Nunn Top2 results

[Heinz van Kempen]

Geschrieben von:/Posted by: Heinz van Kempen at 17 May 2004 06:22:47:

Hi ,
Nunn Blitz Top2 tournament is in progress now with a lot of new interesting releases and some betas (thanks to all authors ). I plan to finish the tournament in four weeks on two equally fast computers, what will of course slow down the Active Chess Tournaments, because there are 380 games here with each position, giving 760 games to each version in the end.
Conditions:
Athlon 2600+
64 MB Hash
5 men EGTB
all Nunn2 positions
Time control: 4 min. + 2 sec.


Position 1
1   Aristarch 4.50            ** ½0 11 11 ½½ 11 10 01 11 11 11 11 11 1½ 10 ½1 11 10 ½1 11  30.0/38
2   Crafty 19.13 DC           ½1 ** ½½ 10 01 10 ½1 11 ½½ 1½ 10 11 10 01 11 11 11 1½ ½½ ½½  26.0/38
3   Thinker 4.6b              00 ½½ ** 1½ 01 ½1 11 01 01 0½ 10 1½ 0½ 1½ ½1 ½1 11 11 11 11  25.0/38
4   Patriot 0.172-Light       00 01 0½ ** 00 01 ½1 10 ½1 00 01 10 11 11 11 11 10 1½ 11 11  23.0/38
5   Gothmog 0.4.8f            ½½ 10 10 11 ** 00 00 01 01 01 10 ½0 11 11 0½ 10 1½ 11 11 11  22.5/38  
6   Little Goliath 2000 v3.9  00 01 ½0 10 11 ** 10 ½0 ½½ ½0 ½1 11 1½ 00 ½1 11 10 11 11 1½  22.5/38  
7   Terra 3.3B11              01 ½0 00 ½0 11 01 ** 11 ½0 01 11 11 10 11 0½ 10 00 01 11 11  22.0/38
8   King of Kings 2.56        10 00 10 01 10 ½1 00 ** 11 10 11 00 01 0½ 1½ 11 ½1 ½0 11 ½1  21.0/38  
9   Crafty 19.12 SE Capa      00 ½½ 10 ½0 10 ½½ ½1 00 ** 11 11 ½0 10 1½ ½0 ½½ 11 10 1½ 11  21.0/38  
10  Slow Chess 2.93           00 0½ 1½ 11 10 ½1 10 01 00 ** ½0 01 00 11 11 ½0 1½ 1½ ½1 10  20.0/38
11  LambChop 10.99            00 01 01 10 01 ½0 00 00 00 ½1 ** 10 1½ ½0 11 11 ½1 10 0½ 10  17.0/38  
12  Jonny 2.63b               00 00 0½ 01 ½1 00 00 11 ½1 10 01 ** ½1 01 10 10 01 10 00 11  17.0/38  
13  The Baron 1.3.1b2         00 01 1½ 00 00 0½ 01 10 01 11 0½ ½0 ** 0½ ½0 1½ 10 1½ 10 1½  16.5/38  
14  PostModernist 1010a       0½ 10 0½ 00 00 11 00 1½ 0½ 00 ½1 10 1½ ** 0½ 11 01 10 ½½ 10  16.5/38  
15  Amy 0.8.7b                01 00 ½0 00 1½ ½0 1½ 0½ ½1 00 00 01 ½1 1½ ** 00 1½ ½½ 1½ 10  16.0/38
16  Naum 1.0                  ½0 00 ½0 00 01 00 01 00 ½½ ½1 00 01 0½ 00 11 ** 11 ½½ ½½ 11  15.0/38
17  Fruit X 05/09             00 00 00 01 0½ 01 11 ½0 00 0½ ½0 10 01 10 0½ 00 ** ½1 11 1½  14.5/38
18  Movei 00_8_198            01 0½ 00 0½ 00 00 10 ½1 01 0½ 01 01 0½ 01 ½½ ½½ ½0 ** 10 01  14.0/38
19  Booot 3.3                 ½0 ½½ 00 00 00 00 00 00 0½ ½0 1½ 11 01 ½½ 0½ ½½ 00 01 ** 0½  11.0/38
20  Frenzee 157               00 ½½ 00 00 00 0½ 00 ½0 00 01 01 00 0½ 01 01 00 0½ 10 1½ **   9.5/38
Position 2
1   Little Goliath 2000 v3.9  ** ½½ 10 01 1½ 01 ½0 11 ½½ 11 10 1½ 1½ 11 11 1½ 01 ½½ 11 11  26.5/38
2   Aristarch 4.50            ½½ ** ½0 ½½ 1½ ½1 1½ 01 01 1½ 10 01 1½ 11 10 1½ 11 ½½ 11 11  25.5/38  
3   Crafty 19.13 DC           01 ½1 ** ½½ 0½ ½0 0½ 11 ½1 ½1 11 11 10 10 ½1 01 ½1 11 11 1½  25.5/38  
4   Thinker 4.6b              10 ½½ ½½ ** ½½ 1½ 01 10 ½1 ½½ ½1 1½ 01 1½ 01 01 11 1½ 11 11  25.0/38
5   Crafty 19.12 SE Capa      0½ 0½ 1½ ½½ ** 10 0½ 1½ ½1 11 10 11 0½ ½1 1½ 11 1½ 10 ½1 11  24.5/38  
6   Patriot 0.172-Light       10 ½0 ½1 0½ 01 ** 0½ 10 01 10 11 01 11 11 11 1½ ½1 ½1 11 01  24.5/38  
7   Gothmog 0.4.8f            ½1 0½ 1½ 10 1½ 1½ ** 00 10 10 01 ½½ 11 ½½ 11 00 01 11 10 10  21.5/38
