Choosing the main lines that form the core of your opening repertoire is the part that will require the most work. After that, I would stick with analyzing your own games, and only worrying about opening issues that arise there. In many cases it's as simple as one player blundered or played an obviously weak move. In other cases, more work is required.
Until fairly recently, I was following the strategy of "analyze your games, find out where the game left theory, decide what you'd play next, and move on." I have recently come to realize that if your opponent played a "proper" move, it's important to follow the "what you'd play next" for some length afterwards. If the deviation occurs before development is complete, follow the line you'd play in the future until the approximate beginning of the middlegame. If you actually made it into the middlegame, extend the "trunk" out a distance so that you know where you want to take the variation. You need a view of the road ahead to provide context for the move. I find that if I only try to understand and remember the move I should have played, I often have to make the mistake several times before it really sinks in.
I don't consider myself to have a very good memory, but by following this procedure of using a board and determining a main line I'm finding myself able to recall variations and ideas much better than before. Just the other day, I reproduced a main line from the Italian game out to move 12, when previously I would have made an error around move 7. I had made no attempt to memorize that line beyond thinking about Black's moves and reading through Marin's explanatory text. This is the line, plus what I was thinking as I played it:
1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. c3 Nf6 5. d4 exd4 6. cxd4 Bb4+ 7. Bd2 Bxd2+ (Black's last two moves avoid losing a tempo. See the chapter "On Exchanging" in Nimzowitsch's My System) 8. Nbxd2 d5 (now I break up White's center!) 9. exd5 Nxd5 (and saddle him with an IQP! {isolated queen's pawn}) 10. Qb3 Na5 (forking queen and bishop) 11. Qa4+ Nc6 (threatens queen and offers a draw by repetition). My opponent finally played a move that avoided the repetition. After the game, I mapped out how I would respond against his move and left it at that.
This is a case where minimal opening study was involved. I checked out a minor branch from the trunk of my opening, which only took a few minutes.
Then there's the following case, where I faced a novelty in the Nimzo-Indian. It's not much of a novelty...I'm sure my opponent simply didn't understand what he was playing. However, by trying to answer "what would I play next" I found that I had my work cut out for me.
My opponent's "novelty" was: 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Qc2 d5 5.a3 Bxc3+ 6.bxc3, an odd blending of the Samisch (4.a3) and Classical (4.Qc2) lines. A point of 4.Qc2 is that the queen can capture on c3 and avoid doubled c-pawns. Therefore, this appears to be a Samisch where white has "wasted" time by playing Qc2.
I am in the very early stages of working out a Nimzo repertoire for black. Since not many of my opponents allow it as White, it hasn't rated as highly on my "things to do" list. I have some repertoire books as guides but I'm still in the process of deciding what my main lines will be.
First, as a starting point, let's look at the Samisch variation. After 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4, White challenges the Black bishop with 4.a3 Bxc3 5.bxc3. Quoting John Watson from Mastering the Chess Openings, Volume 2 (which I will be referring to for analysis):
The Samisch variation is in many ways the most instructive of all Nimzo-Indian lines. It seems odd to force Black into ...Bxc3+, a move he is likely to play anyway, and thus to accept the weak doubled c-pawns while losing time....Indeed, 4.a3 was one of the earliest methods of play versus the Nimzo-Indian and many of the best players of the time...were infatuated with possession of the bishop-pair....The Samisch Variation is the ideal starting point for discussing the Nimzo-Indian because it contains a majority of the fundamental themes that arise from the opening.
I have been using repertoire books by Dearing (
Play the Nimzo-Indian, my favorite), Emms (
Easy Guide to the Nimzo-Indian), and to a lesser extent Alburt et al (
Chess Openings for Black, Explained) for guidance. I have also used the general guide
Mastering the Nimzo-Indian (with the Read and Play Method) by Kosten, which is more concerned with pawn structures and typical middlegame plans than specific lines.
My tentative main line against the Samisch was that of Dearing and Emms. I will just give it without a move-by move explanation...the point here is that my repertoire did not feature an immediate ...d5:
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. a3 Bxc3+ 5. bxc3 c5 6. e3 O-O 7. Bd3 Nc6 8. Ne2 b6 9. e4 Ne8 10. O-O Ba6 11. f4 f5.
