Dr.Deeb,the Hamsters 0.2 and the Dragon....

[Tony Thomas]

I guess we shouldnt tell him that Romi is starting to kill hamsters again about 67% of the time.

Tsk... you are scoring about 13% against each hamster, nothing to be proud of really...

But, Romi is just using one hand to battle those determined little beast, one finger for each hamster!

Since, I finally saw Tony's results for NG1 and figured out what must have been the cause for the strange results, changes have been made. Ninja Girl 2 is now playing a hundred game match vs. Hamsters 0.2 and after 46 games complete the score is +23 -10 =13 in favor of Romi. The good news is (bad if you are a hamsters fan) is that Romi is very strong in the second half of a Sherwin50.pgn match. If all goes according to established trends I fully expect Romi to finally get a 70%+ score vs. Hamsters 0.2. And if indeed that is the case then we shall see about hamsters 0.3.

OBTW, Romi now only spills the blood of virtual Hamsters. She says; "it's just as much fun since the blood and guts seem so real, but when I pull the wire out of the back of my head there's no mess to clean up!"

I'll post the results when done.

Edit: Now +25 -10 =13. Romi at the half way point is usually +5, not +15. I sure hope the second half goes as expected! Retaliation expected!!

Did you do anything that would reduce the node speed of Romi? Romi used to search slightly more than a million nodes on my computer towards the middle of the game, but now she is barely doing half a million nodes. So far the node king on my computer is Pharaon, doing around 20 million nodes per second (once in a while) in a 1minute+1 second match

[Michael Sherwin]

I guess we shouldnt tell him that Romi is starting to kill hamsters again about 67% of the time.

Tsk... you are scoring about 13% against each hamster, nothing to be proud of really...

But, Romi is just using one hand to battle those determined little beast, one finger for each hamster!

Since, I finally saw Tony's results for NG1 and figured out what must have been the cause for the strange results, changes have been made. Ninja Girl 2 is now playing a hundred game match vs. Hamsters 0.2 and after 46 games complete the score is +23 -10 =13 in favor of Romi. The good news is (bad if you are a hamsters fan) is that Romi is very strong in the second half of a Sherwin50.pgn match. If all goes according to established trends I fully expect Romi to finally get a 70%+ score vs. Hamsters 0.2. And if indeed that is the case then we shall see about hamsters 0.3.

OBTW, Romi now only spills the blood of virtual Hamsters. She says; "it's just as much fun since the blood and guts seem so real, but when I pull the wire out of the back of my head there's no mess to clean up!"

I'll post the results when done.

Edit: Now +25 -10 =13. Romi at the half way point is usually +5, not +15. I sure hope the second half goes as expected! Retaliation expected!!

Did you do anything that would reduce the node speed of Romi? Romi used to search slightly more than a million nodes on my computer towards the middle of the game, but now she is barely doing half a million nodes. So far the node king on my computer is Pharaon, doing around 20 million nodes per second (once in a while) in a 1minute+1 second match

Hey Tony,

Ummm, no, nothing to slow it down, if anything it is slightly faster now. If what you say is accurate then I suppose that means Romi should be 50 to 80 elo better than your test results indicate. Maybe, I sent a debug version to you by mistake. Is the exe file size also way bigger? I will check my computer to see if the zip that I sent to you still exist and test it if it does. I'll email you as soon as I find something out.

Mike

[Tony Thomas]

I guess we shouldnt tell him that Romi is starting to kill hamsters again about 67% of the time.

Tsk... you are scoring about 13% against each hamster, nothing to be proud of really...

But, Romi is just using one hand to battle those determined little beast, one finger for each hamster!

Since, I finally saw Tony's results for NG1 and figured out what must have been the cause for the strange results, changes have been made. Ninja Girl 2 is now playing a hundred game match vs. Hamsters 0.2 and after 46 games complete the score is +23 -10 =13 in favor of Romi. The good news is (bad if you are a hamsters fan) is that Romi is very strong in the second half of a Sherwin50.pgn match. If all goes according to established trends I fully expect Romi to finally get a 70%+ score vs. Hamsters 0.2. And if indeed that is the case then we shall see about hamsters 0.3.

OBTW, Romi now only spills the blood of virtual Hamsters. She says; "it's just as much fun since the blood and guts seem so real, but when I pull the wire out of the back of my head there's no mess to clean up!"

I'll post the results when done.

Edit: Now +25 -10 =13. Romi at the half way point is usually +5, not +15. I sure hope the second half goes as expected! Retaliation expected!!

Did you do anything that would reduce the node speed of Romi? Romi used to search slightly more than a million nodes on my computer towards the middle of the game, but now she is barely doing half a million nodes. So far the node king on my computer is Pharaon, doing around 20 million nodes per second (once in a while) in a 1minute+1 second match

Hey Tony,

Ummm, no, nothing to slow it down, if anything it is slightly faster now. If what you say is accurate then I suppose that means Romi should be 50 to 80 elo better than your test results indicate. Maybe, I sent a debug version to you by mistake. Is the exe file size also way bigger? I will check my computer to see if the zip that I sent to you still exist and test it if it does. I'll email you as soon as I find something out.

Mike

Executable is the same size, the node speed has been steadily decreasing for the past few versions, tho not significant. I only watched about 10 games of the Beta NG1, and all of them showed less than usual node counts. May be you changed the way you compile things, also you might want to try a Bryan Hoffman, JA, or DC compiles, those dudes surely no how to pull out 20% speed increase out of nowwhere.

[Michael Sherwin]

Hey Tony,

The zip that I sent is okay and runs fast as ever. Maybe you have another program that is stealing time in some way. When the new release version is ready for WBEC 15 I will ask if someone will do a pgo compile for it!

Mike

[Tony Thomas]

Hey Tony,

The zip that I sent is okay and runs fast as ever. Maybe you have another program that is stealing time in some way. When the new release version is ready for WBEC 15 I will ask if someone will do a pgo compile for it!

Mike

I think that might be the case. I played a game against it few minutes ago, Romi kicked my ass in around 20 moves. Also, she was searching around her regular node count. If I notice the above mentioned problem, I will note the opponent and see if Romi search the same amount of nodes against that opponent in all her matches.

[Michael Sherwin]

Final result:

RomiChessNG2 - Hamsters2 : 64.0/100 50-22-28 (011==11===1=1==111=011111000=011=111==0010111011==11===1=01010=1011011==1110111=11=101=11100100=1==0) 64% +100

See an expanded write-up in the tamf.