Your network blocks the Lichess assets!

lichess.org
Donate

Visualizing the Italian Game: A Data-Driven Decision Tree Approach

@cuendillar said in #5:

I think you're really giving 3...Be7 short shrift here. It's a solid developing move that basically prevents white from entering sharp, critical lines as 4.Ng5 hangs a piece and 4.b4 (or 4.c3 + 5.d4) hits empty air. It also stops any Bg5-Nd5xNf6 attempts at opening up in front of blacks king. Yes, it's passive and black often get a bit less space without much in the way of tactical counterplay but he also doesn't really have any concrete weaknesses so it's nice if you just wanna play a relaxing game of chess - or come to the board tired after work/school/long previous round.

Apparently it's very rare in the database, which is kind of surprising to me.

@cuendillar You made me curious, so I actually checked the data – and you're absolutely right to call this out.

Here's what I found for 3...Be7:

  • Stockfish evaluation: +0.5 (only slightly better for White)
  • Stockfish ranking: 3rd best move for Black!
  • Games at 1000 ELO: virtually zero
  • Games at 1600-2000 ELO: virtually zero

So we have Stockfish's 3rd choice that almost nobody plays. That's remarkable!!!

Your positional arguments explain why it's objectively solid:

  • 4.Ng5 hangs a piece
  • 4.b4 and early d4 ideas lose their punch
  • Bg5-Nd5 plans are neutralized
  • Black gets a fortress with no real weaknesses

And according to me, the rarity is actually a PRACTICAL ADVANTAGE at club level: your opponents won't have preparation, won't know the typical plans, and might waste time figuring out what you're doing.

This is a perfect example of what data-driven analysis can miss: a move can be excellent precisely BECAUSE it's unpopular. The database reflects what people play, not what they should play.

So thanks for the insight – this is exactly the kind of discussion that makes chess fascinating!

@cuendillar said in #5: > I think you're really giving 3...Be7 short shrift here. It's a solid developing move that basically prevents white from entering sharp, critical lines as 4.Ng5 hangs a piece and 4.b4 (or 4.c3 + 5.d4) hits empty air. It also stops any Bg5-Nd5xNf6 attempts at opening up in front of blacks king. Yes, it's passive and black often get a bit less space without much in the way of tactical counterplay but he also doesn't really have any concrete weaknesses so it's nice if you just wanna play a relaxing game of chess - or come to the board tired after work/school/long previous round. > > Apparently it's very rare in the database, which is kind of surprising to me. @cuendillar You made me curious, so I actually checked the data – and you're absolutely right to call this out. Here's what I found for 3...Be7: - Stockfish evaluation: +0.5 (only slightly better for White) - Stockfish ranking: 3rd best move for Black! - Games at 1000 ELO: virtually zero - Games at 1600-2000 ELO: virtually zero So we have Stockfish's 3rd choice that almost nobody plays. That's remarkable!!! Your positional arguments explain why it's objectively solid: - 4.Ng5 hangs a piece - 4.b4 and early d4 ideas lose their punch - Bg5-Nd5 plans are neutralized - Black gets a fortress with no real weaknesses And according to me, the rarity is actually a PRACTICAL ADVANTAGE at club level: your opponents won't have preparation, won't know the typical plans, and might waste time figuring out what you're doing. This is a perfect example of what data-driven analysis can miss: a move can be excellent precisely BECAUSE it's unpopular. The database reflects what people play, not what they should play. So thanks for the insight – this is exactly the kind of discussion that makes chess fascinating!

@MyChessPosters I mean, I agree. e5 is better than the Alekhine and the Pirc, for obvious development issues.

Tho, I believe that most under 1000 players that pick e5, do it not because they are aware of the above statistics, but because to them it sounds obvious, that after e4, you play e5... That's literally elementary chess, and the first thing anyone (even your dad) would show you, when you are little and you start learning chess...

e5 is like a Tsunami... It is deceiving. From a far it looks like a regular wave... But from the movies you have seen that Tsunami are towering 50 meter high waves, so when you see a tsunami in the distance, you pay no attention to it... In reality what you should be doing is running for your life! But instead, you stay there unaware of what is happening...

