Test Tourney 369

[vigo64]

Test Tourney 369
26 engines
Dual Opteron 244, WinXP x64
4RR, time control 3' + 1"
hash table ~128MB, ponder off
all 3, 4 and 5 men TBs, cache 32MB
SMP engines use 2 CPU's
GUI: Winboard 4.5.0
all .debug are saved

Code: Select all

 1  Houdini 2.0 x64              2934  98.0/100
 2  Houdini 1.05a x64            2882  97.5/100
 3  RedQueen 1.0.0               2558  85.5/100
 4  RedQueen 0.9.8 JA            2613  85.0/100
 5  iCE 0.2 build 1085           2447  73.5/100
 6  Alex v2.14_x64               2381  67.5/100
 7  Alex v2.14a_win32            2410  66.5/100
 8  Ifrit_m1_2_Beta_23_9_2011    2417  65.0/100 
 9  Betsabe II 1.08 JA           2371  61.5/100
10  FireFly v2.5.13 (x64)        2353  59.0/100
11  Ifrit_j4_4_Beta_9_9_2011_JA  2398  58.0/100
12  Betsabe II 1.08 C JA         2371  57.5/100
13  Obender 3.2.x.s.7.5a JA      2351  57.0/100
14  FireFly v2.5.12 (x64)        2304  55.5/100
15  ChessKISS 1.3                2213  44.5/100
16  ChessKISS 1.3a               2270  43.5/100
17  OBender3.10.02a JA           2228  41.0/100
18  iCE 0.1 build 1120           2262  40.5/100
19  Arabian Knight v1.47         2218  39.0/100
20  Arabian Knight v1.40         2224  34.5/100
21  Ziggurat v0.18               2162  28.0/100
22  Neurone X                    1964  14.0/100
23  Ziggurat v0.16               2052  13.5/100
24  Neurone IX                   1882   9.0/100
25  Evolution Chess 0.6.15.dev   1822   5.0/100
26  Evolution Chess 0.6.13.dev   1678   0.5/100

To download the games please visit my site:

http://vigo.altervista.org --> Test_Tourney.htm

Greetings,
Ciro

[Kurt]

In my opinion, tourneys including different versions of the same program, do not make much sense.
Kurt

[tpetzke]

In my opinion, tourneys including different versions of the same program, do not make much sense.

why, for me as a developer I appreciate those testings. I get valuable and independent information out of it. Programers are not good testers of their own engine.

I suspected iCE 0.2 to be stronger than 0.1, because in my limited tests it scored much better, but I'm not a experienced tester and have only limited testing resources. This test tourney proves it is stronger. So for me it does make sense.

Thomas...

[Andres Valverde]

In my opinion, tourneys including different versions of the same program, do not make much sense.

why, for me as a developer I appreciate those testings. I get valuable and independent information out of it. Programers are not good testers of their own engine.

I suspected iCE 0.2 to be stronger than 0.1, because in my limited tests it scored much better, but I'm not a experienced tester and have only limited testing resources. This test tourney proves it is stronger. So for me it does make sense.

Thomas...

+1