However, my opponent played the Classical move order with 4.Qc2. In that case, my main line is tentatively that of Dearing. Here also, I'll just give the main line without commentary, and point out that here an immediate ...d5
is played:
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 d5 5. cxd5 Qxd5 6. Nf3 Qf5 7. Qxf5 exf5 8.a3 Be7 9. Bf4 c6 10. e3 Be6 11. Bd3 Nbd7 .Now, back to the Samisch. In my game, it was as if my opponent had played the Samisch 4.a3 Bxc3+ 5.bxc3 d5 6.Qc2. However, none of my repertoires play 5...d5 against the Samisch! After the game, I tried to determine how one should play against this supposed inaccuracy by White. Referring to Watson and Kosten, I looked into Samisch lines that featured the ...d5 move lacking in my repertoire.
First, here is an example taken from Watson of inaccurate play on White's part by Botvinnik (!). In Botvinnik-Kotov, Groningen 1946, after
4. a3 Bxc3+ 5. bxc3 d5 6. cxd5 exd5 7. Bg5?!
7... c5! Watson says, "The problem is that after the natural
8. e3,
8...Qa5 forces some awkward defence like
9. Qc2 Ne4 10. Bf4 cxd4 11. exd4Nc6."
Now, let's try the same thing against my opponent's move order. After
4.Qc2 d5 5.a3 Bxc3+ 6.bxc3, Black is a tempo up if they can show that 4.Qc2 was a waste of time. So, let's make use of that free time by doing something useful, such as castling. However, after
6... O-O 7. cxd5 exd5 8. Bg5, we reach a position nearly identical to the previous one:
However, this may actually be respectable. We have transposed into a known position, according to my Chessbase. For example, if we continue as in Botvinnik-Kotov with
8...c5 9. e3, the queen already protects c3, and 9...Qa5 doesn't have the same punch as in Botvinnik-Kotov. For example, in Petrosian-Korchnoi, Moscow 1971, Black instead played
9... Nbd7.
What this shows is that, if both sides play stereotypically against the "waste of time" move Qc2, we find that it's not necessarily a waste of time at all! The refutation of this move order (if any) must lie elsewhere.
At this point, spending hours trying to prepare a line against a "goofy" novelty wouldn't be productive. However, in the course of trying to figure out what was going on in this position, I learned quite a bit... I've only included some of this work here. What becomes obvious is that
the general ideas that you find in Watson and Kosten (or Kmoch, for that matter) are of more use than repertoire books such as Dearing and Emms. For example, in these Samisch-like lines featuring ...d5 (which Watson calls the "Botvinnik approach"), we often arise at a pawn structure that Kosten calls the "light-squared blockade", as shown in the following "skeleton" position:
If Black can keep White's pawns on dark squares and exchange the light-squared bishops, White will lose the bishop pair and be stuck with a bad bishop. White is going to try and push e4.
Studying main lines not only ensures that you're playing sensible moves against your opponent's sensible moves...it also gives you experience in playing the typical structures and plans that arise in the opening. When your opponent plays something odd, you're going to have to rely on tactical awareness (yes, tactics are important!) plus your experience with and understanding of the position. In this particular case, I'm still not sure what the "proper" response to my opponent would have been, but I'm sure that the more I study and play these positions the better I'll be able to handle them.
In particular, the game made me question what I want to play against the Samisch. I'm attracted to this "light-squared blockade" pawn structure. If I like the move ...d5 against other variations, why not here? The short answer is: none of my repertoire books cover it! Therefore, I am going to say "heck with the books" and try to figure out my own ...d5 lines against the Samisch. If I had an established repertoire, I would not tear it all down and start again just because of one odd game. However, since I'm just starting out I can take the approach of "play the move you like the best" and build the repertoire around that. Remember, books are only there to guide you.
To summarize, when you analyze your games afterwards:
- If your opponent played a reasonable (book) response that you're unfamiliar with, work out what you would play in the future as a "main line" and then get on with life.
- If your opponent played an odd novelty, try to refute it. If a chess engine doesn't find an obvious tactical or gross positional problem with it, try to use your understanding of the opening and general principles to find a good response. In the process of finding the answer, you may learn quite a bit.
