4 ms·

It's still a systematic "issue"... If it is virtually impossible to find a great or optimal move (even given enough time) without having to memorize tons of spe

It's still a systematic "issue"... If it is virtually impossible to find a great or optimal move (even given enough time) without having to memorize tons of specific situations first that's kinda annoying. It's a similar thing with competitive programming. At first the problems are solvable by thinking, but at some point you just need to memorize (or even copy paste) a solution that someone else figured out first... Boring.

smatija 3y ago
You can say this about any human endeavour - to learn language, you need to learn idioms. To write, you need tropes, to program you need patterns, etc. Every sport has its plays. I think learning games of past masters in chess is akin to learning culture when studying language. Standing on the shoulders of the giants and all that jazz. Especially since learning openings is not pure memorization of moves - its about learning ideas, stories, patterns. For example: Najdorf variation of sicilian begins with 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 [0]. Idea behind 5. ... a6 is that black wants to play e5+d5 to gain full center, but 5. ... e5 right now losses due to 6. Bb5+. So black prevents it with 5. ... a6. Now white has a lot of choices - 6. Bg5 is very concrete, you need some memorization, 6. Bc4 is sharp but a bit more rational, 6. Be3 is very popular today, ... All of them have some benefits and drawbacks. I play 6. Be2, since it gives white some pressure, while having very classical idea-based play and black players often lack preparation for it. Black can now play 6. ... e6 (transposing to Scheveningen sicilian) or 6. ... e5, staying in the Najdorf waters. Lets look at 6. ... e5. White plays 7. Nb3, 7. Nf3?! is worse since white will soon want to play f4 or f3 and knight would block that pawn. Now is the interesting part: the only good move in this position is 7. ... Be7 and it's hard to see why. A lot of my online games are already decided in next few moves. Point is, that 7. ... Nbd7 (with idea of developing light squared bishop on b7) interferes with black control of d5, so white can gain bind on d5 via playing a4 (preventing b5) and then either Bc4 or Nd2-Nc4-Ne3 takes full control of d5. This is crucial, since if black is unable to achieve d5 pawn break he has nothing to play for in these structures, so game is totally in white hands. Secondly 7. ... Be6 fails very concretely to 8. f4 Qc7 9. g4 exf4 10. g5!, which was pioneered by Julio Kaplan, a former junior chess world champion and software chess pioneer. We get to a very open position where black king is stuck in the center, so white gets a very pleasant attack for free. I hope this gives you a bit of a counterpoint towards why learning theory is not only annoying. [0] - in algebraic notation, for those interested in learning it: https://en.wikipedia.org/wiki/Algebraic_notation_(chess) https://en.wikipedia.org/wiki/Algebraic_notation_(chess)
- gjm11 3y ago
  In that last "very concrete" line, the 'fish seems to think 8 ... Qc7 is a mistake, and after e.g. 8 ... Be7 instead it puts W at just +0.2, which is perfectly respectable but not obviously better for W than after 7 ... Be7. If you try the same g4; fg g5 idea, then it thinks B is actually slightly better after either ... Nfd7, or even better ... f3; Bxf3 Nfd7. I'm a patzer so don't claim to know what the key differences are, but one is that in the variation Stockfish prefers the white pawn on g5 is hanging so he has to either spend a move playing something like h4 or else sacrifice it.
  - smatija 3y ago
    There is another finesse: the point of f4 is that White is threatening f5, attacking the bishop. Black then has a choice between retreating (which losses time) or Bxb3 (which gives White bishop pair). The point of Qc7 is that black can answer f5 with Bc4 - offering trade of bishops instead, which is a bit disappointing for White. This is why this resource with g4 is important. So if Black plays Qc7 White needs to know about this g4 temporary pawn sac. If Black plays exf4 then White recaptures with Bxf4, gaining a nice advantage in development. Finally if Black plays anything else (like Be7) White gains comfortable advantage with f5.
    - gjm11 3y ago
      Stockfish isn't convinced that f4 Be7; f5 Bd7 does give W a comfortable advantage; it reckons +0.2, same as if B plays ... Be7 after Nb3 as you recommend. Again, I'm a patzer so if your intuition/calculation says W is much better I don't have any refutation other than "computer says no". W's position sure looks more fun to play, with prospects of a kingside attack, but that's often true in the Sicilian even when things are objectively fairly balanced. If I let it sit for a while, it vacillates between wanting to play g4, a3, Be3, h4, etc., and a quieter approach with Bg5xf6, Qd3, Nd2. At depth 46 it prefers the latter but only gives it +0.1 even though its PV shows W winning the d-pawn. Aaaand at depth 48 it's switched back to Operation Pawn Storm, though it wants to prepare with a3 and Bf3 before playing g4 and h4. And back to the piece manoeuvres and winning the d-pawn at depth 50. Still only +0.1, though. (But computers are awfully good defenders, and maybe this is one of those positions that's much better for the attacking side when humans rather than computers are playing.)

