Write Derek at drksmart @ gmail.com
Write Phil at phil.bencomo @ gmail.com
by Derek Smart
Yesterday, faithful reader Todd S. sent me an email reminding me (as sadly, I needed it) that Baseball Prospectus just put this season's PECOTA cards up (subscription required for all cards, except catchers and members of the White Sox).
He also sent along a few quick calculations of the NL Central's projected VORP numbers which have inspired me to see if I can sort through what's out there to come up with some team comparison's, and perhaps even some breakdowns of what happens when certain players receive playing time over certain other players (and you know who you are...).
However, since I've been out of practice for a few days, and I'm still trying to figure out if I've got the chops to sort through the numbers and do what I'd like to do (reconciling playing time issues can be a real bear on a team level, in particular), I thought I'd pass along some observations I had while perusing some of the Cubs' cards.
Those of you familiar with PECOTA know about the Most Comparable lists that accompany each player card, and while there is at least a modicum of shared intent with the Most Similar lists you'll find on a BBRef player page, the way the two different lists are compiled are very different. Still, like with the BBRef pages, it's always fun to see who PECOTA thinks a particular player is most like.
It so happened that the first player card I visited belonged to Kerry Wood, he of the lifelong comparisons - legitimate or not, harmful or not - to the likes of Nolan Ryan and Roger Clemens, but when I worked my way down to the bottom of the page, I was surprised to find out that Wood's #1 comp according to PECOTA was none other than the Red Baron, Rick Sutcliffe. Circa 1985.
So, if I'm looking for some hope regarding Wood's seasons to come, I now know where not to look.
PECOTA thinks Patterson The Younger is nearly ready to play right now, giving him a Weighted Mean batting line of .261/.324/.419, while playing excellent defense at second base, this despite having all of 37 PAs - and lousy ones, at that - above A ball.
It is the issue with any projection system: if there isn't much data, the projection's accuracy takes a big hit, and although I haven't seen Young Eric play ball, his PECOTA projection strikes me as a prime example of that phenomenon.
On the opposite end of the spectrum, one of the more glaring underprojections I saw among the Cubs was that for Matt Murton, with a Weighted Mean batting line of .281/.343/.418. Now, I realize that I'm in the grips of a severe case of the mancrushies when it comes to Murton, but that's simply not the player I saw last year. Granted, I don't expect the .321/.386/.521 he posted with the club last year, but even knowing I'm biased, I think the truth lies somewhere in between.
All fear Cleanliver Goodbody! And when I say "All," I mean "All You Cub Fans." However, it's not the Weighted Mean batting line that should send shivers down your collective spines (although he comes in at a dismal .248/.292/.375), it's his Collapse Rate, the quickie definition of which is the percent chance that his production will decrease by 20% relative to the weighted average of his work in the previous three seasons. Goodbody's Collapse Rate? A charming, 48%. That, by itself, is worth a sleepless night.
The exact opposite of Collapse Rate is Breakout Rate - in other words, it's the percent chance that a player's production will increase by 20% relative to the weighted average of his work in the previous three seasons. I like to think of it as PECOTA's "Hope Index."
Interestingly, the system sees Greg Maddux and Mark Prior as excellent breakout candidates, putting their chances at 32% and 29% respectively, and while Carlos Zambrano only has a 15% chance to bust out beyond what he's done recently, he's been so good that the number actually seems kinda high.
What's interesting about Big Z's card is his Attrition Rate, which is the chance that his batters faced will decrease by at least 50% relative to his baseline forecast (his card doesn't have batters faced, but his Weighted Mean IP forecast is 223.7). Many of us have screamed about the use and abuse of Zambrano over the years, and since this number deals with playing time, it's something of a proxy for injury risk (as well as time lost due to ineffectiveness). So, bearing in mind all of our kvetching in the past, what's Big Z's Attrition Rate? The Big 0.
I'll quit for now, but there's a lot to be gleaned from studying these puppies, so you'd best believe there's more to come before the offseason's through.