Your network blocks the Lichess assets!

lichess.org
Donate

You’re studying chess openings the wrong way (do this instead)

#1 Reach middlegames you’re happy to play, and
#2 Learn something from every game.

This is good clarity. But, winning the game and learning from the game are not the same -- and often somewhat at odds.

The more you are happy to play the position, the less you may gain from the experience (and I am speaking as someone who contentedly plays some of the same openings now for 10 or 15 years of blitz :-).

If the goal were to build a better chess player, we should take an aggressive "tilt" to this recipe -- like 80% learning, and 20% comfort (=winning right now).

#1 Reach middlegames you’re happy to play, and #2 Learn something from every game. This is good clarity. But, winning the game and learning from the game are not the same -- and often somewhat at odds. The more you are happy to play the position, the less you may gain from the experience (and I am speaking as someone who contentedly plays some of the same openings now for 10 or 15 years of blitz :-). If the goal were to build a better chess player, we should take an aggressive "tilt" to this recipe -- like 80% learning, and 20% comfort (=winning right now).

@quacktagon said ^

It’s nice to see more and more people pushing the same idea that openings are often overstudied. But I still think there’s an open question here: where do you actually draw the line? When does “enough opening knowledge” end?

As a Sicilian player, for example, there are so many sharp and aggressive lines where you need very precise defense. So how far do you really need to know it? Is it enough to understand the key idea and the best defensive move, or do you need to know the line until things simplify and the attack is clearly over?

That’s where I still see the difficulty. In some variations, especially the chaotic ones, it feels like you simply need concrete knowledge. Otherwise you’re just lost.

I think you should stop studying once you are comfortable and can play on your own. For example, if there is a sharp line in the Siclillian, you should stop studying once the chances of making a mistake are low.

@quacktagon said [^](/forum/redirect/post/tYLaIFbl) > It’s nice to see more and more people pushing the same idea that openings are often overstudied. But I still think there’s an open question here: where do you actually draw the line? When does “enough opening knowledge” end? > > As a Sicilian player, for example, there are so many sharp and aggressive lines where you need very precise defense. So how far do you really need to know it? Is it enough to understand the key idea and the best defensive move, or do you need to know the line until things simplify and the attack is clearly over? > > That’s where I still see the difficulty. In some variations, especially the chaotic ones, it feels like you simply need concrete knowledge. Otherwise you’re just lost. I think you should stop studying once you are comfortable and can play on your own. For example, if there is a sharp line in the Siclillian, you should stop studying once the chances of making a mistake are low.

Interesting article but I’m disappointed it was written with AI... surely you could write it yourself. I’d rather read a less “well” written article that uses an authentic voice than any AI written work.

Interesting article but I’m disappointed it was written with AI... surely you could write it yourself. I’d rather read a less “well” written article that uses an authentic voice than any AI written work.
<Comment deleted by user>

@Kleopatra2019 said ^

As usual, I deepy enjoy your writings

Thanks for reading

@Kleopatra2019 said [^](/forum/redirect/post/fz7lT0AO) > As usual, I deepy enjoy your writings Thanks for reading

@irresoluble said ^

This article is quite well done and will prove of value to players from beginner to (at least) 21-2200 FIDE, my playing strength in my best days.

Glad you found it worthwhile!

@irresoluble said [^](/forum/redirect/post/aE4IL1VH) > This article is quite well done and will prove of value to players from beginner to (at least) 21-2200 FIDE, my playing strength in my best days. Glad you found it worthwhile!

@liYichen said ^

Interesting article but I’m disappointed it was written with AI... surely you could write it yourself. I’d rather read a less “well” written article that uses an authentic voice than any AI written work.

Thanks for the comment - I can see there were certain lines that take on the 'not X, Y' kind of phrasing that's seen a lot these days with AI writing. I've refined the post so those bits should sound more natural as well.

@liYichen said [^](/forum/redirect/post/90BpvqTH) > Interesting article but I’m disappointed it was written with AI... surely you could write it yourself. I’d rather read a less “well” written article that uses an authentic voice than any AI written work. Thanks for the comment - I can see there were certain lines that take on the 'not X, Y' kind of phrasing that's seen a lot these days with AI writing. I've refined the post so those bits should sound more natural as well.

this is so useful and thanks for sharing

this is so useful and thanks for sharing

@marcoen85 said ^

So I tend to look at a lot of model games and stem games. In each particular line to get to GM level you probably need at least five games and analyse them very very well.

Model or stem games are mentioned everywhere (e.g. by Andras Toth) but no one really explains where you should find them and what makes a good stem game. Take, for example, the Two Knights defence, Jan Piński has a book on this opening from 2004. It contains a lot of annotated games. Are any five of them good for use as model games? Maybe none of them are as the annotations are over 20 years old and possible completely false today.

Can you even take those stem games from books or do you have to pick five random games in your opening of choice and fully annotate them yourself?

Yep finding good ones are not easy. Doing this the old-fashioned way in ChessBase is one way. You filter by position, filter by year, look at the older games where notable players are playing it, stem game can point to historical games like those - databases like CB also has annotated games which help there. So with your example, I'd say they're still good for model games even if 20 years old if they contain typical plans and ideas from both sides that you can understand and analyse yourself - in fact, older games can be better for this as looking at modern top games, it's harder to see ideas clearly because the players can forestall them. Keep in mind having stem games and multiple model games is more a recommendation for those 2000+ FIDE and aiming for higher, but the idea is useful for intermediate level and above - understanding typical piece placements and what both sides are aiming for in the middlegame so you can use that in your own games even if the exact position is not the same.

