Chess players love learning new openings. The usual process is to look it up online (or to buy one of the endless courses available) and then to study variations. This, though, has a relatively low success rate. Sure, if you are lucky enough to get your exact variation in game, great, but if not, you are left high and dry. Learning raw variations, by itself, rarely works.
2.h3. Now what do I do? |
The canned response is to say you need to "understand the moves" you are playing ... but what does that really mean? Sure, I'll learn the plans and ideas ... but which plans? What ideas? How do I know which plans are important, and how do I use those ideas to help me when I face 2.h3?
Here's my suggestion. Rather than this vague "understand the moves", you need to know three critical things:
- What do you do if your opponent does nothing? (eg, what is your dream scenario)
- What do you do if your opponent tries to kill you? (eg, what is the most aggressive option)
- What do you do against your opponent's main thematic plan? (eg, what is your opponent trying to do)
That's it. If you can answer these three questions, then you can play your opening against just about anything with total confidence. Conversely, if you don't know these, then you don't know your opening, regardless of how many moves you've memorized or Chessable reviews you've done. THIS is opening understanding.
Example: 1.e4 e5 as Black
First, White Does Nothing
There are two different types of "nothing", if that makes sense. First, there's "White plays near random moves like a lvl-1 bot" type of nothing, and second, there's "White makes normal developing moves but applies zero pressure" type of nothing. This second type is what I mean by "nothing". Something like this:
This is a fantasy position, but hopefully you see what I mean: White has played all the normal developing moves ... and put absolutely no pressure on Black. How should we respond?
Our absolute main idea is to play ...d5. Why? Because it fights for the center and gives us complete freedom of piece play. All of our pieces will be able to go to their best squares. If we can play that and White doesn't have a way to cause us problems, then we do it. This can happen in one of three ways.
First, if White plus super slow, we can prepare it with ...c6, like here:
1.e4 e5 2.Bc4 Nf6 3.d3 c6! |
That immediately challenges White's center and, according to Lichess, scores the best for Black. Makes sense, because it's following our main plan against slow play: get in ...d5! Here, we play it straight in the opening, but it is also our main middlegame idea as well. For example:
In this standard position, we have played ...Nd4 to free up our c-pawn, which can now advance and fight for the center. We might need to play ...Be6 and/or ...g5 to fully implement this idea, but it's our guiding thought: get in ...d5!
Second, we can prepare ...d5 tactically. This is typically through a timely ...Nxe4 shot, allowing a follow-up fork, something like this:
This is a very common motif. Black wins back his piece and gets the coveted ...d5 break in. This works anytime a Bishop is on c4 (quite common) and only a piece can recapture on e4 (somewhat common). Very useful to keep in the back pocket.
Third, we can play ...d5 without any preparation, tactical or otherwise. This can be risky, because White starts the game up a tempo and we run the risk of opening the game for White's pieces. This generally works best if White wastes time (eg, a super early h3 or a3) or if White does something that destabilizes his own center (generally an early c3):
1.e4 e5 2.c3 d5! |
Black is in a position recapture with his Queen and White has no Nc3 as a way to punish because of his c3-pawn. Black got his main pawn move in and can now do whatever he wants. Typically, that means developing our Knights to normal squares (c6 and f6), our Bishops to the most advanced diagonals (likely c5 and g4) and our heavy pieces to the center. We then try to apply pressure to White's weakest spot, likely f2 or the d-file, depending how White plays. Either way, opening over, huge success.
From a theoretical standpoint, nothing I've said here is groundbreaking. Practically speaking, though, it hugely simplifies our task. If White is doing normal developing, not making any threats, then the only thing we have to look for is ...d5. That's our only thought. If we can do it, great. If not, prepare it. Super simple. That's the value of opening understanding.
Second, White Tries to Kill Us
The most unpleasant part of playing 1.e4 e5 as Black is White's various gambit tries. White will typically do some combination of play d4, sacrifice a pawn, throw all his pieces at f7 and then mate us. When Black stumbles, the game is over in under 12 moves. You need to be prepared.
Here's our approach. First, we will always take the first pawn but refuse the second. Second, our primary response should be ...d5. Third, development is king. If we stay at least equal in development, we should be okay.
