2011 CAIRO Projections v.0.1 Now Available
A few weeks ago, I began constructing my own forecasting system to help project players’ future performances in order to estimate 1) their true talent level, and 2) make an educated guess as to how they will perform in the next season. I had a few ideas in mind as to what I wanted to do- use about five years of data (weighted, with the most recent season weighted the heaviest), regressed to the population mean, along with component aging factors and component park factors.
Sure enough, somebody has done this already. SG of The Replacement Level Yankees Weblog has released his CAIRO forecasts (version 1.0) for the 2011 season. CAIRO uses four years, however, but it also uses Major League Equivalencies to translate Minor League performance in an attempt to estimate the player’s likely performance at the Major League level. I guess you could say that this is the type of system I wanted to put together- luckily for me, this means I won’t have to do any of the heavy lifting myself!
By the way, I’d like to mention one huge pet peeve of mine before sending you to the link. It’s pretty common to see responses to a player’s forecast along the lines of “what the ****? How can they say (insert player name) is going to have only ___? I can’t take this seriously- this is pure bull****!” And boy, it gets annoying. There are a few things to keep in mind when it comes to projection systems. The first is that since we’re dealing with human inputs, there’s a good likelihood that there will be a bug here or there in the initial releases of the system. This happens. Secondly, projection systems do not take in to account esoteric knowledge about the player’s recent weight loss, improved vision, revamped mechanics or anything along those lines. It’s nothing more than an educated guess, and more often than not it will be exceptionally pessimistic due to regression and the like (unless you’re the Bill James Forecasts).
I hope you enjoy these as much as I do.