Testing of current Romi beta (about 3 weeks old)

[Tony Thomas]

I finished testing the beta but I do not know if it is stronger or weaker than the previous version. Though the newer version scored lower the results are within the error margin which suggests that there might not be much of a strength difference between both versions. One thing worth mentioning is the fact that the new Romi performed as good or better than Chesstiger, Naum, Zappa, Arasan and Francesca but in the mean time the newer version performed worse against almost all of the weaker opponents.

Code: Select all

7 RomiChessDK6              : 2446  420 (+188,= 77,-155), 53.9 %

Danasah 2.85                  :  30 (+ 20,=  5,-  5), 75.0 %
Phalanx Reborn                :  30 (+ 14,=  3,- 13), 51.7 %
Delphil 1.6c                  :  30 (+ 13,=  9,-  8), 58.3 %
Francesca MAD 0.13            :  30 (+  7,= 10,- 13), 40.0 %
Lime 6.3                      :  30 (+ 25,=  3,-  2), 88.3 %
Djinn 0.925x                  :  30 (+ 11,=  8,- 11), 50.0 %
Zeus 1.28                     :  30 (+ 18,=  4,-  8), 66.7 %
Zappa 1.1                     :  30 (+  9,=  4,- 17), 36.7 %
GreKo 5.2                     :  30 (+ 25,=  4,-  1), 90.0 %
NanoSzachy 2.7                :  30 (+ 19,=  4,-  7), 70.0 %
Arasan 9.5                    :  30 (+ 12,=  9,-  9), 55.0 %
Horizon_4_3_173               :  30 (+ 13,=  6,- 11), 53.3 %
Naum 2.0                      :  30 (+  1,=  4,- 25), 10.0 %
ChessTiger2007UCI             :  30 (+  1,=  4,- 25), 10.0 %

Code: Select all

9 RomiBetaNG1               : 2439  420 (+185,= 75,-160), 53.0 %

Danasah 2.85                  :  30 (+ 15,=  4,- 11), 56.7 %
Phalanx Reborn                :  30 (+ 11,=  5,- 14), 45.0 %
Delphil 1.6c                  :  30 (+ 15,=  7,-  8), 61.7 %
Francesca MAD 0.13            :  30 (+ 11,=  6,- 13), 46.7 %
Lime 6.3                      :  30 (+ 22,=  4,-  4), 80.0 %
Djinn 0.925x                  :  30 (+ 12,=  4,- 14), 46.7 %
Zeus 1.28                     :  30 (+ 21,=  5,-  4), 78.3 %
Zappa 1.1                     :  30 (+ 12,=  4,- 14), 46.7 %
GreKo 5.2                     :  30 (+ 23,=  5,-  2), 85.0 %
NanoSzachy 2.7                :  30 (+ 15,=  4,- 11), 56.7 %
Arasan 9.5                    :  30 (+ 13,=  7,- 10), 55.0 %
Horizon_4_3_173               :  30 (+ 11,= 10,-  9), 53.3 %
Naum 2.0                      :  30 (+  4,=  3,- 23), 18.3 %
ChessTiger2007UCI             :  30 (+  0,=  7,- 23), 11.7 %

[Michael Sherwin]

Hi Tony,

just saw this post. Thanks for the test! It is just totally bafflelng how Romi can score as well or better than before against the 4 top engines and then do worse vs the weaker ones. if you would have run Zappa and Fran Mad first then it would have been a huge let down as the test went on.

[Rubinus]

Zappa 1.1
What games lost bug?

[Tony Thomas]

Hi Tony,

just saw this post. Thanks for the test! It is just totally bafflelng how Romi can score as well or better than before against the 4 top engines and then do worse vs the weaker ones. if you would have run Zappa and Fran Mad first then it would have been a huge let down as the test went on.

Good thing I waited until the last minute to test the stronger engines. Pavel, I dont know if I understand your post correctly, but Zappa didnt lose any games due to bugs, they were true Romi wins.