If you follow this approach, then most of White's killer tries get disarmed quickly. For example, the King's Gambit:
1.e4 e5 2.f4 exf4 3.Nf3 d5! |
1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.c3 d5! |
Notice how both of these follow the rules. We take the first pawn, but we ignore the second. Our focus instead is playing ...d5 immediately. This completely defuses these gambits and gives us good, equal play.
White has some gambits that stop us from playing ...d5 immediately. For example, the Scotch gambit:
1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Bc4!? |
White wins 63% of the time if we get greedy! |
What's the best scoring move for Black? ...Nf6. How do we remember that? Simple. It follows all our rules: we don't take the second pawn and we develop instead. It also prepares ...d5. Indeed, if after 5...Nf6! 6.e5, Black only has one good move, 6...d5!
1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Bc4!? Bc5 5.c3 Nf6! 6.e5 d5! |
This should not be hard, though, because of our guiding thought: play ...d5. In almost any position, that's what we want to do. Here it works perfectly. The game isn't over, of course, and it's not as if Black is better, but we have certainly navigated the early dangers.
The other gambit that stops us playing ...d5 immediately is the Evans Gambit. This is very dangerous, and it might be an example of where declining the gambit with ...Bb6 makes the most sense:
1.e4 e5 2.Nf3 Nf6 3.Bc4 Bc5 4.b4!? |
As we see, Black scores better by declining the pawn. It also has the advantage of not giving White what he likely wants, a wild open game. Nonetheless, to be consistent, we can still follow our guiding thought process: take the first pawn, but don't take the second. Funnily enough, it leads to the identical percentages!
By not accepting the second pawn, we keep more lines closed and give White less chance to attack us. Unfortunately, we have to play ...d6 instead of ...d5, and we have to be careful, because normal development with ...Nf6 is actual a trap.
Notice what happened here: in all other variations, our rules applied simply and got us a good position. Here, though, we have to be careful. Why? Because the Evans Gambit is really dangerous. It's critical. If you play 1.e4 e5, it's not enough to just know the plans and ideas in general, but you need to understand this specific position. The upshot? We know what we need to study: this. Almost everything else is already taken care of. This is the only thing we need to do a bit of work on.
In other words, plans and understanding can only take you so far. If you play things like the Sicilian Dragon or the Pirc, White's most dangerous plan IS to try to checkmate you as fast as possible. It's not enough to just know about it generally. You have to know it in depth (or at least, you need to know enough to avoid some common pitfalls). In contrast, with 1...e5, there's precious few positions you NEED to know, but this is definitely one of them.
Said another way, our opening understanding helps us understand what we need to spend time on. Evans gambit is serious, everything else here not so much.
Third, White's Most Thematic Plan
Perhaps just by coming this far you can already guess White's best and most thematic plan. The most dangerous ideas have revolved c3+d4 in some capacity. Black's best response is to play ...d5. So what does White want to do? Play c3+d4 while also stopping ...d5 from Black. There are two ways to do that: the Italian and the Spanish, 3.Bc4 (the Bishop controls d5) or 3.Bb5 (a potential pin on the Knight).
And yes, this is why the Italian and the Ruy Lopez / Spanish are the most common and most challenging moves to face. It's because it allows White's best plan while discouraging Black's best response. Nonetheless, our rules still follow.
There's too much nuance here for me to explain adequately in a short time, and I've spent nearly two hours on this post, so I'm going to cheat and simplify the analysis. Nevertheless, look at the following: this is one of the mainlines of the Italian:
What happened? White played his thematic c3 and d4. How did Black respond? According to all our rules! He took the first pawn, he played ...d5 as soon as it was possible, and he focused on development, not wasting time with ...h6 or anything else like that.
Or how about this, possibly the mainline of the Spanish:
What happened? White was all set to play c3 and d4 ... but Black has managed to play ...d5! And he has full development! Indeed, Black has three pieces developed and has castled, whereas White still has his entire Queenside undeveloped. True, playing this way costs Black a pawn, but his resulting play is so strong and easy that it doesn't matter. Black basically never loses in this position at the elite level, and he scores much better at the amateur level.
And I hope you understand why. It's because he's following all of our main plans and ideas. When you understand the main ideas, the moves practically play themselves.
