Monday, April 6, 2009

Trusting Your Instincts

Actually, this post is about NOT trusting your instincts. If you're using the right data and doing the right calculations, you should generally trust the numbers, rather than your instincts. Most people find that very difficult to do, and if you're able to do it, then you're going to have an edge over your opponents. In the Sporting News Salary Cap Challenge, my calculations told me that Garret Atkins and Alexei Ramirez belong in my lineup. My instincts tell me that both are overrated players going into this season. However, the calculations are able to take things into account that are too subtle for my intuition to factor in - for example, the exact details of the scoring system for the contest format. I don't believe that my instincts are based on something that the calculations can't reflect (for example an injury, or reduced playing time), so I'll follow the numbers and gain an edge that other people might pass up.

Another example of that is in Draftbug, where my instincts told me that the four hitters who have clear advantages at their position today are Sizemore, Wright, V. Martinez, and Kinsler. The numbers confirm Sizemore and Wright, but show that Kinsler actually is a substantially worse pick for today than Pedroia (except that Pedroia may be rained out), and that under Draftbug's scoring system Martinez only has a very small edge on Mike Napoli. As usual, I'll trust the numbers, except where I know they're unaware of something important (like the weather report).


Crunch Bunch said...

I'm enjoying draftbug so far. Won the daily freeroll and finished 2nd in a 6 player game. Would have finished first if Peavy had outpitched Halladay.

As for day 2, I see one SP totally undervalued for some strange reason who will be on my teams.

I'll be interested to see how I do in draftbug since I don't use spreadsheets but I do use stats extensively (RHP/LHP, parks, player history, and other fantasy sites). In the long run, spreadsheets > my instincts I'm sure.

Alex said...

Cruch - I'm sure you'll do fine, since you're very knowledgeable, but the more you try to quantify things, the better you'll do. Check out my latest Hardball Times article for some more discussion of this topic.