/20 0.71s} Ra4+ {+0.72/24 2.0s} 99. Kh5 {-0.05/21 1.3s}
Ra5+ {+0.72/22 1.1s} 100. Kh4 {-0.05/21 1.1s} Ne5 {+0.68/23 1.1s}
101. Rb7+ {-0.05/22 1.2s} Kh6 {+0.68/24 1.6s} 102. Rb6+ {-0.07/24 0.78s}
Kh7 {+0.65/22 1.3s} 103. Rf6 {-0.07/23 0.68s} Kg7 {+0.65/21 1.1s}
104. Rb6 {-0.07/21 0.71s} Nf3+ {+0.67/24 2.1s} 105. Kg4 {-0.06/23 1.1s}
Ng5 {+0.67/21 0.84s} 106. Rb7+ {-0.06/24 0.88s} Nf7 {+0.62/23 0.85s}
107. Kh4 {-0.07/21 1.0s} Ra4+ {+0.62/24 1.1s} 108. Kh5 {-0.06/22 1.1s}
Ra6 {+0.62/25 1.4s} 109. Rb5 {-0.08/22 2.8s} Rh6+ {-M124/1 1.2s}
110. Kg4 {-0.07/21 0.23s} Rf6 {+0.62/22 1.1s} 111. Ra5 {-0.04/23 2.1s}
Rg6+ {+0.62/22 1.7s} 112. Kh4 {-0.02/24 0.52s} Rc6 {+0.62/24 1.1s}
113. Kh5 {-0.02/23 0.56s} Rh6+ {+0.61/22 1.4s} 114. Kg4 {-0.02/21 0.24s}
Rb6 {+0.61/22 1.5s} 115. Kh5 {-0.02/24 1.3s} Re6 {+0.62/23 0.81s}
116. Ra4 {-0.02/23 1.1s} Rh6+ {+0.62/25 1.6s} 117. Kg4 {-0.02/22 0.25s}
Rb6 {+0.62/23 1.5s} 118. Kh5 {-0.02/24 1.0s} Rb5+ {+0.66/23 1.0s}
119. Kh4 {-0.02/24 0.88s} Rd5 {+0.62/26 1.9s} 120. Ra7 {-0.02/25 5.2s}
Rd6 {+0.62/25 1.6s} 121. Kh5 {-0.02/23 0.42s} Rh6+ {+0.62/26 1.6s}
122. Kg4 {-0.02/23 0.23s} Rb6 {+0.61/24 0.69s} 123. Kh5 {-0.02/24 0.77s}
Rb3 {+0.62/26 1.5s} 124. Kh4 {-0.02/23 0.47s} Re3 {+0.63/23 1.1s}
125. Kg4 {-0.02/22 0.92s} Rc3 {+0.62/23 0.99s} 126. Ra5 {-0.02/24 0.89s}
Rc1 {+0.62/24 1.5s} 127. Ra7 {-0.02/23 1.5s} Rc4+ {+0.62/22 1.4s}
128. Kh5 {-0.03/23 1.2s} Rc5+ {+0.62/26 1.3s} 129. Kh4 {-0.03/24 0.60s}
Rc3 {+0.63/24 1.3s} 130. Ra6 {-0.03/22 0.60s} Rc4+ {+0.62/24 1.8s}
131. Kh5 {-0.01/21 0.48s} Rc5+ {+0.62/25 1.6s} 132. Kh4 {-0.01/24 0.95s}
Rc3 {+0.62/24 2.0s} 133. Kg4 {-0.01/25 0.95s} Rd3 {+0.39/20 1.5s}
134. Kf4 {-0.01/23 0.70s} Rd4+ {0.00/21 1.9s} 135. Ke3 {-0.01/23 0.41s}
Rg4 {0.00/22 0.69s} 136. Kf2 {-0.01/23 0.55s} Ng5 {0.00/23 0.69s}
137. Kg2 {-0.01/23 1.1s} Ne4 {0.00/25 0.90s} 138. Kf3 {0.00/26 2.2s}
Nf6 {0.00/27 0.82s} 139. Rxf6 {+0.01/28 0.95s, Draw by adjudication: SyzygyTB}
1/2-1/2

Thanks for reporting those. Typical for first time public releases.

Dio wrote:

I have played 1400 games with Rebel 16 so far.
I use Cutechess (GUI), 8 games are played simultaneously.