smatija 3y ago

You can say this about any human endeavour - to learn language, you need to learn idioms. To write, you need tropes, to program you need patterns, etc. Every sport has its plays. I think learning games of past masters in chess is akin to learning culture when studying language. Standing on the shoulders of the giants and all that jazz. Especially since learning openings is not pure memorization of moves - its about learning ideas, stories, patterns. For example: Najdorf variation of sicilian begins with 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 [0]. Idea behind 5. ... a6 is that black wants to play e5+d5 to gain full center, but 5. ... e5 right now losses due to 6. Bb5+. So black prevents it with 5. ... a6. Now white has a lot of choices - 6. Bg5 is very concrete, you need some memorization, 6. Bc4 is sharp but a bit more rational, 6. Be3 is very popular today, ... All of them have some benefits and drawbacks. I play 6. Be2, since it gives white some pressure, while having very classical idea-based play and black players often lack preparation for it. Black can now play 6. ... e6 (transposing to Scheveningen sicilian) or 6. ... e5, staying in the Najdorf waters. Lets look at 6. ... e5. White plays 7. Nb3, 7. Nf3?! is worse since white will soon want to play f4 or f3 and knight would block that pawn. Now is the interesting part: the only good move in this position is 7. ... Be7 and it's hard to see why. A lot of my online games are already decided in next few moves. Point is, that 7. ... Nbd7 (with idea of developing light squared bishop on b7) interferes with black control of d5, so white can gain bind on d5 via playing a4 (preventing b5) and then either Bc4 or Nd2-Nc4-Ne3 takes full control of d5. This is crucial, since if black is unable to achieve d5 pawn break he has nothing to play for in these structures, so game is totally in white hands. Secondly 7. ... Be6 fails very concretely to 8. f4 Qc7 9. g4 exf4 10. g5!, which was pioneered by Julio Kaplan, a former junior chess world champion and software chess pioneer. We get to a very open position where black king is stuck in the center, so white gets a very pleasant attack for free. I hope this gives you a bit of a counterpoint towards why learning theory is not only annoying. [0] - in algebraic notation, for those interested in learning it: https://en.wikipedia.org/wiki/Algebraic_notation_(chess) https://en.wikipedia.org/wiki/Algebraic_notation_(chess)

gjm11 3y ago
In that last "very concrete" line, the 'fish seems to think 8 ... Qc7 is a mistake, and after e.g. 8 ... Be7 instead it puts W at just +0.2, which is perfectly respectable but not obviously better for W than after 7 ... Be7. If you try the same g4; fg g5 idea, then it thinks B is actually slightly better after either ... Nfd7, or even better ... f3; Bxf3 Nfd7. I'm a patzer so don't claim to know what the key differences are, but one is that in the variation Stockfish prefers the white pawn on g5 is hanging so he has to either spend a move playing something like h4 or else sacrifice it.
- smatija 3y ago
  There is another finesse: the point of f4 is that White is threatening f5, attacking the bishop. Black then has a choice between retreating (which losses time) or Bxb3 (which gives White bishop pair). The point of Qc7 is that black can answer f5 with Bc4 - offering trade of bishops instead, which is a bit disappointing for White. This is why this resource with g4 is important. So if Black plays Qc7 White needs to know about this g4 temporary pawn sac. If Black plays exf4 then White recaptures with Bxf4, gaining a nice advantage in development. Finally if Black plays anything else (like Be7) White gains comfortable advantage with f5.
  - gjm11 3y ago
    Stockfish isn't convinced that f4 Be7; f5 Bd7 does give W a comfortable advantage; it reckons +0.2, same as if B plays ... Be7 after Nb3 as you recommend. Again, I'm a patzer so if your intuition/calculation says W is much better I don't have any refutation other than "computer says no". W's position sure looks more fun to play, with prospects of a kingside attack, but that's often true in the Sicilian even when things are objectively fairly balanced. If I let it sit for a while, it vacillates between wanting to play g4, a3, Be3, h4, etc., and a quieter approach with Bg5xf6, Qd3, Nd2. At depth 46 it prefers the latter but only gives it +0.1 even though its PV shows W winning the d-pawn. Aaaand at depth 48 it's switched back to Operation Pawn Storm, though it wants to prepare with a3 and Bf3 before playing g4 and h4. And back to the piece manoeuvres and winning the d-pawn at depth 50. Still only +0.1, though. (But computers are awfully good defenders, and maybe this is one of those positions that's much better for the attacking side when humans rather than computers are playing.)