Page 2 of 4

Re: Is the Blackmar-Diemar Gambit playable?

Posted: Thu Jul 05, 2018 11:07 pm
by Dann Corbit
BeyondCritics wrote:
Thu Jul 05, 2018 10:46 pm
mclane wrote:
Thu Jul 05, 2018 7:41 pm
Yes.playable.
What do you play against 5..g6?
This analysis is too shallow for serious use, but Bg5 is suggested at depth 37:

rnbqkb1r/ppp1pp1p/5np1/8/3P4/2N2N2/PPP3PP/R1BQKB1R w KQkq - acd 37; acs 441; bm Bg5; cce -83; ce -54; pm Bg5; pv Bg5 Bg7 Bc4 Nc6 Qd2 O-O O-O-O a6 a3 b5 Bb3 Rb8 Ne5 Nxe5 dxe5 Qxd2+ Rxd2 Ng4 Nd5 Nxe5 Nxe7+ Kh8 Nxc8 Rbxc8 Bf4 Rfe8 Rhd1 c6 Bxe5 Bxe5 Bxf7 Re7 Rd8+ Rxd8 Rxd8+ Kg7 Be8 c5 g3 Re6 c3 Bf6 Rd7+ Kh6 Bf7 Re1+ Kc2 c4 Bd5 Re2+ Kb1 Rxh2 Ra7 Bg5 Rxa6; white_wins 42; black_wins 73; draws 50; Opening Blackmar-Diemer Gambit: Bogoljubow Variation. 1.d4 d5 2.e4 dxe4 3.Nc3 Nf6 4.f3 exf3 5.Nxf3 g6; CaxtonID: 181; ECO: D00;

Re: Is the Blackmar-Diemar Gambit playable?

Posted: Fri Jul 06, 2018 12:07 am
by jdart
IM Gary Lane has a book on the gambit, with lots of analysis (https://www.amazon.com/Blackmar-Diemer- ... 0713477253).

Re: Is the Blackmar-Diemar Gambit playable?

Posted: Fri Jul 06, 2018 6:50 am
by mclane
A computer software cannot play a gambit.
It has no clue about the sense of a gambit.
It will defend material although the idea is to sac material.

It is not playing chess. It emulates playing chess.

Re: Is the Blackmar-Diemar Gambit playable?

Posted: Fri Jul 06, 2018 7:07 am
by Uri Blass
mclane wrote:
Fri Jul 06, 2018 6:50 am
A computer software cannot play a gambit.
It has no clue about the sense of a gambit.
It will defend material although the idea is to sac material.

It is not playing chess. It emulates playing chess.
chess programs also sacrifice material for positional compensation.
Every strong program does it and programs sometimes make wrong sacrifices

After 1.e4 d5 2.exd5 my stockfish suggest 2...Nf6

Re: Is the Blackmar-Diemar Gambit playable?

Posted: Fri Jul 06, 2018 10:09 am
by mclane
how many chess engines (book turned off) you know play d4 d5 c4 ??

of course chess engines do to a certain degree sac for compensation.
but in a gambit, the idea is different. the compensation e.g. in the BDG is much deeper then the chess engines could smell.

its not about compensation for a material win. its about a compensation to mate the opponent.

the BDG is a good example. Kings gambit and queens gambit or englund gambit and budapest gambit
or any other gambit also show that the chess engines fail to emulate chess understanding.

and this behaviour has IMO not changed over long term of computer chess development.

the programs compute deeper and they kill humans. but they do not play better in gambits. its because they have no clue that the target is mating . instead they play for material wins.


lets hope the latest developments in computerchess help to solve this emulation lack.

Re: Is the Blackmar-Diemar Gambit playable?

Posted: Fri Jul 06, 2018 11:05 am
by F. Bluemers
A minor point:
how many chess engines (book turned off) you know play d4 d5 c4 ??
Phalanx has/had special code for these pawnformations.
Best
Fonzy

Re: Is the Blackmar-Diemar Gambit playable?

Posted: Fri Jul 06, 2018 1:17 pm
by mclane
BeyondCritics wrote:
Thu Jul 05, 2018 10:46 pm
mclane wrote:
Thu Jul 05, 2018 7:41 pm
Yes.playable.
What do you play against 5..g6?
6.Bc4

Re: Is the Blackmar-Diemar Gambit playable?

Posted: Fri Jul 06, 2018 3:37 pm
by todd
1. d4 Nf6 2. Nc3 d5 3. e4 Nxe4 is nice if you can get it, but beware - white can play 1. d4 Nf6 2. f3 d5 3. e4 dxe4 4. Nc3 and you've been tricked into the dxe4 variation.

Re: Is the Blackmar-Diemar Gambit playable?

Posted: Fri Jul 06, 2018 4:17 pm
by tpoppins
todd wrote:
Fri Jul 06, 2018 3:37 pm
1. d4 Nf6 2. Nc3 d5 3. e4 Nxe4 is nice if you can get it, but beware - white can play 1. d4 Nf6 2. f3 d5 3. e4 dxe4 4. Nc3 and you've been tricked into the dxe4 variation.
Black doesn't have to play along. Instead of 3...dxe4 he can meet 3.e4 with 3...e5 when 4.dxe5 runs into 4...Nxe4.

Re: Is the Blackmar-Diemar Gambit playable?

Posted: Fri Jul 06, 2018 6:13 pm
by Dann Corbit
mclane wrote:
Fri Jul 06, 2018 10:09 am
how many chess engines (book turned off) you know play d4 d5 c4 ??

of course chess engines do to a certain degree sac for compensation.
but in a gambit, the idea is different. the compensation e.g. in the BDG is much deeper then the chess engines could smell.

its not about compensation for a material win. its about a compensation to mate the opponent.

the BDG is a good example. Kings gambit and queens gambit or englund gambit and budapest gambit
or any other gambit also show that the chess engines fail to emulate chess understanding.

and this behaviour has IMO not changed over long term of computer chess development.

the programs compute deeper and they kill humans. but they do not play better in gambits. its because they have no clue that the target is mating . instead they play for material wins.


lets hope the latest developments in computerchess help to solve this emulation lack.
My 64 core machine will play that using Stockfish at 40 moves in two hours.