Fianchetto nonsense ?

[Ras]

Here is a nice game CT800 vs. Mephisto MM4, both at 20s per move.

The CT800 starts out in a calm way, then takes a pawn during the opening. The gambit is sound because it is compensated by enough tempos, and the MM4 nicely pushes to develop its pieces. The CT800 exchanges one of its knights for the important light squared bishop and enjoys the pair of bishops. The MM4 is clearly ahead in development, but needs some time to coordinate its pieces. Meanwhile, the CT800 completes its development while holding back its centre pawns. Now the MM4 has turned the tempo compensation into spatial advantage in the centre.

The CT800, however, is fine with less centre space and opts for a setup with both of its bishops fianchettoed. The MM4 builds up pressure in the centre, eying the weak white d-pawn, but the CT800 uses its pair of bishops to stab the centre from the flanks, embracing some ideas of "hypermodern" chess strategy. This results in tactical entanglements where the CT800 tries to win a minor piece, which the MM4 can only fend off at the cost of two more pawns. Instead of directly heading for the endgame, the CT800 first launches a piece attack and finally succeeds in winning a minor piece in addition to the pawns. Since the MM4 has run out of wood, the endgame is hopeless, and the MM4 resigns.

[pgn][White "CT800"]
[Black "MM4"]
[Opening "English Opening"]
[WhiteElo "2100"]
[BlackElo "1953"]
[Date "2016.08.20"]
[Result "1-0"]
[TimeControl "1/20"]

1. Nf3
c5 2. c4 Nc6 3. d3 Nf6 4. g3 d5 {Now it is a reversed
Sicilian setup.} 5. cxd5 Nxd5 6. Bg2
e5 7. Qb3 Na5 8. Qa4+ Bd7 9. Qe4 Bd6 {Nb4 would have
kept the game equal.} 10. Nxe5 Nf6 11. Qe3 O-O 12. Nxd7
{Makes the bishop on g2 much stronger.}
Qxd7 13. Qd2 Nc6 14. Nc3 Rad8 15. O-O Rfe8 16. b3 Be5
17. Bb2 Nd4 18. e3 Nf5 19. Rfd1 b5 20. Rac1 h6 {a6
would have been better.} 21. Bh3 Bxc3 {Loosing more
material. Instead, g6 would have been the move.}
22. Bxc3 Qe6 23. Qb2 Nd5 24. Bxg7 Ndxe3 {Qg6 was much better.}
25. Re1 Qg6 26. Bf6 Rd6 27. Be5 {That bishop belonged on c3,
but still, the game is won.} Rde6 28. Rxc5 f6 29. Bc3 Qh5
30. fxe3 Qxh3 31. e4 Nd6 32. Bxf6 {White is now three pawns
up, the endgame is clear.} Rf8 33. Be5 Nf7 34. Bf4
{Closing the f-file for the moment.} a6
35. Rec1 Qg4 36. Rc7 Rd8 37. Rc8 Qf3 38. Rxd8+
{Trading down material and putting the knight out of the game.}
Nxd8 39. Qd4 Nf7 40. Qd7 Rb6 {Putting the rook off the game.}
41. Rf1 Qh5 42. Bc7 Qg6 {The rook has no escape square.}
43. Rxf7 Qxf7 44. Qxf7+ Kxf7 45. Bxb6 h5 46. Kg2 Ke6
47. d4 Kd7 48. Kh3 Kc6 49. Bc5 a5 50. Kh4 Kd7 51. d5 a4
52. e5 axb3 {Black resigned.} 1-0[/pgn]

[Henk]

I think in this position bishop is better on e2 than on g2. If so then it looks like fianchetto is only good if bishop is not blocked by an immobile pawn. But is it possible to prevent that in a closed/half open position ?

[d] r1bq1rk1/ppp1bppp/2n2n2/3pp3/4P3/2PP1N2/PP1NBPPP/R1BQ1RK1 b - - 0 7

how did that position arise? black already has an advantage.

of course, bishop on g2 is always better than bishop on e2 (at least in 95% of cases), no matter if there is a blocked own central pawn or not.

with Bg2, after d5, white could simply have captured ed5.

engines do not understand fianchettoe, simply because:

- involves longer search
- they try to apply fianchettoe term, but their move ordering and search routines do not like that, they like some quick captures instead, some quick attacks, etc.
- when do you think SF will successfully apply a fianchettoe term, when they are testing with a 2-moves book, having a big portion of lines like: 1.h4 a5 2.g4 b5 ? ; of course, SF will never succeed with this, as fianchettoe is simply impossible/counterproductive above. if you have 30/40% of lines like that, the term will fail of course.