In 3 games (against different opponents) it happened that Rebel 16.0 stopped calculating already on search depth 1,
mate values and normal evaluations were displayed alternately.
I then had to restart the GUI to continue the match.

Have other testers had the same experience ?

I think it’s me trying to be too clever at reporting mates. Am looking into it.

All reports are -M124, which will make trapping it easier.
-124 is outside maximum depth, so theoretically not possible, which gives me a good clue what it is (bounds error in hash table where depth is held in 8 bit, and depth gets used to adjust mate scores returned from the hash).

Shall I break from doing the pondering code to fix this?

Three things
-M124
Sudden death still too slow
Add ponder

pohl4711 wrote:

I had no freezes, but something is definitly wrong with Rebel. Look at move 109 by Rebel 16 (black)...
Drawn position, but a thinking depth of 1 after 1.2 seconds and a bogus evaluation: Rh6+ {-M124/1 1.2s}
(This happens again and again in my testrun) (I use cutechess-cli)

(And Rebel 16a moves still too fast. Now, 7min 5secs per game in my testings (better than Rebel 16 (6min 50sec))
Normal average value in my ratinglist is 7min 45secs

So,  I aborted the Rebel testruns. Makes no sense right now. Engine is buggy. And time management is still bad.

[Event "CPU-engine testing"]
[Site "?"]
[Date "2022.11.30"]
[Round "6"]
[White "Berserk 10 avx2"]
[Black "Rebel 16 avx2"]
[Result "1/2-1/2"]
[ECO "C24"]
[Opening "Bishop's Opening"]
[PlyCount "277"]
[Termination "adjudication"]
[TimeControl "180+1"]
[Variation "Urusov gambit"]

1. e4 {book} e5 {book} 2. Bc4 {book} Nf6 {book} 3. d4 {book} exd4 {book}
4. Nf3 {book} Nxe4 {book} 5. Qxd4 {book} Nf6 {book} 6. Nc3 {-0.64/25 15s}
c6 {+0.47/20 8.0s} 7. Bg5 {-0.50/20 2.2s} d5 {+0.85/21 6.9s}
8. O-O-O {-0.56/20 3.3s} Be6 {+0.47/22 4.0s} 9. Rhe1 {-0.24/24 7.7s}
Be7 {+0.31/20 2.7s} 10. Qh4 {-0.32/23 4.5s} h6 {+0.32/20 7.2s}
11. Bd3 {-0.01/27 13s} Rg8 {0.00/21 7.1s} 12. Rxe6 {+0.01/27 15s}
fxe6 {-0.03/19 3.6s} 13. Bg6+ {+0.03/25 4.8s} Kd7 {-0.56/21 2.5s}
14. Bf4 {-0.01/23 5.6s} Kc8 {-0.28/19 4.6s} 15. Nd4 {+0.01/28 6.8s}
Nfd7 {-0.85/22 4.5s} 16. Qh3 {+0.01/29 4.7s} Bg5 {0.00/20 2.3s}
17. Nxe6 {0.00/28 1.5s} Bxf4+ {0.00/21 3.0s} 18. Nxf4 {-0.01/25 1.6s}
Qg5 {0.00/23 3.9s} 19. g3 {0.00/27 1.5s} Rd8 {0.00/23 2.9s}
20. Rd2 {+0.01/27 4.6s} Qf6 {0.00/21 2.9s} 21. Ncxd5 {0.00/27 7.8s}
cxd5 {-0.82/19 2.2s} 22. Nxd5 {-0.01/26 1.4s} Qf8 {0.00/23 2.2s}
23. Bf5 {-0.02/27 1.6s} Nc6 {0.00/23 4.0s} 24. Bxd7+ {+0.09/24 3.1s}
Rxd7 {0.00/24 1.9s} 25. Nf6 {+0.15/26 4.8s} gxf6 {0.00/25 1.9s}
26. Rxd7 {+0.17/24 1.7s} Kb8 {0.00/26 4.6s} 27. Qe6 {+0.15/24 2.4s}
a5 {+0.37/20 2.0s} 28. Qd6+ {+0.01/27 3.5s} Qxd6 {+0.55/21 2.0s}
29. Rxd6 {+0.02/20 1.3s} Kc7 {+0.40/20 2.0s} 30. Rxf6 {+0.11/22 2.2s}
Rh8 {+0.55/21 2.6s} 31. h4 {+0.01/22 2.3s} b5 {0.00/20 5.4s}
32. h5 {+0.39/20 2.1s} a4 {+0.16/22 1.9s} 33. a3 {+0.29/26 7.8s}
Ne5 {0.00/23 3.2s} 34. Kd2 {+0.35/23 2.5s} Ng4 {0.00/25 4.1s}
35. Rf5 {+0.29/25 2.6s} Kb6 {0.00/22 4.5s} 36. Rf4 {+0.12/22 2.1s}
Rg8 {0.00/23 3.3s} 37. Ke2 {+0.46/24 3.2s} Ne5 {0.00/21 4.3s}
38. b3 {+0.47/24 2.1s} Ra8 {0.00/23 2.2s} 39. Rf6+ {+0.39/24 2.4s}
Kb7 {0.00/23 2.4s} 40. bxa4 {+0.62/21 1.2s} bxa4 {0.00/24 1.5s}
41. Rxh6 {+0.69/23 2.8s} Nc4 {0.00/25 1.5s} 42. Kd3 {+0.47/26 6.7s}
Nxa3 {0.00/20 1.6s} 43. Re6 {+0.44/23 3.3s} Nb5 {0.00/22 1.6s}
44. Re1 {+0.46/22 1.4s} Rd8+ {-0.07/20 4.4s} 45. Kc4 {+0.46/22 1.1s}
Kc6 {0.00/23 1.6s} 46. Kb4 {+0.33/22 2.3s} Nd4 {-0.16/21 1.5s}
47. Rc1 {+0.50/22 2.6s} Rb8+ {-0.31/21 1.5s} 48. Kc4 {+0.46/21 1.3s}
Nb5 {-0.13/22 1.5s} 49. h6 {+0.46/23 3.6s} Nd6+ {-0.23/19 1.7s}
50. Kd3 {+0.38/23 1.3s} Rh8 {-0.15/21 1.5s} 51. Rh1 {+0.36/23 2.2s}
a3 {-0.15/21 1.5s} 52. h7 {+0.35/24 8.7s} a2 {0.00/20 1.5s}
53. Ra1 {+0.22/25 2.4s} Rxh7 {0.00/19 1.5s} 54. Rxa2 {+0.20/23 1.2s}
Rh8 {-0.16/20 1.5s} 55. Ra6+ {+0.33/23 3.6s} Kc7 {-0.08/17 2.1s}
56. Ke3 {+0.35/23 4.5s} Re8+ {-0.03/21 1.9s} 57. Kf3 {+0.31/21 0.99s}
Rc8 {0.00/22 2.1s} 58. c4 {+0.33/21 1.4s} Kd7 {0.00/22 1.4s}
59. Ra2 {+0.14/22 6.9s} Rxc4 {0.00/20 2.1s} 60. Re2 {+0.14/22 1.8s}
Kc6 {+0.12/20 1.4s} 61. Rd2 {+0.13/23 3.0s} Ne4 {+0.14/21 3.5s}
62. Rd3 {+0.15/23 1.6s} Ra4 {+0.18/21 2.0s} 63. Ke3 {+0.13/24 1.8s}
Nc5 {+0.14/21 1.4s} 64. Rd2 {+0.13/24 1.5s} Re4+ {+0.09/23 3.0s}
65. Kf3 {+0.13/22 0.24s} Rc4 {+0.07/20 1.3s} 66. Ke3 {+0.12/24 5.0s}
Rb4 {+0.02/22 1.7s} 67. Ra2 {+0.12/21 2.3s} Kd5 {+0.09/22 3.2s}
68. Kf3 {+0.09/23 3.4s} Rb3+ {+0.08/22 2.4s} 69. Kg4 {+0.08/21 1.2s}
Ne4 {0.00/23 2.8s} 70. Kf4 {+0.07/21 0.80s} Nd6 {0.00/20 1.8s}
71. Ra5+ {+0.08/22 2.8s} Ke6 {+0.11/19 1.3s} 72. Ra6 {+0.07/23 6.0s}
