Problem 21466

[drahacikfm]

I propose this problem should be deleted.  The solution given is 1...exf3 2.Bxf3 Nxd5 and the problem stops.  But what is the point after 3.Bxd5 or exd5?  Black has not shown the point yet, and I sure don't see it myself.

I put the position after 1...exf3 2.Bxf3 on Rybka, and it gives
2...Nxd5 -2,90
2...Ne5 -1.66
2...Qe8+ -1.56

I don't think there's enough difference between the best move and the 2nd and 3rd best moves.

16 people have tried this problem so far, and nobody got it right.

[richard]

I've had a quick look at this and my version of toga II is giving odd evaluations for this.  I've noticed in the past that toga can sometimes get confused when using multipv mode (the UCI mode to show multiple lines) and I think this may be the case here.  Looking at what the engine is doing this looks more likely a bug in toga multipv than my problem generator.

I'd love to be able to use Rybka as my evaluation engine, however I need it to work on linux and mac os x to be useful for me.  I just did a quick search and it seems the next version will probably run on linux and there are ways of getting the current versions running on linux and the mac:
http://www.rybkachess.com/index.php?auswahl=Installation

I have a few other things to get done first but I'm planning a manual run through all the 1 star rated problems to see which ones are most in need of culling.  Adding comments to the problem as you have done is very helpful in this regard.

Regards,
Richard.

[drahacikfm]

If you do decide to change your evaluation engine, note that not all evaluations are created equal!  Rybka is well-known to have very low evaluations.  On my Rybka 2.2 (not the latest version), if you take the initial position of a chess game, and remove a Black knight, it says White has an advantage of only about +1.5.  Other engines will say +3.0 or so.  This version of Rybka seems to have evaluations about half as much as other engines.  So you would have to adjust the parameters in your generator accordingly.

[richard]

Interesting, I hadn't heard that (I haven't read much about Rybka, other than it seems to be the leading chess engine right now). Do you feel it's lower evaluations are justified by deeper positional analysis of the situation or is it just a generally pessimistic engine? :-)

Regards,
Richard.

[drahacikfm]

I don't know exactly, but I heard one explanation that Rybka values an extra pawn as 0.5 in the opening and 1.0 in the endgame.  So in the opening a knight is only worth 1.5.

[slacker00]

I don't understand this problem either.

If you click the computer icon to get a further continuation, the suggested line continues with 3. 0-0, which really floored me.  Does black have a big attack that we're not seeing?  I mean, for White to castle instead of equalizing material with 3.BxN or 3.PxN, black would have to  have something much more that what  we're seeing.

[tama]

OK i gust got this problem and got it wrong because i didnt c the first move being a 2 square push from a pawn, because of this i failed. I wonder if you could make the first move take like 2 sec longer to apear? I even studied this problem in the past and recognized it but didnt remember the line because i didnt c the first move.

[richard]

tama: If you are often missing the first move you could turn on highlight last move which will show you where the piece moved from and to.

Richard.