ozziejoe wrote:Pro Deo will recognize games it has lost and when faced a second (third, fourth etc.) time it will adapt its playing style until it has found new and good settings for the type of opening it is playing, or is forced to play.
Well, as you can see below, ECO learning is on, but i see virtually no evidence of learning files being changed (the learn file is still only 1kb).
Maybe learning only occurs within the eco database. Thus, the moment you have a move outside of that database, no learning?
Anyway, here are the paremeters. I'll let prodeo run alittle longer, and see if it can break some of its relentless losing to everybody. I am just looking for some vague sign of learning. Otherwise, I'm going back to romi (which I know utilizes alot of info for its learning).
Pro Deo 1.5, executable made on 06-06-2007
Personality Path personal/PRODEO.ENG (0)
Personality description Pro Deo 1.5 (June 2007)
ozziejoe wrote:I am just looking for some vague sign of learning. Otherwise, I'm going back to romi (which I know utilizes alot of info for its learning).
Go Romi!
She ain't just kangaroo meat you know!
If you are on a sidewalk and the covid goes beep beep
Just step aside or you might have a bit of heat
Covid covid runs through the town all day
Can the people ever change their ways
Sherwin the covid's after you
Sherwin if it catches you you're through
ozziejoe wrote:I'll let prodeo run alittle longer, and see if it can break some of its relentless losing to everybody. I am just looking for some vague sign of learning.
Have you seen any sign of non learning? As if ProDeo repeats the same lose twice? If not, then probably you're just playing it against too strong opponents.
Anyway, Romi's learning behavior seems more aggressive, as an additional plus is that it should be getting more time on her games (if she moves instantly) it's like if she used the time of her opponents on her future games.
ProDeo's learning is more different than it seemed at first sight.
Your beliefs create your reality, so be careful what you wish for.
No file seems to be increasing in size to the extent i would expect if you were remebering lines and changing weights. I don't think shredder 10 should be out of this world for prodeo. With a non-retarded learning function, prodeo should get an occasional draw. It just gets slaughtered. I suspect I am not usiing it correctly. Anyway, I have moved to another method. See general forum
Romi definitely has something to teach other engines in terms of learning. You are right, the better the oponent plays (and the longer time it takes), the more romi benefits. this is a great feature. I reckon if you let romi learn one position for a long time, romi can score 100 to 200 elo points above her baseline strength in that position.
so, in conclusion: Romi so sexy. Me love her long time
