PECOTA — or “Player Empirical Comparison and Optimization Test Algorithm” for those of you who don’t like clever acronyms — is the statistical projection system owned and operated by the good people at Baseball Prospectus.
While the specific means by which PECOTA predicts what players and teams will do is proprietary — and while the parts that are public are way over my head — the upshot is that PECOTA looks at the career path followed by every player in baseball history, compares current players to those guys, takes its best computerly guess at how the current guys will progress as well, and based on that spits out a set of predictions each year. All prediction systems are going to be subject to some error here and there — you only get out of these things what you put into them — but PECOTA has actually been pretty accurate as far as these things go.
But it does cause an utter s—storm every once in a while.
Last year, for example, PECOTA famously predicted that Orioles’ catcher Matt Wieters would put up one of the best offensive seasons in the history of catchers. That didn’t come to pass, of course, and a lot of people who like to beat up on stat people beat up on the Baseball Prospectus guys over that. Oh, if we were all judged by our worst moments.
I mention all of that because PECOTA’s projected standings for 2010 just came out, and I don’t think it’s a stretch to say that Baseball Prospectus will be the only reputable organization picking the Yankees to finish in third place and out of the playoffs. Oh sure, it’s close — only three games separate them from the first place Rays — but on the outside looking in they are.
Like I said, I’m a moron about such things, but if I had to guess this projection is due in part to the fact that there simply aren’t that many teams in baseball history with as many older players continuing to play elite baseball as the Yankees have, and thus the projection penalizes them a bit more than might be warranted in real life (remember; this is an algorithm talking, not some sportswriter). That is, if you can consider 93 wins a penalty. For what it’s worth, the AL East looks like a tossup in these projections, and the difference between the Rays, Sox and Yankees probably fall within the margin of error.
Not that it will stop some newspaper writer from going crazy in the next day or two and using it as an opportunity to tell stat guys to get out of their mother’s basement and go watch a real baseball game.
Craig covered the bulk of Rob Manfred’s quotes from earlier. The commissioner was asked about robot umpires and he’s not a fan. Via Jeff Passan of Yahoo Sports:
Manfred was wrong to blame the player’s union’s “lack of cooperation” on proposed rule changes, but he’s right about robot umps and the strike zone. The obvious point is that robot umps cannot yet call balls and strikes with greater accuracy than umpires. Those strike zone Twitter accounts, such as this, are sometimes hilariously wrong. Even the strike zone graphics used on television are incorrect and unfortunate percentage of the time.
The first issue to consider about robot umps is taking jobs away from people. There are 99 umps and more in the minors. If robot umpiring was adopted in collegiate baseball, as well as the independent leagues, that’s even more umpires out of work. Is it worth it for an extra one or two percent improvement in accuracy?
Personally, the fallibility of the umpires adds more intrigue to baseball games. There’s strategy involved, as each umpire has tendencies which teams can strategize against. For instance, an umpire with a more generous-than-average strike zone on the outer portion of the plate might entice a pitcher to pepper that area with more sliders than he would otherwise throw. Hitters, knowing an umpire with a smaller strike zone is behind the dish, may take more pitches in an attempt to draw a walk. Or, knowing that information, a hitter may swing for the fences on a 3-0 pitch knowing the pitcher has to throw in a very specific area to guarantee a strike call or else give up a walk.
The umpires make their mistakes in random fashion, so it adds a chaotic, unpredictable element to the game as well. It feels bad when one of those calls goes against your team, but fans often forget the myriad calls that previously went in their teams’ favor. The mistakes will mostly even out in the end.
I haven’t had the opportunity to say this often, but Rob Manfred is right in this instance.
ESPN’s Howard Bryant is reporting that Major League Baseball has approved a rule allowing for a dugout signal for an intentional walk. In other words, baseball is allowing automatic intentional walks. Bryant adds that this rule will be effective for the 2017 season.
MLB has been trying, particularly this month, to improve the pace of play. Getting rid of the formality of throwing four pitches wide of the strike zone will save a minute or two for each intentional walk. There were 932 of them across 2,428 games last season, an average of one intentional walk every 2.6 games. It’s not the biggest improvement, but it’s something at least.
Earlier, Commissioner Rob Manfred was upset with the players’ union’s “lack of cooperation.” Perhaps his public criticism was the catalyst for getting this rule passed.
Unfortunately, getting rid of the intentional walk formality will eradicate the chance of seeing any more moments like this: