Philidor IV

[Sean Evans]

Black needed to develop more pieces, white was dominating with development.

[Lyudmil Tsvetkov]

of course, as Gerold suggests, Ng4. I do not see any compensation for the pawn.

I do - the open g file and the rook on said file. Probably, that isn't enough compensation, but at least some. The main issue for Black is that castling kingside is out of question, and castling queenside takes more time.

DeepShredder 13 evaluates the position after 5. ... Nxg4 with -0.38 in favour of Black, and after 5. ... h6 with +0.15 in favour of White, both at depth 33. That's a difference of 0.53 pawns. So the compensation is worth 0.47 pawns, according to DeepShredder.

g is a semi-open, and not an open file.

also, if black has 40cps edge after the knight capture, there is certainly no compensation.

I guess Shredder assesses both lines accurately here.

if I have to go further, I guess Ng4 simply wins the game, but that is not easy to prove.

[Ras]

also, if black has 40cps edge after the knight capture, there is certainly no compensation.

If there were no compensation at all, I would expect black to have an edge of 100cps, the material value of the pawn.

I've been digging a bit deeper through the variants to include this gambit as passive knowledge in the CT800's book. The issue is that Black has the better endgame if nothing changes, but reaching that is not easy.

Black cannot castle kingside, and in most variants, not queenside, either. This in turn seriously hampers the rooks' mobility. In fact, with absolutely correct play, Black has the edge, but it is quite a narrow path from having the edge to losing the game.

Looks like a good gambit especially for blitz games, or for tournament sitations where White must make the full point at all cost.

[Henk]

[pgn]
[Event "Bundesliga 2008/09"]
[Site "Various GER"]
[Date "2008.10.04"]
[EventDate "2008.??.??"]
[Round "1.1"]
[Result "1-0"]
[White "Alexey Shirov"]
[Black "Lukasz Cyborowski"]
[ECO "C41"]
[WhiteElo "2741"]
[BlackElo "2546"]
[PlyCount "45"]

1. e4 d6 2. d4 Nf6 3. Nc3 e5 4. Nf3 Nbd7 5. g4 Nxg4 6. Rg1
Ngf6 7. Bc4 h6 8. Be3 Nb6 9. Bb3 exd4 10. Qxd4 Bg4 11. Rg3 c5
12. Qd3 Rc8 13. e5 dxe5 14. Nxe5 Bh5 15. Qf5 Bd6 16. Ne4 Qe7
17. Nxd6+ Qxd6 18. Rxg7 Rc7 19. Bxf7+ Bxf7 20. Rd1 Nbd5
21. Nxf7 Qxh2 22. Nxh8 Rxg7 23. Qc8+ 1-0
[/pgn]

[Henk]

Actually there were at least ten videos from Shirov about Philidor defense on Youtube. But I can't find them anymore. Probably he deleted them.

[Lyudmil Tsvetkov]

Actually there were at least ten videos from Shirov about Philidor defense on Youtube. But I can't find them anymore. Probably he deleted them.

or someone, who hates Shirov, deleted them, or someone, who was appalled by the gambit...

only playable gambit for white is the Evans gambit.
no playable gambits for black.

[Henk]

Other video:

https://www.youtube.com/watch?v=AhB1WpKFFck

[Ras]

Shredder suggests 12. ... Qa5 instead of Qc7, with half a pawn advantage for Black. That loses control over d6, which was why Qc7 was suggested, but for the moment, White cannot get that square anyway.

Even 12. ... Qc7 is not as bad as the video suggests, it's about even. After 13. 0-0-0, Black must not play ... a5, that brings White into clear advantage, and after 14. a4, ... b4 is completely losing.

But Black has several moves with minor advantage here: 13. ... g6, g5 or even Bb7.

The main problem with declining the gambit with h6 instead of Nxg4 is that Black still has no options in the centre and still cannot castle kingside because the pawn shelter gets weak after White plays g5. Plus that White can open up the centre with dxe5 at any time. Which is what White also does if Black accepts the gambit, but at least, Black can hold on and has the pawn as endgame insurance.

[Ras]

On the other hand, I have been digging deeply through the variations of the accepted gambit, and with optimum play by both sides, it looks pretty equal. The issue is just that the paths are very narrow for both sides, which makes the opening sharp.

Actually, I would not recommend this opening for a human versus a strong engine. But with human vs. human, White may play this if White is either a superior player or better prepared. The only drawback is that in human games, Black is unlikely to go for Philidor these days.

It's also a good option for strong players against dedicated units (just like e.g. the Blackmar-Diemer-Gambit) as these will probably not have this opening in the book - with at least one exception, of course.

[AdminX]

What should black play against this gambit ? Don't tell me he has to exchange it's queen against a rook, bishop and a pawn.

[d] r1bqkb1r/pppn1ppp/3p1n2/4p3/3PP1P1/2N2N2/PPP2P1P/R1BQKB1R b KQkq g3 0 5

That depends on what type of player you are. Strategic, Tactical, or Positional. As for me, I like the initiative and active positions so I might try d5 here based of the principal that an attack on the wing is countered by an attack in the center. Mind you I am not adverse to capturing with Nxg4, it just depends on the mood I am in.

PS: I do play the Black Lion but rarely does white play g4 in my games with black, because I normally get in g5 first and it is supported by my own pawn on h6.