Using Houdini v1.5

[overtond]

Hi Robert,

Without going in to technical feedback {that I for one would not understand} have you written the code so that H1.5 makes use of previous analysis? I ask as I have read in many places that when H1.5 is used to play against other engines it is very strong in terms of results obtained, however, when a position is offered to H1.5 such that no previous information was available, the difference between H1.5 and other top engines {for example, R4} is less significant.

Is this something you are aware of?

ATB
David

[Laskos]

Hi Robert,

Without going in to technical feedback {that I for one would not understand} have you written the code so that H1.5 makes use of previous analysis? I ask as I have read in many places that when H1.5 is used to play against other engines it is very strong in terms of results obtained, however, when a position is offered to H1.5 such that no previous information was available, the difference between H1.5 and other top engines {for example, R4} is less significant.

Is this something you are aware of?

ATB
David

Yes, Houdini 1.5 has pattern recognition and learning from previous versions games against all the engines. It uses adaptive Monte Carlo with multidimensional parameter adapter specific to each engine it's playing. If you give Houdini 1.5 a new position to analyze, its level is much lower than Rybka 4, which is known to have the best analysis of the new positions. Sorry, new positions are not the realm of non-Rybka 4 engines.

Kai

[zullil]

Yes, Houdini 1.5 has pattern recognition and learning from previous versions games against all the engines. It uses adaptive Monte Carlo with multidimensional parameter adapter specific to each engine it's playing.
Kai

It can also solve the Saturday New York Times crossword puzzle in less than 300 msec.

No wonder Mr. Houdart wants to keep the source code to himself.

[zamar]

Yes, Houdini 1.5 has pattern recognition and learning from previous versions games against all the engines. It uses adaptive Monte Carlo with multidimensional parameter adapter specific to each engine it's playing. If you give Houdini 1.5 a new position to analyze, its level is much lower than Rybka 4, which is known to have the best analysis of the new positions. Sorry, new positions are not the realm of non-Rybka 4 engines.

Kai

This was hilarious Thanks!

[lucasart]

Yes, Houdini 1.5 has pattern recognition and learning from previous versions games against all the engines. It uses adaptive Monte Carlo with multidimensional parameter adapter specific to each engine it's playing. If you give Houdini 1.5 a new position to analyze, its level is much lower than Rybka 4, which is known to have the best analysis of the new positions. Sorry, new positions are not the realm of non-Rybka 4 engines.

Kai

This was hilarious Thanks!

Hilarious indeed !!!

[Dr.Wael Deeb]

Yes, Houdini 1.5 has pattern recognition and learning from previous versions games against all the engines. It uses adaptive Monte Carlo with multidimensional parameter adapter specific to each engine it's playing.
Kai

It can also solve the Saturday New York Times crossword puzzle in less than 300 msec.

No wonder Mr. Houdart wants to keep the source code to himself.

Amazing....thanks for pointing this out,I wasn't aware of this option....
Dr.D

[Houdini]

Without going in to technical feedback {that I for one would not understand} have you written the code so that H1.5 makes use of previous analysis? I ask as I have read in many places that when H1.5 is used to play against other engines it is very strong in terms of results obtained, however, when a position is offered to H1.5 such that no previous information was available, the difference between H1.5 and other top engines {for example, R4} is less significant.

Is this something you are aware of?

David,

Despite the funny answers other people have given, I actually think that it is a very interesting question.

Houdini does not make use of previous analysis other than the classical transposition table that every engine uses. Yet your analysis contains an element of truth, Houdini is very good at solving positions but the difference with other engines appears to be less than match results might suggest.

The answer may lie in the nature of the positions that are tested on the chess forums. Most positions that are shown here are of the "chess puzzle" nature: there is a single winning move which the engine has to find as quickly as possible. This kind of positions requires a "win finder" strategy, the engine must look for a single winning move. Some engines are very good at this, for example Spark is very impressive in tactical positions.

In a real game these "chess puzzle" positions are quite rare. In most positions there are several good moves, and it's the positional balance that decides which of the several, more or less equivalent moves is chosen. This is where Houdini shines. In a game between Houdini and Spark, Houdini would only very rarely allow Spark to reach a position in which its strong tactical win finder abilities are relevant.

Does this answer your question?

Robert

[overtond]

Thanks for responding Robert and yes, your answer has helped a great deal.

From what I have read elsewhere re: the performance of H1.5 on the STS tests shows your program is better at static analysis {by static I mean introducing a position to the program rather than the program arriving at a position from its own control} then anything else available at the present time. Imagine my confusion when I also read elsewhere that H1.5 was only good at blitz and that when Infinite Analysis was used on static positions that it wasn't as good.

I should stop taking notice of some of these opinions; unfortunately that's hard to do if you want to keep up-to-date with what's going on. Sometimes hard to filter out the noise from the data that really matters.

Thanks again for taking the time to answer.
David