Question to Larry Kaufman about Rybka

[mcostalba]

Depends on what you call "enough". The eval is greatly simplified, and some search parameters are changed, but is that enough to say that they really understand the program? My own feeling is no.

At the end of the day currently it seems Robbo is 50 ELO higher the Rybka 3, I would guess that improving a monster like R3 of 50 ELO is not a kid's joke. You cannot go monky style through the sources and hope to get something 50 ELO stronger at the end of the show.

[Zach Wegner]

Depends on what you call "enough". The eval is greatly simplified, and some search parameters are changed, but is that enough to say that they really understand the program? My own feeling is no.

From what I have seen the answer is an emphatic "yes". In order to make so many changes and not decrease the strength, but rather improve it, you really have to understand what is going on.

[lkaufman]

"I tested komdo against robbolito at fixed depth of 1 ply and 2 plies and it seems that robbolito is clearly stronger at small fixed depth.

I think that you should ask the question why komodo is weaker.
I think that komodo tends to prune too much. "

Robbo crushes Komodo by nearly 200 Elo at 1 or 2 ply, but Robbo also crushes other normal programs like Fritz at those depths. This gap goes away by the time you get out to around 5 ply. It's true that Komodo would see a few more tactics with less restrictions on the quiescence search, but our tests indicate that in general the slowdown would make us weaker. That's not to say that we couldn't improve the quiescence search a bit, maybe five or ten Elo, but this is not the answer to the Robbo mystery.
So does anyone have a clue as to why Robbo is so much stronger than standard engines at 1 or 2 ply but not at 5 ply?

[mcostalba]

"I tested komdo against robbolito at fixed depth of 1 ply and 2 plies and it seems that robbolito is clearly stronger at small fixed depth.

I think that you should ask the question why komodo is weaker.
I think that komodo tends to prune too much. "

Robbo crushes Komodo by nearly 200 Elo at 1 or 2 ply, but Robbo also crushes other normal programs like Fritz at those depths. This gap goes away by the time you get out to around 5 ply. It's true that Komodo would see a few more tactics with less restrictions on the quiescence search, but our tests indicate that in general the slowdown would make us weaker. That's not to say that we couldn't improve the quiescence search a bit, maybe five or ten Elo, but this is not the answer to the Robbo mystery.
So does anyone have a clue as to why Robbo is so much stronger than standard engines at 1 or 2 ply but not at 5 ply?

Perhaps because is not at 1 or 2 ply this is only for the other engine but not for Robbo.

hint: how do you take in account extensions ?....ask Don for the answer

[Uri Blass]

Depends on what you call "enough". The eval is greatly simplified, and some search parameters are changed, but is that enough to say that they really understand the program? My own feeling is no.

A similar question for you. Do you Larry understand Rybka's search and why is it so efficient?

I am not larry but I can give my opinion.

My opinion is that the reason is a smart choice of moves not to prune at the qsearch.

I guess that the moves that rybka does not prune at small depths also allow her to earn rating by more aggresive reductions(when the same reductions are counter productive with Komodo's qsearch).

Uri

[BubbaTough]

It would take a lot of time and effort to unravel what they are doing.

I sugegst to look at ippolit big ipp_eng.c single file, with a good C++ IDE is not so difficult to browse and a lot of code is redundant becasue is duplicated for white/black and for the different searches, so once you do the effort for one kind of search and one color you have almost done it, remaining code is almost the same with little variations.

Ahh, that would make it much easier! I only looked at Robo stuff. If I ever dive back into that mess, I will take up your suggestion.

-Sam

[lkaufman]

Definitely not. If I did, either the ideas would already be in Komodo and it would be as strong as Robbo, or else I wouldn't have worked on Komodo at all as a matter of principle. But of course I do understand some of ideas that are now talked about on this forum.

[Milos]

Definitely not. If I did, either the ideas would already be in Komodo and it would be as strong as Robbo, or else I wouldn't have worked on Komodo at all as a matter of principle. But of course I do understand some of ideas that are now talked about on this forum.

Thx for sincere answer. I also don't know what makes Rybka search so efficient since I haven't looked at Rybka source code, and didn't spend time trying to disassemble it. I can only assume what it is (there I agree with Uri's assumption).
However, my firm belief is that publishing of Ippo sources doesn't brings us any closer to the key of Rybka strength.

[Ferdy]

"I tested komdo against robbolito at fixed depth of 1 ply and 2 plies and it seems that robbolito is clearly stronger at small fixed depth.

I think that you should ask the question why komodo is weaker.
I think that komodo tends to prune too much. "

Robbo crushes Komodo by nearly 200 Elo at 1 or 2 ply, but Robbo also crushes other normal programs like Fritz at those depths. This gap goes away by the time you get out to around 5 ply. It's true that Komodo would see a few more tactics with less restrictions on the quiescence search, but our tests indicate that in general the slowdown would make us weaker. That's not to say that we couldn't improve the quiescence search a bit, maybe five or ten Elo, but this is not the answer to the Robbo mystery.
So does anyone have a clue as to why Robbo is so much stronger than standard engines at 1 or 2 ply but not at 5 ply?

Superb move ordering . In a position where there are many playable moves, that engine is able to estimate the best move or close to the best move most of the time.

[Edsel Apostol]

"I tested komdo against robbolito at fixed depth of 1 ply and 2 plies and it seems that robbolito is clearly stronger at small fixed depth.

I think that you should ask the question why komodo is weaker.
I think that komodo tends to prune too much. "

Robbo crushes Komodo by nearly 200 Elo at 1 or 2 ply, but Robbo also crushes other normal programs like Fritz at those depths. This gap goes away by the time you get out to around 5 ply. It's true that Komodo would see a few more tactics with less restrictions on the quiescence search, but our tests indicate that in general the slowdown would make us weaker. That's not to say that we couldn't improve the quiescence search a bit, maybe five or ten Elo, but this is not the answer to the Robbo mystery.
So does anyone have a clue as to why Robbo is so much stronger than standard engines at 1 or 2 ply but not at 5 ply?

It has a better tuned eval function and superior qsearch.