In a "Monte Carlo" test, indeed White comes ahead with about 70%. But earlier in this thread we have already seen that this kind of test fails when the defender has to play carefully. The 70% could be realistic for players below 1500, but around 1800-2000 I'd already expect a 60%, and even less (maybe 55%) in upper Elo regions.
Rybka may be an extreme case, but I'd think that many programs would overestimate White's chances: the bishop pair alone translates to +0.30, - maybe more, if the software sees that White can even take the remaining black bishop with his knight. Plus the doubled pawn, plus three pawn islands vs two. An end result of +1.14 may be almost meaningless, if the software is dumb enough not to consider other rules: bishops are worse in exploiting such pawn weaknesses, a doubled pawn in the center safely protected by Black's pieces is almost a bonus for the party with the knights. And so on. The main point in my last post was that the ideas of the Bulgarian Ouch or the Fyfe are difficult to understand for the engines.
Vass wrote on 09/10/12 at 16:55:53:
Well, Stefan, I don't believe that white is almost +\- in the final position. My assessment is +\=. I think your human understanding is right. Yes, white has the two bishops, more space (after a2-a4, b2-b4) and so on, but if you continue to play further you'll see that little by little the engine evaluation goes down a bit.
Anyway, this variation is another example for the engines' wrong assessment in the openings. I would like to point at the move 8.f4!? here. Perfectly playable in my opinion and yet not an engine will show it as its first option. And only if you insist on it and give it to the engine to calculate your idea behind it, then suddenly...as if darkness comes to light..it says: "A-ha, so.. This line...you meant!.."
What to say?!.. Calculators!
Yes, your proposal 8.f4 is truly human. Only humans play so actively when the opponent has the better development.
I had looked at a similar idea, where Black has played Qd7 instead of h5-h4. So with the pawn still on h5 and the queen already on d7, the move f2-f4 seemed too risky. Maybe the aggressive advance is playable in the present situation?
One of your lines goes:
8. f4 exf4 9. gxf4 h3 10. Bf3 Nh7 11. Qe2 Qh4 12. Qf2 Qxf4 13. Nd5 Qe5 14. Ne2 Bxd5 15. exd5 Nb4 16. Bf4. However, instead of 10...Nh7, Black has the stronger
10...Nd7! 11.f5 Nde5 12.fxe6 fxe6!. Who stands better? I'd estimate that in a poll 70% will favour Black. [Note that 11.Qe2 Qh4+ 12.Qf2 Qxf4 13.Nd5 Qe5 14.Ne2, in analogy to Vass' analysis, runs into 14...Bc5!.]
Pages: 
C25: The Everlasting Fyfe Gambit (Read 198239 times)
