Chessbase article regarding Fritz 15

[Laskos]

The "style of play" of Fritz 15 is almost indistinguishable from that of Rybka 4.

I don't think you can conclude that at all from the dendrogram. "Style of play" is very subjective and not entirely measurable in my opinion.

Even if the "style" is similar, that does not mean that substantial code changes were not made. Of course any engine he writes will retain some of the Rybka flavour, it is inevitable. But over time the two engines may start to diverge more.

"Style of play" is almost exactly what this dendrogram is showing. It shows the similarity of what engines' play in given positions and time control. "Style of play" is described by the succession of played moves in given positions, it's that simple. Now, Vas can be right, and he changed somewhat the search. That wouldn't be clear on this dendrogram, as it shows mostly the eval part.

[Laskos]

Hi Kai,

the reason that I don't read it.
Experience with commercial material, it's allways the same.

I try in the time I am working for Schach-Welt (German newspaper by GM Hickl) to make an interview to the current Fritz Version at this time. I made a lot of interviews with chess programmers at this time.

From ChessBase ...
We have our once interview ...
5 questions and answers on ChessBase site.

I have the questions to that interview on my harddisk ...
Unfortunately, much more interesting as that what we got ... in my opinion.

But fact ist that ChessBase do a good work with ChessBase database and Fritz GUI and all the other things. Good that we have the commercials, also in Winboard times. The secret is a good mix I think.

Best
Frank

It's a bit annoying when Chessbase tries again to appeal to those less informed on computer chess, just because their buyers' base is such. More so when the article and interview is by someone very competent like Albert Silver. He even suggests that Fritz 15 peculiarity with respect to Rybka 4 consists of "Wild Tactics and attacking" and "The boa constrictor".

[velmarin]

It's a bit annoying when Chessbase tries again to appeal to those less informed on computer chess, just because their buyers' base is such. More so when the article and interview is by someone very competent like Albert Silver. He even suggests that Fritz 15 peculiarity with respect to Rybka 4 consists of "Wild Tactics and attacking" and "The boa constrictor".

Albert Silver only makes two questions, the correct two, of course they sell their products Chessbase.
It is a beautiful and correct article.
Nothing to do with the Komodo propaganda in this forum.
foolish games with material difference, that if it is a crap.

[Laskos]

Nothing to do with the Komodo propaganda in this forum.
foolish games with material difference, that if it is a crap.

I don't know if it is Komodo propaganda, or just that many people are interested in the strongest engines. At least Komodo people don't take the audience for illiterates in computer chess. As for material handicap games, to some, including me, they are quite interesting. In fact the most interesting topics to me in that somewhat boring (to me) sub-forum of Tournaments and Matches.

[MMoisan]

http://en.chessbase.com/post/fritz-15-e ... nd-answers

I hope that this is not seen as commercial exhortation.
It just gives more information about the engine for those interested.

For those who are not interested, there is no need for silly comments. Just ignore the post.

Nice to hear that Fritz 15.1 is coming sometime soon. The rest of the article is composed either of usual marketing blah-blah or misleading statements. Also, if they talk of "styles of play" and cherry pick some games to illustrate them, in the context of differences from Rybka 4, I would suggest to have a look at my older dendrogram:



The "style of play" of Fritz 15 is almost indistinguishable from that of Rybka 4.

Hi Kai,

Am I reading the dendrogram correctly if I say that Gull is closer to Stockfish than from Rybka or Houdini? I thought that Gull was a clone from Ippolit. It doen't show that much in your test.

Thanks!,

Marc

[Laskos]

http://en.chessbase.com/post/fritz-15-e ... nd-answers

I hope that this is not seen as commercial exhortation.
It just gives more information about the engine for those interested.

For those who are not interested, there is no need for silly comments. Just ignore the post.

Nice to hear that Fritz 15.1 is coming sometime soon. The rest of the article is composed either of usual marketing blah-blah or misleading statements. Also, if they talk of "styles of play" and cherry pick some games to illustrate them, in the context of differences from Rybka 4, I would suggest to have a look at my older dendrogram:



The "style of play" of Fritz 15 is almost indistinguishable from that of Rybka 4.

Hi Kai,

Am I reading the dendrogram correctly if I say that Gull is closer to Stockfish than from Rybka or Houdini? I thought that Gull was a clone from Ippolit. It doen't show that much in your test.

Thanks!,

Marc

No, the plot is read the other way around: nearest common ancestor from left to right gives the closest engines. Here Fritz 15 and Rybka 4 are the closest. Gull 3, Stockfish, Komodo 9 are almost unrelated to anything.

[MMoisan]

Oh, I see. Thank you for answering.

[Ozymandias]

It's a useful article, mostly because it shows that Vas is back in action. Once his test-bed is setup we might start to see some elo.

It felt fabricated to me, not like a real interview. Just a couple of quotes and pictures, pasted together. BTW, the pics are from an old piece (http://en.chessbase.com/post/rajlich-bu ... e-for-sure) which looks more serious, even though it was an April Fools' joke.

[Modern Times]

It's a useful article, mostly because it shows that Vas is back in action. Once his test-bed is setup we might start to see some elo.

Yes, I think the article is saying that Vas's involvement with Fritz is long term, which is fantastic. A Fritz 15.1 is mentioned, and then we will see further strength improvement with subsequent versions. A long way to catch up to Komodo and Stockfish, but if anyone can do it, Vas can.