The reason for all of this, is the weak focusing points, like f7 where white would sacrifice (for no reason sometimes), the knight on f6, that burn's black's time trying to develop it and white just kicks it away... pawn gambits like the scotch, or the Evans, the deceiving threat of 2.Nc3 (that leads to the Vienna) instead of 2.Nf3 that new players would not even fathom how much difference it makes (new players seem to think that 2.Nf3 or 2.Nc3 are pretty much the same...), all the early f4 stuff, all the Queens jumping on h5 and g4, and all the unsound sacrifices, that eventually find "compensation" after black being unable to defend against them...

Yes, the Sicilian is the best for new players, or the Caro Kann...

If all those players in this statistics, actually followed my rule, those statistics would quickly change for the better!

@MyChessPosters I mean, I agree. e5 is better than the Alekhine and the Pirc, for obvious development issues. Tho, I believe that most under 1000 players that pick e5, do it not because they are aware of the above statistics, but because to them it sounds obvious, that after e4, you play e5... That's literally elementary chess, and the first thing anyone (even your dad) would show you, when you are little and you start learning chess... e5 is like a Tsunami... It is deceiving. From a far it looks like a regular wave... But from the movies you have seen that Tsunami are towering 50 meter high waves, so when you see a tsunami in the distance, you pay no attention to it... In reality what you should be doing is running for your life! But instead, you stay there unaware of what is happening... The reason for all of this, is the weak focusing points, like f7 where white would sacrifice (for no reason sometimes), the knight on f6, that burn's black's time trying to develop it and white just kicks it away... pawn gambits like the scotch, or the Evans, the deceiving threat of 2.Nc3 (that leads to the Vienna) instead of 2.Nf3 that new players would not even fathom how much difference it makes (new players seem to think that 2.Nf3 or 2.Nc3 are pretty much the same...), all the early f4 stuff, all the Queens jumping on h5 and g4, and all the unsound sacrifices, that eventually find "compensation" after black being unable to defend against them... Yes, the Sicilian is the best for new players, or the Caro Kann... If all those players in this statistics, actually followed my rule, those statistics would quickly change for the better!

To me it sounds really odd, that the French scores better than the Caro Kann, for example, since to the untrained eyes, it looks like a better version of the French... But I guess most coaches still teach the French more than the Karo, for no particular reason? Maybe tradition?

Morphy in his time was obliterating players that dared to play e5 against him, because (judging from their games) they were unaware of the dangers!

As an example I will show you this Morphy game, that he beats someone anonymous with Knight odds. If we take each move one by one, Black didn't really do any obvious blunders... But he still got schooled in this miniature! If we take each black move individually (not a good thing to do, but this is how new players think), all of them are "sound" and natural looking, until the tsunami (in this case 18.h4!!) hits!

[Variant "From Position"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/R1BQKBNR w KQkq - 0 1"]

  1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. b4 Bxb4 5. c3 Bc5 6. d4 Bb6 7. O-O d6 8. dxe5 dxe5 9. Qb3 Qf6 10. Bb5 Be6 11. Qa4 Ne7 12. Bg5 Qg6 13. Bxe7 Kxe7 14. Bxc6 bxc6 15. Nxe5 Qf6 16. Nxc6+ Kf8 17. e5 Qg5 18. h4 Qg4 19. Qa3+ Kg8 20. Ne7+ Kf8 21. Ng6+ Kg8 22. Qf8+ Rxf8 23. Ne7#
To me it sounds really odd, that the French scores better than the Caro Kann, for example, since to the untrained eyes, it looks like a better version of the French... But I guess most coaches still teach the French more than the Karo, for no particular reason? Maybe tradition? Morphy in his time was obliterating players that dared to play e5 against him, because (judging from their games) they were unaware of the dangers! As an example I will show you this Morphy game, that he beats someone anonymous with Knight odds. If we take each move one by one, Black didn't really do any obvious blunders... But he still got schooled in this miniature! If we take each black move individually (not a good thing to do, but this is how new players think), all of them are "sound" and natural looking, until the tsunami (in this case 18.h4!!) hits! [Variant "From Position"] [FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/R1BQKBNR w KQkq - 0 1"] 1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. b4 Bxb4 5. c3 Bc5 6. d4 Bb6 7. O-O d6 8. dxe5 dxe5 9. Qb3 Qf6 10. Bb5 Be6 11. Qa4 Ne7 12. Bg5 Qg6 13. Bxe7 Kxe7 14. Bxc6 bxc6 15. Nxe5 Qf6 16. Nxc6+ Kf8 17. e5 Qg5 18. h4 Qg4 19. Qa3+ Kg8 20. Ne7+ Kf8 21. Ng6+ Kg8 22. Qf8+ Rxf8 23. Ne7#

