1800 my @ss

[AdminX]

Who are you kidding, I set Rybka Human to Handicap 1800 and this was the game it played ...

I know of no 1800 that plays this weak. Oh well, thank heaven for Delfi Trainer.

[Event "?"]
[Site "?"]
[Date "2008.13.08"]
[Round "?"]
[White "Rybka Human"]
[Black "Ted Summers"]
[Result "*"]
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 d5 5. cxd5 exd5 6. Qf5 Bxf5 7. b3 Bxc3+ 8.
Kd1 Ne4 9. f3 Nf2# *

[Charles B.]

Setting the handicap through Aquarium?

[AdminX]

Setting the handicap through Aquarium?

Yes, So I see now it's a GUI Bug.

[Charles B.]

I know you just wanted to post that great game.

[AdminX]

I know you just wanted to post that great game.

LMAO, I was truely Shocked

[bob]

Who are you kidding, I set Rybka Human to Handicap 1800 and this was the game it played ...

I know of no 1800 that plays this weak. Oh well, thank heaven for Delfi Trainer.

[Event "?"]
[Site "?"]
[Date "2008.13.08"]
[Round "?"]
[White "Rybka Human"]
[Black "Ted Summers"]
[Result "*"]
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 d5 5. cxd5 exd5 6. Qf5 Bxf5 7. b3 Bxc3+ 8.
Kd1 Ne4 9. f3 Nf2# *

This is a tough challenge. I learned how to handicap crafty last year, and posted the results. But the problem is that my handicapping was run on our cluster, and if you used hardware significantly faster or slower, then the numbers were not so good.

It is a _very_ difficult thing to do, even for a human. Just try playing against someone and play like an 1800 player, assuming you are much stronger. It is not so easy. For a computer, it is nearly impossible, particularly to do this in a generic way that works on all platforms...

[Dr.Wael Deeb]

Who are you kidding, I set Rybka Human to Handicap 1800 and this was the game it played ...

I know of no 1800 that plays this weak. Oh well, thank heaven for Delfi Trainer.

[Event "?"]
[Site "?"]
[Date "2008.13.08"]
[Round "?"]
[White "Rybka Human"]
[Black "Ted Summers"]
[Result "*"]
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 d5 5. cxd5 exd5 6. Qf5 Bxf5 7. b3 Bxc3+ 8.
Kd1 Ne4 9. f3 Nf2# *

This is a tough challenge. I learned how to handicap crafty last year, and posted the results. But the problem is that my handicapping was run on our cluster, and if you used hardware significantly faster or slower, then the numbers were not so good.

It is a _very_ difficult thing to do, even for a human. Just try playing against someone and play like an 1800 player, assuming you are much stronger. It is not so easy. For a computer, it is nearly impossible, particularly to do this in a generic way that works on all platforms...

Thanks Bob,an intersting point of view regarding this delicate issue....

[Ovyron]

Still, people have reported that the built in handicap mode of Rybka 3 works fine (the one Ted Summers showed is the buggy version from the Aquarium GUI.)

[geots]

Who are you kidding, I set Rybka Human to Handicap 1800 and this was the game it played ...

I know of no 1800 that plays this weak. Oh well, thank heaven for Delfi Trainer.

[Event "?"]
[Site "?"]
[Date "2008.13.08"]
[Round "?"]
[White "Rybka Human"]
[Black "Ted Summers"]
[Result "*"]
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 d5 5. cxd5 exd5 6. Qf5 Bxf5 7. b3 Bxc3+ 8.
Kd1 Ne4 9. f3 Nf2# *

This is a tough challenge. I learned how to handicap crafty last year, and posted the results. But the problem is that my handicapping was run on our cluster, and if you used hardware significantly faster or slower, then the numbers were not so good.

It is a _very_ difficult thing to do, even for a human. Just try playing against someone and play like an 1800 player, assuming you are much stronger. It is not so easy. For a computer, it is nearly impossible, particularly to do this in a generic way that works on all platforms...

Thanks Bob,an intersting point of view regarding this delicate issue....

There have been a number of Chessbase articles over the years about the inherent problems involved in trying to "dumb down" a chess engine. Especially when you try to apply an ELO scale to it. Extremely difficult to be very accurate, if possible at all.


Best,

[Dr.Wael Deeb]

Who are you kidding, I set Rybka Human to Handicap 1800 and this was the game it played ...

I know of no 1800 that plays this weak. Oh well, thank heaven for Delfi Trainer.

[Event "?"]
[Site "?"]
[Date "2008.13.08"]
[Round "?"]
[White "Rybka Human"]
[Black "Ted Summers"]
[Result "*"]
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 d5 5. cxd5 exd5 6. Qf5 Bxf5 7. b3 Bxc3+ 8.
Kd1 Ne4 9. f3 Nf2# *

This is a tough challenge. I learned how to handicap crafty last year, and posted the results. But the problem is that my handicapping was run on our cluster, and if you used hardware significantly faster or slower, then the numbers were not so good.

It is a _very_ difficult thing to do, even for a human. Just try playing against someone and play like an 1800 player, assuming you are much stronger. It is not so easy. For a computer, it is nearly impossible, particularly to do this in a generic way that works on all platforms...

Thanks Bob,an intersting point of view regarding this delicate issue....

There have been a number of Chessbase articles over the years about the inherent problems involved in trying to "dumb down" a chess engine. Especially when you try to apply an ELO scale to it. Extremely difficult to be very accurate, if possible at all.


Best,

Indeed,no doubt,but I hear that Delfi does best,I mean it's approach is almost real....