ChessBase is probably behind the times a bit though (though for annotated games and filtering it's strong, probably are alternatives now), and to look for model players who've played multiple games with a particular line you want to play, one site is qchess they have a tool https://qchess.net/modelgames and probably with stem games and finding annotated game collections places like chessgames.com are possible too.

So for stem games, the oldest ones with notable players - model games, it's not finding the perfect ones but analysing some in depth (guess-the-move is good here too) in a way that when you play it in serious games e.g. OTB classical, you already have understanding of typical piece placements and plans that help you in various middlegames rather than specific moves or variations memorised.

@marcoen85 said [^](/forum/redirect/post/nRsD1MT2) > > So I tend to look at a lot of model games and stem games. In each particular line to get to GM level you probably need at least five games and analyse them very very well. > > Model or stem games are mentioned everywhere (e.g. by Andras Toth) but no one really explains where you should find them and what makes a good stem game. Take, for example, the Two Knights defence, Jan Piński has a book on this opening from 2004. It contains a lot of annotated games. Are any five of them good for use as model games? Maybe none of them are as the annotations are over 20 years old and possible completely false today. > > Can you even take those stem games from books or do you have to pick five random games in your opening of choice and fully annotate them yourself? Yep finding good ones are not easy. Doing this the old-fashioned way in ChessBase is one way. You filter by position, filter by year, look at the older games where notable players are playing it, stem game can point to historical games like those - databases like CB also has annotated games which help there. So with your example, I'd say they're still good for model games even if 20 years old if they contain typical plans and ideas from both sides that you can understand and analyse yourself - in fact, older games can be better for this as looking at modern top games, it's harder to see ideas clearly because the players can forestall them. Keep in mind having stem games and multiple model games is more a recommendation for those 2000+ FIDE and aiming for higher, but the idea is useful for intermediate level and above - understanding typical piece placements and what both sides are aiming for in the middlegame so you can use that in your own games even if the exact position is not the same. ChessBase is probably behind the times a bit though (though for annotated games and filtering it's strong, probably are alternatives now), and to look for model players who've played multiple games with a particular line you want to play, one site is qchess they have a tool https://qchess.net/modelgames and probably with stem games and finding annotated game collections places like chessgames.com are possible too. So for stem games, the oldest ones with notable players - model games, it's not finding the perfect ones but analysing some in depth (guess-the-move is good here too) in a way that when you play it in serious games e.g. OTB classical, you already have understanding of typical piece placements and plans that help you in various middlegames rather than specific moves or variations memorised.

@quacktagon said ^

It’s nice to see more and more people pushing the same idea that openings are often overstudied. But I still think there’s an open question here: where do you actually draw the line? When does “enough opening knowledge” end?

As a Sicilian player, for example, there are so many sharp and aggressive lines where you need very precise defense. So how far do you really need to know it? Is it enough to understand the key idea and the best defensive move, or do you need to know the line until things simplify and the attack is clearly over?

That’s where I still see the difficulty. In some variations, especially the chaotic ones, it feels like you simply need concrete knowledge. Otherwise you’re just lost.

I think the other aspect is, you learn the most from actually playing it, the more serious and long time controls the better because then you and your opponent are both trying to play the best moves from the opening and you learn from playing out the middlegame. Too many people try to memorise and 'know' things before even playing something multiple times. Quantity is also good in terms of playing it because you learn from that and then you can compare to what's popular and then you can compare your thoughts and impressions during games with what's commonly played at your level & higher and try to understand each move, much easier to get into when you've experienced the positions yourself.

So for me the 'knowledge' should come from playing it first, studying it from other places is always secondary knowledge, they help the most when you combine it with your primary knowledge and experience where you deepen your understanding.

But yes, with tactical lines and depending on your playing level of course you'll have to know more before playing, so good to test online or against computer/training partners ideally before competition. Sort of a necessary thing if you're playing sharp lines, but even then probably in general people are trying to book up too much and might be better to use that time to work on other parts of the game where it'll help regardless of opening. I was put off some lines like Najdorf because the Bg5 lines require knowing concrete lines to survive, as you mention!

@quacktagon said [^](/forum/redirect/post/tYLaIFbl) > It’s nice to see more and more people pushing the same idea that openings are often overstudied. But I still think there’s an open question here: where do you actually draw the line? When does “enough opening knowledge” end? > > As a Sicilian player, for example, there are so many sharp and aggressive lines where you need very precise defense. So how far do you really need to know it? Is it enough to understand the key idea and the best defensive move, or do you need to know the line until things simplify and the attack is clearly over? > > That’s where I still see the difficulty. In some variations, especially the chaotic ones, it feels like you simply need concrete knowledge. Otherwise you’re just lost. I think the other aspect is, you learn the most from actually playing it, the more serious and long time controls the better because then you and your opponent are both trying to play the best moves from the opening and you learn from playing out the middlegame. Too many people try to memorise and 'know' things before even playing something multiple times. Quantity is also good in terms of playing it because you learn from that and then you can compare to what's popular and then you can compare your thoughts and impressions during games with what's commonly played at your level & higher and try to understand each move, much easier to get into when you've experienced the positions yourself. So for me the 'knowledge' should come from playing it first, studying it from other places is always secondary knowledge, they help the most when you combine it with your primary knowledge and experience where you deepen your understanding. But yes, with tactical lines and depending on your playing level of course you'll have to know more before playing, so good to test online or against computer/training partners ideally before competition. Sort of a necessary thing if you're playing sharp lines, but even then probably in general people are trying to book up too much and might be better to use that time to work on other parts of the game where it'll help regardless of opening. I was put off some lines like Najdorf because the Bg5 lines require knowing concrete lines to survive, as you mention!