I play Giuoco piano because I have lost the only game I played two knights as black

I play Giuoco piano because I have lost the only game I played two knights as black

@MyChessPosters said in #11:

This is a perfect example of what data-driven analysis can miss: a move can be excellent precisely BECAUSE it's unpopular. The database reflects what people play, not what they should play.

For tournament players, unfortunately this is where the mind games begin! Tournament players need to understand what results they intend to play for...

Some openings are more solid and offer less winning chances.
Some openings offer better winning or drawing chances but are harder to learn.
Some openings have surprise value which offer better chances against unprepared opponents but otherwise are not as good.

3... Be7 may be easier to learn, but in exchange you'll get difficult positions unless catching your opponent by surprise. To me it feels like playing a French defense but also giving up control of the center, so it's risky.

@MyChessPosters said in #11: > This is a perfect example of what data-driven analysis can miss: a move can be excellent precisely BECAUSE it's unpopular. The database reflects what people play, not what they should play. For tournament players, unfortunately this is where the mind games begin! Tournament players need to understand what results they intend to play for... Some openings are more solid and offer less winning chances. Some openings offer better winning or drawing chances but are harder to learn. Some openings have surprise value which offer better chances against unprepared opponents but otherwise are not as good. 3... Be7 may be easier to learn, but in exchange you'll get difficult positions unless catching your opponent by surprise. To me it feels like playing a French defense but also giving up control of the center, so it's risky.

@T_deus_ex_T interesting tsunami metaphor :)!

Your list of e5 traps is real. But doesn't the Sicilian have its own tsunami? Smith-Morra, Alapin, Grand Prix Attack, Bc4+Qb3 cheapos...

My hypothesis: the Sicilian's 50% score at 1000 ELO might be SELECTION BIAS. Players who choose 1...c5 are probably more serious students who've done homework. The 1...e5 crowd includes everyone playing "the obvious move."

So whatever you play, prepare for the common traps at your level. Different opening = different tsunami, not no tsunami!

@T_deus_ex_T interesting tsunami metaphor :)! Your list of e5 traps is real. But doesn't the Sicilian have its own tsunami? Smith-Morra, Alapin, Grand Prix Attack, Bc4+Qb3 cheapos... My hypothesis: the Sicilian's 50% score at 1000 ELO might be SELECTION BIAS. Players who choose 1...c5 are probably more serious students who've done homework. The 1...e5 crowd includes everyone playing "the obvious move." So whatever you play, prepare for the common traps at your level. Different opening = different tsunami, not no tsunami!

@DevIsTheBest said in #14:

I play Giuoco piano because I have lost the only game I played two knights as black

@DevIsTheBest Ha ha! That's the most honest opening selection process I've ever heard. "I lost once, never again." ^^

The good news: the data backs your instinct. Giuoco Piano is more forgiving – fewer tactical landmines, more strategic play. Two Knights requires knowing EXACTLY how to handle 4.Ng5!

Smart adaptation!

@DevIsTheBest said in #14: > I play Giuoco piano because I have lost the only game I played two knights as black @DevIsTheBest Ha ha! That's the most honest opening selection process I've ever heard. "I lost once, never again." ^^ The good news: the data backs your instinct. Giuoco Piano is more forgiving – fewer tactical landmines, more strategic play. Two Knights requires knowing EXACTLY how to handle 4.Ng5! Smart adaptation!