It’s late January, which can only mean one thing: it’s time for Baseball Prospectus to release its PECOTA projections for the 2010 season! Not enthused? Well that’s ok, because there are some common misconceptions about what a projection system is. I also encourage any fan to subscribe to BP, as it really is a great site for people who enjoy the statistical side of the game. (As a side note, I can’t go into specific player statistics because it’s subscriber only content and I don’t want to get in trouble for being a stat pirate, if there is such a thing)
What it is:
-PECOTA takes into account numerous player information, matches them with comparable players, and then predicts what the player will do in the subsequent season(s) based on that information. How accurate are the projections? Statistics have shown that there is a .7-.8 correlation between predicted and actual stats, which is strong, but not perfect. Basically there is too strong of correlation to completely dismiss PECOTA but too weak to trust it as the absolute truth.
-PECOTA should be used to get a general idea of what a fan can reasonably expect to see from a player next season.
-When looking at any one specific player, the projection may seem to understate or overstate what the player will likely do. This is because the PECOTA projection system divides its projection into ten categories of percentiles, and uses the 50th percentile score, which is obviously the midpoint. You’ll see what I mean by this a little later.
-As far as the Cardinals are concerned, PECOTA expects them to score 745 runs and give up 702 runs. Last year the Cardinals scored 730 runs and gave up 640, so while the system expects the offense to improve slightly, it expects some regression in the pitching/defense department. It may seem hard to believe that with a full season of Holliday the Cardinal offense is only projected to score 15 more runs, but remember the system only uses the 50% score when running simulations. Elite players such as Pujols and Holliday routinely end up performing in their 70-80 and even in their 90th percentile projections for the season, which would obviously raise the amount of runs the Cardinals would score. For the pitching, PECOTA is very conservative in their projections for co-aces Chris Carpenter and Adam Wainwright. Since Carpenter basically missed 2 years out of the last 3 and Wainwright had his first break-out year in 2009, the projections for the two are probably low. Still, it’s nice to know as a Cardinal fan that conservative estimates of the pitching/defense have the team only giving up around 700 runs.
-Using the projected runs scored and runs allowed, PECOTA expects the Cardinals to win the Central division with an 86-76 record. Like I have said before, these projections are probably a tad conservative, but still it’s nice to know that if things don’t go as well as most fans expect the Cardinals are still predicted to be 10 games over the .500 mark and win their division by a comfortable 5 games over the Reds and 9 games over the Cubs.
-Once again it’s important to not put too much stock into these projections, but it is nice to know that the Cardinals are considered the favorite going into the year. They still have to play the games on the field, but Tony LaRussa has a roster that has the most talent in the division to work with.