Rb2 {0.00/24 2.3s} 73. Ke3 {+0.03/24 5.0s} Ke7 {+0.03/18 1.3s}
74. Ra4 {+0.04/21 0.67s} Kf6 {+0.04/19 3.2s} 75. Ra6 {+0.03/23 2.5s}
Rb3+ {+0.04/22 1.2s} 76. Kf4 {+0.02/23 0.88s} Rb4+ {0.00/24 1.0s}
77. Ke3 {+0.02/21 0.26s} Ke7 {0.00/24 1.0s} 78. Ra8 {+0.02/20 0.78s}
Nc4+ {+0.17/20 2.2s} 79. Kd4 {+0.02/21 0.97s} Nb6+ {+0.14/19 1.3s}
80. Kc3 {0.00/19 0.61s} Nd5+ {+0.12/21 1.6s} 81. Kc2 {+0.17/20 0.62s}
Rb7 {+0.09/18 3.1s} 82. Kd3 {+0.05/21 2.2s} Rb2 {+0.12/20 2.3s}
83. Ke4 {+0.01/19 0.84s} Nf6+ {+0.09/20 2.6s} 84. Kf3 {+0.05/20 0.41s}
Nd7 {0.00/22 1.0s} 85. Ke3 {+0.06/19 0.53s} Ne5 {+0.17/21 2.6s}
86. Ke4 {+0.03/21 1.0s} Re2+ {+0.14/22 2.5s} 87. Kf5 {0.00/23 2.7s}
Nf3 {0.00/21 1.4s} 88. Ra7+ {0.00/25 1.5s} Kd6 {+0.08/19 1.0s}
89. Kf6 {0.00/21 0.22s} Rxf2 {+0.41/21 1.3s} 90. Kf5 {0.00/21 0.51s}
Nd4+ {+0.48/19 2.4s} 91. Kg5 {0.00/22 0.49s} Ne6+ {+0.38/20 0.97s}
92. Kh5 {0.00/20 0.46s} Ke5 {+0.46/21 2.7s} 93. Ra5+ {-0.01/21 0.60s}
Kf6 {+0.49/20 1.2s} 94. Kg4 {-0.01/23 0.64s} Nd8 {+0.56/22 1.9s}
95. Ra3 {-0.01/23 0.97s} Kg7 {+0.57/20 0.93s} 96. Rf3 {-0.03/20 0.94s}
Ra2 {+0.70/21 2.3s} 97. Rb3 {-0.09/19 0.94s} Nf7 {+0.69/20 2.3s}
98. Rb6 {-0.03/20 0.71s} Ra4+ {+0.72/24 2.0s} 99. Kh5 {-0.05/21 1.3s}
Ra5+ {+0.72/22 1.1s} 100. Kh4 {-0.05/21 1.1s} Ne5 {+0.68/23 1.1s}
101. Rb7+ {-0.05/22 1.2s} Kh6 {+0.68/24 1.6s} 102. Rb6+ {-0.07/24 0.78s}
Kh7 {+0.65/22 1.3s} 103. Rf6 {-0.07/23 0.68s} Kg7 {+0.65/21 1.1s}
104. Rb6 {-0.07/21 0.71s} Nf3+ {+0.67/24 2.1s} 105. Kg4 {-0.06/23 1.1s}
Ng5 {+0.67/21 0.84s} 106. Rb7+ {-0.06/24 0.88s} Nf7 {+0.62/23 0.85s}
107. Kh4 {-0.07/21 1.0s} Ra4+ {+0.62/24 1.1s} 108. Kh5 {-0.06/22 1.1s}
Ra6 {+0.62/25 1.4s} 109. Rb5 {-0.08/22 2.8s} Rh6+ {-M124/1 1.2s}
110. Kg4 {-0.07/21 0.23s} Rf6 {+0.62/22 1.1s} 111. Ra5 {-0.04/23 2.1s}
Rg6+ {+0.62/22 1.7s} 112. Kh4 {-0.02/24 0.52s} Rc6 {+0.62/24 1.1s}
113. Kh5 {-0.02/23 0.56s} Rh6+ {+0.61/22 1.4s} 114. Kg4 {-0.02/21 0.24s}
Rb6 {+0.61/22 1.5s} 115. Kh5 {-0.02/24 1.3s} Re6 {+0.62/23 0.81s}
116. Ra4 {-0.02/23 1.1s} Rh6+ {+0.62/25 1.6s} 117. Kg4 {-0.02/22 0.25s}
Rb6 {+0.62/23 1.5s} 118. Kh5 {-0.02/24 1.0s} Rb5+ {+0.66/23 1.0s}
119. Kh4 {-0.02/24 0.88s} Rd5 {+0.62/26 1.9s} 120. Ra7 {-0.02/25 5.2s}
Rd6 {+0.62/25 1.6s} 121. Kh5 {-0.02/23 0.42s} Rh6+ {+0.62/26 1.6s}
122. Kg4 {-0.02/23 0.23s} Rb6 {+0.61/24 0.69s} 123. Kh5 {-0.02/24 0.77s}
Rb3 {+0.62/26 1.5s} 124. Kh4 {-0.02/23 0.47s} Re3 {+0.63/23 1.1s}
125. Kg4 {-0.02/22 0.92s} Rc3 {+0.62/23 0.99s} 126. Ra5 {-0.02/24 0.89s}
Rc1 {+0.62/24 1.5s} 127. Ra7 {-0.02/23 1.5s} Rc4+ {+0.62/22 1.4s}
128. Kh5 {-0.03/23 1.2s} Rc5+ {+0.62/26 1.3s} 129. Kh4 {-0.03/24 0.60s}
Rc3 {+0.63/24 1.3s} 130. Ra6 {-0.03/22 0.60s} Rc4+ {+0.62/24 1.8s}
131. Kh5 {-0.01/21 0.48s} Rc5+ {+0.62/25 1.6s} 132. Kh4 {-0.01/24 0.95s}
Rc3 {+0.62/24 2.0s} 133. Kg4 {-0.01/25 0.95s} Rd3 {+0.39/20 1.5s}
134. Kf4 {-0.01/23 0.70s} Rd4+ {0.00/21 1.9s} 135. Ke3 {-0.01/23 0.41s}
Rg4 {0.00/22 0.69s} 136. Kf2 {-0.01/23 0.55s} Ng5 {0.00/23 0.69s}
137. Kg2 {-0.01/23 1.1s} Ne4 {0.00/25 0.90s} 138. Kf3 {0.00/26 2.2s}
Nf6 {0.00/27 0.82s} 139. Rxf6 {+0.01/28 0.95s, Draw by adjudication: SyzygyTB}
1/2-1/2

