The stronger the player, the more important it is that they know the opening well. The number of lines also increases tremendously. While beginners may get away with knowing just the basics, elite players have to analyze and memorize huge amounts of information.
There are two main types of opening preparation: having a narrow repertoire and a wide repertoire. In the first case you have a single response to the main moves by White (i.e. one opening against 1.e4, one against 1.d4, etc.). The pros of such approach are that you know your lines well and have a vast experience playing them. The cons: it’s very easy to prepare for a game against you; it’s hard to tune your tournament strategy in accordance with your performance. E.g. you may be playing a solid opening for Black, which doesn’t help to win a decisive game at all.
A wide repertoire implies having a few lines up your sleeve (e.g. being able to play the KID, Gruenfeld and the Nimzoindian against 1.d4). Pros: flexible tournament strategy; opponents will have a hard time preparing. Cons: not enough experience; necessity to memorize enormous amounts of lines. The third approach is a hybrid of the first two: when you have one (or more) well-analyzed openings and can also choose something different in special cases. Pros: surprise effect. Cons: if you decide to improvise and play an opening from scratch, your knowledge of the system will likely be superficial, thus increasing the probability of making a mistake (or several!).
When choosing an opening for the game we should try to predict our opponent’s intentions, consider the tournament situation and find breaches in his preparation. If you know a line that more or less perfectly suits your goals, you can go ahead and play it. However, in the real world it often seems that something is wrong about your prep. Then you have to think of a different way. As a case study, let’s consider my recent game against ex-Russian chess champion GM Sergey Volkov.
I had Black against him in round 4 of the Polugaevsky Memorial-2011. At this point I had 2.5/3, so my options weren’t limited to playing for a win only. By reviewing my opponent’s games, I discovered that against my main opening vs 1.d4, the Nimzovitch defense, he virtually always opts for a complicated line with 4.f3. He has a lot of experience treating those positions, while I can’t boast the same. However, most of his opponents preferred 4…d5 to “my” move – 4…0-0. At first I was toying with the idea of employing my previous favorite – the Benko gambit. Then I thought that I don’t play it too often myself, and the opponent will probably prepare against it anyway, so it won’t be much of a surprise factor. So, I decided to settle for the Nimzovitch defense and spent quite a lot of time analyzing an interesting line I found (later named by French GM Vladislav Tkachiev “novelty of the week in the world”). The advantage of this approach was that my opponent had never faced this idea before. Also, I reviewed the main variations thoroughly while preparing. The drawbacks were that in some variations a dangerous position could occur, and I wasn’t sure I would be able to memorize all the prep well enough. Anyway, I knew that most experienced chess players try to avoid main lines when caught off guard, since they are scared of having to play against an opponent who has prepared the variation with a chess engine. Therefore, the risk was justified.
As a result, the game proceeded just like I expected. My opponent deviated from a critical line, so I got a comfortable position. In fact, the opening was the main factor for a relatively successful outcome of the whole game. At some point my position was better, and I could even win a piece (although giving White a good compensation for it). Nonetheless, White didn’t risk losing this game too much.