[pgn]
[Event "?"]
[Site "?"]
[Date "?"]
[Round "-"]
[White "?"]
[Black "?"]
[Result "*"]

1. d3 e5 2. g3 d5 3. Bg2 Nf6 4. Nf3 Nc6 5. O-O Be7 6. Nbd2 O-O 7. e4
*
[/pgn]

So white had no opportunity to capture on d5. Now black will probably capture on e4. If it continues with f4 ef gf later on, it will give hanging pawns and bad king safety for white or not ? But perhaps black should not capture on f4 and instead create a block on e5 that makes bishop on g2 a weakness.

[Henk]

If black succeeds in creating a blockade on e5 white might play f5. But with an open d-file a king attack with g2-g4-g5 may at least be risky.

So instead white should continue with c3 or perhaps a4? Position looks very drawish. Perhaps bishop on e2 better after de, de

[Henk]

Probably black better here.

[pgn]
[Event "?"]
[Site "?"]
[Date "?"]
[Round "-"]
[White "?"]
[Black "?"]
[Result "*"]

1. d3 e5 2. g3 d5 3. Bg2 Nf6 4. Nf3 Nc6 5. O-O Be7 6. Nbd2 O-O 7. e4 dxe4
8. dxe4 Bc5 9. c3 a5 10. Qc2 Be6 11. Ng5 Bd7 12. Ngf3 h6 13. Nc4 Qe7
*
[/pgn]

[Henk]

Probably white better here. But don't know if a5 was a good move.
[pgn]
[Event "?"]
[Site "?"]
[Date "?"]
[Round "-"]
[White "?"]
[Black "?"]
[Result "*"]
[FEN "r1bq1rk1/ppp1bppp/2n2n2/3pp3/4P3/2PP1N2/PP1NBPPP/R1BQ1RK1 b - - 0 1"]
[SetUp "1"]

{--------------
r . b q . r k .
p p p . b p p p
. . n . . n . .
. . . p p . . .
. . . . P . . .
. . P P . N . .
P P . N B P P P
R . B Q . R K .
black to play
--------------}
1... dxe4 2. dxe4 a5 3. Bb5 Bc5 4. Qe2 Qe7 5. a4
*
[/pgn]

[Lyudmil Tsvetkov]

Stockfish 8 says this position is slightly better for black

[d] r1bq1rk1/ppp1bppp/2n2n2/3pp3/4P3/3P1NP1/PPPN1PBP/R1BQ1RK1 b - - 3 7


According to Stockfish 8 position below is equal.

[d] r1bq1rk1/ppp1bppp/2n2n2/3pp3/4P3/2PP1N2/PP1NBPPP/R1BQ1RK1 b - - 0 7


So according to Stockfish 8 fianchetto wrong decision in this position.

you know very well that usually the truth is opposite to what SF claims, so please relax.

still, I do not see why white does not capture on d5.

[Lyudmil Tsvetkov]

I think in this position bishop is better on e2 than on g2. If so then it looks like fianchetto is only good if bishop is not blocked by an immobile pawn. But is it possible to prevent that in a closed/half open position ?

[d] r1bq1rk1/ppp1bppp/2n2n2/3pp3/4P3/2PP1N2/PP1NBPPP/R1BQ1RK1 b - - 0 7

how did that position arise? black already has an advantage.

of course, bishop on g2 is always better than bishop on e2 (at least in 95% of cases), no matter if there is a blocked own central pawn or not.

with Bg2, after d5, white could simply have captured ed5.

engines do not understand fianchettoe, simply because:

- involves longer search
- they try to apply fianchettoe term, but their move ordering and search routines do not like that, they like some quick captures instead, some quick attacks, etc.
- when do you think SF will successfully apply a fianchettoe term, when they are testing with a 2-moves book, having a big portion of lines like: 1.h4 a5 2.g4 b5 ? ; of course, SF will never succeed with this, as fianchettoe is simply impossible/counterproductive above. if you have 30/40% of lines like that, the term will fail of course.

