Sunday, January 21, 2007

Playing with the formulas

Alex tried to create a little bit of a fuss over at his web site but it might be a little early for people to be paying attention or caring just yet. Since no one has stumped him yet, I'll take a little time to talk about the formulas in general.

Alex Patton has been writing about player valuation for over 20 years. If you really want to read the nuts and bolts of how he got started, try to find the old How to Win at Rotisserie Baseball book with the blue cover. It has a wonderful essay from Alex about all the trial and error he went through before he came up with the pricing system that he uses today.

To explain the basics of the system:

1) In the raw pricing model, hitters are worth $2184 and pitchers are worth $936 in a 12-team league that uses a $260 budget, or $182/$78 per team.
2) For years, once Alex started predicting, he was using a $175/$85 split on the bids. Recently, he's been using a $169/$91 model.
3) Hitting is worth more than pitching because there are three quantitative hitting categories in 4x4 versus only two quantatative pitching categories. Furthermore, there are 14 hitters times these three categories, leaving 42 "units of improvement". There are 9 pitchers times these two categories, leaving 18 "units of improvement". 42/60 = 70% of your money, or $182, to spend on hitting.

In theory, you should budget based on how your league budgets, which is why Alex started using $175/$85 and then, later, $169/$91. If you budget $2184 for hitting and your competitor budgets $2100, you might very well wind up buying a $200 offense, given the difference in your overall budgets.

This is the most basic of pricing theory. There's obviously more to it and, more importantly, there are obvious ways to play around with pricing that will help you do very well in your league year in and year out.

The most important thing to remember is to budget $3120 for your league if you're in a 12 team league with $260 team budgets. This seems so easy but I've seen people neglect to do it. Everything else tends to fall into place if you remember this simple fact.

1 comment:

Mike said...

You know, I should be (1) disqualified from guessing, and (2) more confident about my answer. Looking at the old masochists' notes (posted at askrotoman.com), and last year's version, the change seems to be in the wins denominator, which always used to be 2.5, is now 2.4. Not a huge adjustment, but Alex used to say it was the one constant. I would imagine it's due (essentially) to the move towards fewer innings pitched by starters, which has finally spread the wins around more and made them more valuable. Worth thinking about, but I can't really post it on Alex's site, can I?