For better or for worse, the opening tends to be the aspect of the game that gets the most attention. More books have been written about that phase than the middlegame and endgame combined. This importance, though, is probably overstated at lower levels, but it becomes more important the higher up the ladder you go.

My repertoire against 1.d4 had long been based around the Semi-Slav, using the "Triangle" move order: 1.d4 d5 2.c4 e6 3.Nc3 c6 (the triangle) 4.Nf3 Nf6 (the Semi-Slav). This gets around the problem of 1.c4, as while 1...d5 isn't playable, all I'm trying to do is set up the Triangle, and so 1.c4 e6 2.Nf3 d5 3.d4 c6 gets me to the same kind of setup.

However, after a long break from chess from 2002 to 2006, I had missed out on a long stretch of theory in the Moscow Gambit Variation of the Semi-Slav (1.d4 d5 2.c4 e6 3.Nc3 c6 4.Nf3 Nf6 5.Bg5 h6 6.Bh4 dxc4 7.e4 g5 8.Bg3 b5). My score in this line hadn't been great to begin with, and here is one of my bigger losses in this line. This was against then-IM Abhijeet Gupta, who finished his GM title soon after this event and went on to win the World Junior Championship later that year.

I figured it wasn't worth the headache for a non-professional player, so I decided to start playing the Slav Defense. The Slav (1.d4 d5 2.c4 c6) is generally a much safer defense, and is less theoretically open to wild changes from game to game. The "problem" with the Slav is that against 1.c4, you have to be willing to play a very different sort of position than the Slav. That's because the Slav response to 1.c4 is with 1...c6, aiming for 2.d4 d5, but instead of 2.d4, White can play 2.e4. Then, after 2...d5 3.exd5 cxd5 4.d4 (4.cxd5 is also played, but isn't so dangerous in my view), we've transposed into the Panov-Botvinnik variation of the Caro-Kann. Since I don't play the Caro Kann against 1.e4, this would mean I'd have to be prepared in a completely different opening system as well.

Normally, when I want to introduce a whole new opening into my repertoire, I like to test it out prior to any serious games. Sometimes that means playing slow games against strong players, but as I've found it tough to find willing people, I tend to try out my new openings on ICC in blitz games.

Before I've fully developed my lines, the blitz games can give me a feeling of whether I feel comfortable in the resulting middlegames, what lines might give me trouble, what some typical plans are, etc. Once I've picked all my lines out, I can then use the blitz games as a test of whether they can hold up and whether I can remember the lines!

In the following game, from the last round of the Montreal International, I had my first true Panov-Botvinnik as black in a slow game. I had already faced this kind of setup twice before (GMs Ehlvest and Kacheishvili played it against me, but not with the main Panov lines in mind), but GM Arkadij Naiditsch (2697 FIDE) was the first to enter the main lines against me. The result was not pretty.

What would you play as White after 15...Qd8?

What would you play as Black after 20.h5?

One loss won't turn me away from the Slav and Panov Botvinnik, but the way I lost this game means that I need to go back and do my homework a little more thoroughly next time. Even though Karpov said it, it's back to "trust but verify."