Eden on the 40 positions

[nczempin]

The first match in the 40-positions-test between Alchess and Eden is finished. The result is 67-13. I'll be very curious to see how that'll change in the next match. And of course the two matches after that.

Okay, here are my current results (Eden 0.0.13 basic Java vs ..., results are from opponent's point of view):

ALChess 1.5b:
67.0-13.0
62.5-17.5

Piranha 0.5:
44.5-35.5
44.5-35.5

Philemon C:
50.5-29.5
54.5-25.5

Apilchess:
34.0-46.0
[second match will start tonight]

After Apilchess, I want to play against a umax version between 4.8 and 1.6, that hgm (hello? ) would estimate to be closest to Eden 0.0.13, yet stronger.

[pedrox]

Results for DanaSah 3.13

Nejmet 3.07:
---------------
Danasah313xp 34,5/60 =1==0=11100=01000=1=01=11101011==10==111000=1==11=0=01111011
Danasah313xp 26,5/60 1001000=010011==0==111000==0=01001=10=011=010=11=00=00==0=11

Dragon 4.6:
-------------
Danasah313xp 33,5/60 11110101=1010000011=11001110=00=1=0111=1110=101=1=000==10==1
Danasah313xp 36,0/60 =1=0011=111=10==0==11=01011=011000=11101=1010=0=0111110=1=11

Pepito 1.59:
--------------
Danasah313xp 21,5/60 1101=====0000100000100=1===00==0=0010100100=1=1=0000=01===00
Danasah313xp 26,0/60 0=1101001=000==0100110101100===00=0=0101000==11=00=0=1110101

Anmon 5.50:
---------------
Danasah313xp 21,5/60 10011==0==00000=1=00=11001000110==00=00=00=100100110==0=0100
Danasah313xp 25,5/60 0110111000=001=1=11====1==00=110100=0100000100==000===010=01

Against Nejmet there is a difference of 8 points of 60 between the first test and the second. I believe that with these results we must give much reason to Bob. Although all we cannot have a cluster.

Games 1' + 1'' I believe that with a time control the 40/4 results are less extreme.

Pedro

[bob]

I don't see the variability drop that much as the time controls stretch out. I've run some 60+60 matches that were just as variable...

It does make it difficult to measure modest improvements when you play modest numbers of games...