under ARENA there are no problems, also with me was alternately displayed the mate value -124.00
and then a balanced evaluation. The engine then hangs on depth 1. The GUI freezes briefly.

After about 3 to 5 minutes the game continues as if nothing had happened.
Sometimes both clocks run at the same time.

I don't see Ponder as very important at the moment.

This bug does not seem to have any influence on the game strength.
Since the bug occurs rarely, it is not easy to take a screenshot.

Translated with www.DeepL.com/Translator (free version)

Dio wrote:

under ARENA there are no problems, also with me was alternately displayed the mate value -124.00
and then a balanced evaluation. The engine then hangs on depth 1. The GUI freezes briefly.

After about 3 to 5 minutes the game continues as if nothing had happened.
Sometimes both clocks run at the same time.

I don't see Ponder as very important at the moment.

This bug does not seem to have any influence on the game strength.
Since the bug occurs rarely, it is not easy to take a screenshot.

Translated with www.DeepL.com/Translator (free version)

Don’t worry about a screen shot, -M124 is enough to track it down

+77 Elo compared with Rebel 15.1a for Rebel 16.0 in our 40/4-tests after 1900 games,
I did two additionell (unofficial) tests with 16a to compare with 16.0
https://cegt.forumieren.com/t1828-testing-rebel-16-0nn-x64-1cpu

Jörg thank you.

Note the time fix is only for sudden death levels.

I know, I am using 4 Minutes +2 Seconds for our 40/4-Tests

Aha [!]

That will make a difference....

Then (but up to you of course) it's better to wait for an official update.

This is what I get with 16a with 10+0.1

Code:

Engine                 Depth    Average
Seer_2.6.0             19.08      0.22  
RubiChess_20220813     21.23      0.22  
Ethereal_13.75-NNUE    19.78      0.22  
Koivisto_8.0           19.16      0.21  
Revenge_3.0            20.70      0.21  
SlowChess_2.9          16.02      0.20  
Berserk_10             18.41      0.20  
rebel-16a              18.74      0.19  

0.19 (average time use) is already a lot better than 16, but needs a small extra push.

Time controls such as 40 moves in 4 minutes and repeating date from a time when chess computers were tested (they had only this timing).

Even Stockfish 15.0 has considerable problems with this old time control (Andreas Matthies, Rubichess), i.e. Stockfish very often gets into time trouble and has to answer even after very short thinking time). Another candidate is LCZero...