Conclusion
If you want to play an opening, you need to know three things:
- What to do in general, when your opponent applies no pressure.
- What to do when your opponent comes straight for your head.
- What to do against your opponent's main thematic plan.
If you can answer all three questions, then you know your opening. Period. No questions. I would also say that it is all you need to know. You don't need to know anything else. Now, some of these questions might require a lot of analysis to fully answer. That is, it requires a lot of theory. That will depend on the opening. Some openings you can largely summarize in just a few rules. Other openings have different answers for every sub-variation. It all depends.
How do you learn this stuff? I wish I had a good answer here. I have never come across anything that organized its opening study this way. I've had to sift and read between the lines to form my own conclusions. Maybe that is a topic for another day. In the meantime, if you play an opening, I'd strongly suggest stopping and asking yourself these questions. How well you can answer them will determine how well you understand it.
I was recently shopping for a book that took this approach to openings, and Eugene Znosko-Borovsky's "How to Play the Chess Openings" seems to fit the bill. “Chess is a game of understanding and not of memory," says the author and promises not to overwhelm me with lines.
ReplyDeleteThanks for mentioning this; I have not come across this source before. It's quite an old book, so I imagine it can help you learn "how to learn" an opening, if that makes sense, but I'm not sure if it will help you learn a particular opening, at least as it is played today. That's an assumption, though, so I'll have to check it out myself.
DeleteOh! Right, I misunderstood. Z-B's book is definitely a "how to learn" kind of book and does not devote many pages to any particular opening.
DeleteSmithy. This is awesome. Together with your “treatise” on Chessable it is helping me reevaluate my chess. I have been filling Sielecki’s Keep it Simple 1.e4. It’s very thorough but I think it might be overkill for my level. I am already studying his KIS for black which I like (Caro and QGD fits my style, plus I guess it’s thematically correct, I.e. control the center and develop). However, the 1.e4 repertoire is a little more complicated—there is a different response to everything—French, Caro, Scandi, Sicilian, Alekhine etc. So taking your guidance I am looking at something more theme based. I like the Four Knights Scotch for 1.e5 but for everything else it seems like the Kings Indian Attack could be a good antidote. Again, thanks for your thoughtful blog posts.
ReplyDeleteThat's certainly possible, but I would caution about taking it to the extreme. In theory, you could play the KIA or the London against absolutely everything, but that's not very effective or fun (unless you absolutely LOVE the KIA). For me at least, the thrill of getting different types of structures and middlegames gives chess a lot of its charm. I naturally did, "After 1.e4 e5, X is my plan, and after 1.e4 e6, Y is my plan," etc, for all the most common openings I faced. It's more total plans, but it's still very manageable. Plus, because I came up with everything myself, it was much easier to remember. It wasn't what a book said, it was what _I_ said. I think that plays an important role in the retention process.
DeleteAnyway, there's nothing "wrong" with your approach, and it can be an excellent stopgap measure. Start with Scotch and the KIA, and over time, if you want, you can learn new systems. So maybe you play the Scotch, the Caro Exchange and then the KIA against everything. And then it becomes the Scotch, the Caro Exchange, the French Advance and the KIA against everything else, etc.
Lastly, just to be clear, my intention isn't so much "learn this system", but rather "learn a plan." A system you can play on auto-pilot; a plan is a bit more amorphous but tends to be long-term. Using the KIA example, the first four or so moves are basically a system, but the PLAN of eventually playing e4-e5 and then launching an attack with h4, Nd2-f1-h2-g4 and often a Bxh6 guides your whole middlegame. That's the real advantage, because a small amount of studying (learning a few plans) can apply for dozens of moves, whereas a system just gives you the first 5 moves and then you are on your own. (Incidentally, this is why I'm not a fan of the Qd8 Scandi: it's a very simple SYSTEM but without any obvious PLAN, which is why many people play the opening just fine but then drift aimlessly after that. Especially at lower levels, I'd argue the resulting plan is most important part of the opening.)
Yup. Brilliant. That is my thinking. I’ve watched some great videos by Daniel Naroditsky on the KIA and he basically said the same thing as you; use it as a rough framework but vary the move order based on what your opponent does.
Delete