8   LambChop 10.99            00 10 00 01 0½ 01 11 ** 11 0½ ½1 01 00 ½½ 11 11 11 01 ½0 ½1  20.5/38
9   Slow Chess 2.93           ½½ 10 ½0 ½0 ½0 10 01 00 ** 1½ ½1 01 01 11 1½ ½½ 0½ ½1 11 ½0  19.5/38
10  Amy 0.8.7b                00 0½ ½0 ½½ 00 01 01 1½ 0½ ** 11 11 11 00 01 11 1½ 11 0½ 00  19.0/38
11  Naum 1.0                  01 01 00 ½0 01 00 10 ½0 ½0 00 ** 11 1½ 10 ½1 11 ½0 01 01 11  18.0/38
12  Frenzee 157               0½ 10 00 0½ 00 10 ½½ 10 10 00 00 ** 11 1½ 1½ 11 0½ 11 01 10  17.5/38
13  Movei 00_8_198            0½ 0½ 01 10 1½ 00 00 11 10 00 0½ 00 ** 01 ½½ ½1 10 11 01 ½1  17.0/38
14  King of Kings 2.56        00 00 01 0½ ½0 00 ½½ ½½ 00 11 01 0½ 10 ** 1½ ½½ 01 ½1 0½ 11  16.0/38
15  Terra 3.3B11              00 01 ½0 10 0½ 00 00 00 0½ 10 ½0 0½ ½½ 0½ ** 1½ 11 01 ½1 11  15.0/38
16  Fruit X 05/09             0½ 0½ 10 10 00 0½ 11 00 ½½ 00 00 00 ½0 ½½ 0½ ** 01 10 11 11  14.5/38
17  The Baron 1.3.1b2         10 00 ½0 00 0½ ½0 10 00 1½ 0½ ½1 1½ 01 10 00 10 ** 00 01 1½  14.0/38
18  PostModernist 1010a       ½½ ½½ 00 0½ 01 ½0 00 10 ½0 00 10 00 00 ½0 10 01 11 ** 0½ 01  12.5/38  
19  Jonny 2.63b               00 00 00 00 ½0 00 01 ½1 00 1½ 10 10 10 1½ ½0 00 10 1½ ** ½0  12.5/38  
20  Booot 3.3                 00 00 0½ 00 00 10 01 ½0 ½1 11 00 01 ½0 00 00 00 0½ 10 ½1 **  11.0/38
For those who are following my tournaments it is no surprise how different scores are concerning each position. Here for example Terra and Frenzee. For both I expect good scores and positions 3 and 4 look good so far for them and also for Amy. Naum is a strong newcomer and the new Aristarch impressing. For Booot it will be a tough tournament, but it is much improved and fighting well. Little Goliath 2000 v.3.9 should be 20-40 ELO points stronger than the Nemesis version, when I look at my old rating list. For Crafty I included two versions, one is the Dann Corbit compile, the other is the SE from Mike Byrne where I had much better results with Capablanca, Fischer and Morphy personalities compared to the native engine for Crafty 19.03 and 19.04. I just received Crafty SE 19.13 and maybe will have the time to run a gauntlet with Crafty 19.13 SE Fischer.
I assume that we have a lot of improved versions in this tournament. More I can tell after running those gauntlets in order to combine engine versions from all tournaments for better rating calculation.
To conclude this longish post I plan to do some statistics for all Nunn tournaments so far, to see which engines are doing best with open, half open and closed positions, who are the favourite opponents for Movei or the most feared opponents for The Baron. Who has the most short wins under 30 moves against engines with rating above 2500. Anyone wants to bet that it will not be Gothmog, the tactical monster? :-)
Suggestions for more statistic values are welcome, but it would be too much work to verify the engine that won the most initially balanced endgames.
And finally I want to have a "best of" download from time to time including the most interesting games with longer time controls from the "died" pyramid and Active Chess tournaments. I replayed most of those games and have some favourites, but proposals are welcome, if someone detected some fantastic games from those available for download.
 


