Author Topic: Week 5: Chapter 9 - Rage Against the Machines (chess)  (Read 1889 times)  

Offline Eltanin Publishing

  • Status: Scheherazade
  • *****
  • Posts: 1266
  • Charlotte, VT
  • Cathy Ryan - Editor-in-Chief
    • View Profile
    • Eltanin Publishing
Week 5: Chapter 9 - Rage Against the Machines (chess)
« on: December 11, 2012, 01:02:13 pm »
This chapter is about chess. And yet, it isn't THAT much about chess. It seems to be more generally about what computers can do better than people and vice versa. I think that the overall point (in addition to pointing out what each does well) is that the best solution is often a human (or humans) working with a computer. If you know what a computer does well and what it doesn't, and take that into consideration, the computer can help you make better predictions (as the example of the two guys with a computer doing best in the online chess tournament). This is similar to the chapter on weather - the computers help a LOT (Silver calls this one of the few computer prediction success stories) - but the humans are able to add a bit more accuracy to the forecast that the computers spit out.

I play a lot of chess, and yet I didn't see this chapter as being THAT much about prediction, or being as relevant to Bayesian logic. Sure, you ask yourself many times, "What's the chances I'll win if I make this move?" or "If I make this move, what are the chances my opponent will respond in a certain way?" but it's not like poker where you can't see your opponent's cards (discussed in the next chapter), and you're trying to predict what s/he has. The only thing hidden is your opponent's motives. They might make a non-obvious move, and you have to try and figure out why they did that. To me, this seems like a different kind of prediction than trying to predict an opponent's poker hand, what the weather might be tomorrow, or whether flu season will be particularly bad.

Typically, I'd argue, there is no "noise" in chess. Moves aren't random* - they're done for a reason. Even if a player is flustered and doesn't know what to do, they'll do something that at least has a vague purpose, such as keeping everything safe. So I don't see how the whole "signal and the noise" theme fits into this chapter.  * I add the asterisk because apparently Deep Blue was programmed to make a nearly-random move (obviously it wouldn't do a move that gives up a piece to capture) if it couldn't decide what move to do. Kasparov thought too hard about WHY Deep Blue made that move, and came to the conclusion that the computer must be able to think REALLY far ahead and saw something he didn't.

So, am I missing something? Was there a stronger "prediction" or "signal and noise" theme in this chapter that I didn't see?

  • Advertisement
  • ***