Getting good by losing

"You will have to lose hundreds of games before becoming a good player," said Capablanca, and this bit of wisdom is accepted by the chess world without much objection. I remember hearing Yasser Seirawan telling a story about being an instructor at a chess camp for kids where he asked the assembled group of children if any of them had lost 10,000 games in the course of their life. One child raised his hand and Seirawan said something like "Young man, you may be on your way to becoming a grandmaster."

I know of no other competitive activity where losses are considered such valuable training. In athletic endeavors such as tennis, swimming, running, or boxing we wouldn't think that a child who was losing regularly was a promising talent; instead we would look to the child who was always winning. Chess is different. If you want to be good, then you have to play against people who are better than you. If you do this regularly you lose a lot.

The losses themselves aren't what make you better; what makes you better is what you learn from them. If all it took to get good was losing lots of games, we would all be grandmasters. If you can learn from your losses, then you start winning and you have to find stronger opponents if you want to continue to improve. When you find stronger opponents you start losing again, and the cycle repeats.

If you could learn as much from games you won as you can from games you lost we wouldn't have to go through all this. Talented players would win their games against weaker opponents, learn from those games and get stronger and take on progressively stronger opponents without having to suffer through many losses. But that's not how it works. Strategies and tactics that work against weaker players don't work against stronger players, so you have to discover new ideas against stronger players and one of the ways you do that is trial and error. You play many games and try many ideas, and most of them won't work and you lose. You study those losses and figure out why. Then you try again until you get it right.

Eventually you hit a plateau, which is what happens when you reach a level where you can't identify, or at least can't correct, whatever it is you're doing wrong. We've discussed the plateau in the blog before; if you want to review that you can see the post here. Sometimes plateaus can be overcome but eventually everyone reaches a plateau they can't get over, and that's the best they will ever be at chess. Many people quit the game if they think they've reached this point, but since you can never know for sure most of us keep plugging away, losing a lot in hopes that we're learning and getting better even if it doesn't show up in the results right away.

There's a balance to be struck. If you lose every game it's not much fun, but if you win every game you're probably not learning and getting better. If you just don't care about getting better or if you're already World Champion I guess that's fine. For the rest of us, we have chosen the way of pain. We win often enough to keep us playing, and lose often enough to learn.