Best Regards
Heinz
http://www.husvankempen.de/nunn/

[Alex Morozov]

Geschrieben von:/Posted by: Alex Morozov at 17 May 2004 08:27:58:
Als Antwort auf:/In reply to: Nunn Top2 results geschrieben von:/posted by: Heinz van Kempen at 17 May 2004 06:22:47:

It is very strong group for Booot. Booot is too young to die

[Kurt Utzinger]

Geschrieben von:/Posted by: Kurt Utzinger at 17 May 2004 08:34:20:
Als Antwort auf:/In reply to: Nunn Top2 results geschrieben von:/posted by: Heinz van Kempen at 17 May 2004 06:22:47:

Hi Heinz
Great work: I very much like systematical testing.
Kurt


Kurt & Rolf Chess

[Heinz van Kempen]

Geschrieben von:/Posted by: Heinz van Kempen at 17 May 2004 09:30:17:
Als Antwort auf:/In reply to: Re: Nunn Top2 results geschrieben von:/posted by: Alex Morozov at 17 May 2004 08:27:58:

It is very strong group for Booot. Booot is too young to die

Hi Alex,
yes, I knew before that it would be very tough, but also saw the big advances in the former tournaments. You can improve most when playing against strong opponents. This is valid for chessplayers and for engines also, as long as the authors are adding experiences from all they can get from different tournaments here like WBEC, RWBC, Chess War, WCCL etc..
Booot already promoted easily to Nunn Active C with version 3.2 and it will not be beaten up completely and not finish last with every position. Not falling behind totally in this tournament would mean another huge rating improvement.
Best Regards
Heinz

[Heinz van Kempen]

Geschrieben von:/Posted by: Heinz van Kempen at 17 May 2004 09:33:52:
Als Antwort auf:/In reply to: Re: Nunn Top2 results geschrieben von:/posted by: Kurt Utzinger at 17 May 2004 08:34:20:

Hi Heinz
Great work: I very much like systematical testing.
Kurt

Hi Kurt,
thanks, I am trying to combine Blitz and Active Chess as much as CPU time allows. Blitz for statistically meaningful comparisons for the fast updating and Active Chess for higher quality. On the other hand a lot of those engines already play better at Blitz than most of us on tournament level, I fear.
Best Regards
Heinz

[Uri Blass]

Geschrieben von:/Posted by: Uri Blass at 17 May 2004 10:57:28:
Als Antwort auf:/In reply to: Nunn Top2 results geschrieben von:/posted by: Heinz van Kempen at 17 May 2004 06:22:47:

Hi ,
Nunn Blitz Top2 tournament is in progress now with a lot of new interesting releases and some betas (thanks to all authors ). I plan to finish the tournament in four weeks on two equally fast computers, what will of course slow down the Active Chess Tournaments, because there are 380 games here with each position, giving 760 games to each version in the end.

Unfortunately this tournament does not give us information if there is an improvement in the engines because all the engines are new versions.
If you want to get rating for new engines then you need to include also some old engines with known rating.
Uri

[Heinz van Kempen]

Geschrieben von:/Posted by: Heinz van Kempen at 17 May 2004 11:08:52:
Als Antwort auf:/In reply to: Re: Nunn Top2 results geschrieben von:/posted by: Uri Blass at 17 May 2004 10:57:28:

Unfortunately this tournament does not give us information if there is an improvement in the engines because all the engines are new versions.
If you want to get rating for new engines then you need to include also some old engines with known rating.
Uri

Hello Uri,
this is one of the purposes of my gauntlets. Here gauntlets with at least 600 games each will be run with Crafty 19.13 Fischer, SmarThink 0.18a r139 and the uncoming Gothmog public release (hopefully). Those will get opponents from all the groups before so that I will get a rating then that is trustworthy. For example 6 opponents from Top2, 3 from Nunn Elite, 3 from Top 1 and three more stronger ones from the other tournaments finished. At least I did so before and when watching the other rating lists mine seems not to differ outside the usual margins where each list is different from others.
Best Regards
Heinz