To begin, I’ll have to admit that I’m basing this off of a very small sample size: four roto leagues - two from 2006 and two from 2007. That being said, I thought that what I found was kind of cool… and possibly very helpful for those of you looking for an edge next year. One might call this a statistical finding… I’m going to call is psychology.

What I did was take a look at the 5×5 standard roto stats and see how, in the leagues, they correlated with a strong finish in the league. The correlation basically tells you how strong of a linear relationship there is between the stat and finishing at the top (or the bottom). I guess I was pretty surprised, as every year the “experts” and analysts tend to preach the value of the stolen base, and let’s face it: players who steal bases always go way too early in the draft. So it was a shock to me to see how little they correlated with the winners of the leagues I looked at. To explain, a correlation I used (r squared) tells you if there is a close correlation between one stat and final standing if the value is close to 1. The closer to 0, the more the stat is random and cannot be used to predict a finish whatsoever. What my test found is that in the four leagues, the correlation for stolen bases was 0.133, the correlation of batting average was also low at .331. The biggest surprise was that the correlation of runs and order of finish was a very strong .851.

Here’s a peek for all you visual learners:

So, now I need to explain why this has more to do with head games than with stats. I believe, that the pre-season experts and all the draft guides over-emphasize the stolen base. They also tend to preach to getting players who hit for average as well as power. They tend to never really talk too much about runs scored. Managers are convinced that stolen bases are the difference maker, and overrate those few guys out there who swipe 30+ bags.  So that means that there are plenty of players out there in the drafts each year who are huge bargains: those are players who score loads of runs, and players who hit for power, but not for average.  I also think it’s better to have a lineup with players who all steal a few bases a year instead of one or two who can combo for 100 swipes.

Who knows how accurate this is, and I’d love to get my hands on data from more leagues to see just how well this holds up over 20 or more leagues, but I know for sure what I’m not going to do next season, and that’s overpay for stolen bases.