As I'm sure many of you know, the Baseball Prospectus for 2015 is in. I'm still digesting all of it and, if you don't have your copy, I definitely urge you to order one.
At the back of the book is perhaps the crown jewel of BP: the PECOTA projections for the upcoming season. For those unfamiliar with PECOTA it stands for Player Empirical Comparison and Optimization Test Algorithm and is described on Wikipedia as a 'backronym' -- a word I had no idea existed -- because of its namesake, Bill Pecota. Nate Silver originally developed it but has since moved onto FiveThirtyEight as a political analyst that makes extremely successful projections.
What you need to know out of all of this is that PECOTA, as its acronym suggests, relies on a comparison algorithm, where it takes data from the previous three seasons and returns players with similar statistics at the same age.
While the Toronto Blue Jays have Marcus Stroman tied for 13th with Jon Lester for the upcoming season in WARP (Wins Above Replacement Player -- BP's iteration of WAR), there was a name that caught my eye on the WARP-Improvments list. As you may have guessed thanks to the title of this article, Marco Estrada is expected to make the second biggest improvement by WARP of any pitcher in the major leagues.
According to Baseball Prospectus, Marco Estrada was good for -0.9 WARP last season. Nate Silver's computer baby expects Estrada to improve by just over 2 WARP and have a 1.2 win season in 2015. Due to the nature of PECOTA, it is a relatively conservative projection system; it doesn't get especially excited or depressed about any player. So, Estrada's expected improvement, though it may seem not incredibly impressive, is still telling.
That projection puts him behind Jered Weaver and ahead of fellow underperformers Matt Cain, Tim Hudson, Tim Lincecum and C.J. Wilson by expected improvements. But why? Why does PECOTA seem to like Estrada? Beside those names, Estrada feels like an odd man out. All five of those names could, at one point in their career, be described as aces. I know that, to many people, ace involves longevity and that's where Cain and Lincecum fall short. The others fall short of the 'ace' narrative mostly thanks to one or more below average seasons over an otherwise impressive career.
That leaves us with Marco Estrada who, at no point, has been described as an ace. With a season of 2.3 WARP in PECOTA's memory, a season of 1.2 in 2015 isn't far-fetched. Including last year's bad year, Estrada has been worth 0.8 WARP on average to his team the past three seasons. Not a terribly far cry from his projection.
PECOTA goes on to compare Estrada to Brett Myers, Jeff Karstens and Randy Wolf. As Estrada is going into his age 31/32 season, it is important to note Estrada is being compared to the 31-32 year old versions of those pitchers. For Myers, that was two seasons ago. Myers was coming off a career-high 3.2 fWAR the season before and hit 1.2 fWAR (Estrada's projection) the following year. Interesting to the Estrada dialogue: Myers started 33 games that season. In the land of unfair comparisons: Myers hasn't started a game since April 19, 2013.
Jeff Karstens is only one year older than Marco Estrada, so there's not much crystal-balling to be done there.
Randy Wolf on the other hand started games as recently as last year in his age 37 season. In fact, from 2007 to 2011, Wolf started 152 games, averaging over 30 per season. Those are his age 31 to 35 seasons.
Obviously we can't pick and choose who we want Marco Estrada to become, he is Marco Estrada. However, PECOTA projections have made a somewhat bold claim here. Mostly because of its selectively short memory. Where Randy Wolf had 7 above average seasons from age 23-29. Estrada didn't have a noticeably above average season until he was 28.
This is by no means a condemnation of Baseball Prospectus, PECOTA or Marco Estrada. If PECOTA projects something, it's definitely addressing a certain perspective. It is sometimes important to determine where the perspective came from and whether or not to believe it.
Time will tell with Estrada and, frankly, I'm optimistic. But, after looking at why PECOTA comes to its conclusions, my optimism is at least slightly time sensitive.
. . .
All statistics courtesy of FanGraphs.