[pgn]
[Event "?"]
[Site "?"]
[Date "?"]
[Round "-"]
[White "?"]
[Black "?"]
[Result "*"]

1. d3 e5 2. g3 d5 3. Bg2 Nf6 4. Nf3 Nc6 5. O-O Be7 6. Nbd2 O-O 7. e4
*
[/pgn]

So white had no opportunity to capture on d5. Now black will probably capture on e4. If it continues with f4 ef gf later on, it will give hanging pawns and bad king safety for white or not ? But perhaps black should not capture on f4 and instead create a block on e5 that makes bishop on g2 a weakness.

6.Nc3 is much stronger; white should try to win that

[Henk]

Stockfish 8 says this position is slightly better for black

[d] r1bq1rk1/ppp1bppp/2n2n2/3pp3/4P3/3P1NP1/PPPN1PBP/R1BQ1RK1 b - - 3 7


According to Stockfish 8 position below is equal.

[d] r1bq1rk1/ppp1bppp/2n2n2/3pp3/4P3/2PP1N2/PP1NBPPP/R1BQ1RK1 b - - 0 7


So according to Stockfish 8 fianchetto wrong decision in this position.

you know very well that usually the truth is opposite to what SF claims, so please relax.

still, I do not see why white does not capture on d5.

Because it is blacks turn.

[Henk]

I think in this position bishop is better on e2 than on g2. If so then it looks like fianchetto is only good if bishop is not blocked by an immobile pawn. But is it possible to prevent that in a closed/half open position ?

[d] r1bq1rk1/ppp1bppp/2n2n2/3pp3/4P3/2PP1N2/PP1NBPPP/R1BQ1RK1 b - - 0 7

how did that position arise? black already has an advantage.

of course, bishop on g2 is always better than bishop on e2 (at least in 95% of cases), no matter if there is a blocked own central pawn or not.

with Bg2, after d5, white could simply have captured ed5.

engines do not understand fianchettoe, simply because:

- involves longer search
- they try to apply fianchettoe term, but their move ordering and search routines do not like that, they like some quick captures instead, some quick attacks, etc.
- when do you think SF will successfully apply a fianchettoe term, when they are testing with a 2-moves book, having a big portion of lines like: 1.h4 a5 2.g4 b5 ? ; of course, SF will never succeed with this, as fianchettoe is simply impossible/counterproductive above. if you have 30/40% of lines like that, the term will fail of course.

[pgn]
[Event "?"]
[Site "?"]
[Date "?"]
[Round "-"]
[White "?"]
[Black "?"]
[Result "*"]

1. d3 e5 2. g3 d5 3. Bg2 Nf6 4. Nf3 Nc6 5. O-O Be7 6. Nbd2 O-O 7. e4
*
[/pgn]

So white had no opportunity to capture on d5. Now black will probably capture on e4. If it continues with f4 ef gf later on, it will give hanging pawns and bad king safety for white or not ? But perhaps black should not capture on f4 and instead create a block on e5 that makes bishop on g2 a weakness.

6.Nc3 is much stronger; white should try to win that

May be.

Nd2-c4 attacks pawn e5.

Also I probably would not play 4. Nf3. It is because Stockfish says so.

[Henk]

Stockfish 8 says this position is slightly better for black

[d] r1bq1rk1/ppp1bppp/2n2n2/3pp3/4P3/3P1NP1/PPPN1PBP/R1BQ1RK1 b - - 3 7


According to Stockfish 8 position below is equal.

[d] r1bq1rk1/ppp1bppp/2n2n2/3pp3/4P3/2PP1N2/PP1NBPPP/R1BQ1RK1 b - - 0 7


So according to Stockfish 8 fianchetto wrong decision in this position.

you know very well that usually the truth is opposite to what SF claims, so please relax.

still, I do not see why white does not capture on d5.

Because it is blacks turn.

[pgn]
[Event "?"]
[Site "?"]
[Date "?"]
[Round "-"]
[White "?"]
[Black "?"]
[Result "*"]

1. d3 Nc6 2. e4 e5 3. Nf3 Nf6 4. Be2 d5 5. Nbd2 Be7
*
[/pgn]

This is how position was created. Indeed White already had opportunity to capture on d5.