A duality of bi-partisan narrow-mindedness popped up this past week to remind us all that there remain pockets of willful ignorance on both sides of the Sabermetric divide, even while most of us have managed without too much effort to balance intuition with statistical analysis.
First, in reviewing the spinning wheel of trading deadline transactions, some of the seamhead community relied on their projection systems to determine whether teams should be buyers or sellers.
Here are two examples from Baseball Prospectus:
"...maybe baseball executives aren't as smart as we think. That's one way to read the White Sox's decision to hang onto Samardzija at the deadline, despite their 9 percent playoff odds at the time of the deadline."
(With respect to the Blue Jays) "... if PECOTA has the talent levels of everyone pegged correctly, they probably can't do enough to realistically make themselves AL East favorites."
But as the chart below demonstrates, BP’s projection system, PECOTA, provided as little insight into 2012’s final standings as Donald Trump on immigration (or anything else). In this small sample, the system bombed on the Orioles, Red Sox, A’s, Angels, Rays, Indians, A's, Mets, Braves and Reds.
2012 American League
Team Opening Day Projection July 4th Projection Final Win Total
New York Yankees 91.7 95.1 95
Boston Red Sox 83.5 83.8 69
Tampa Bay Rays 83.7 83.6 90
Baltimore Orioles 75.3 80.5 93
Toronto Blue Jays 76.0 78.7 73
Chicago White Sox 87.2 88.2 85
Cleveland Indians 83.6 83.6 68
Detroit Tigers 86.6 83.1 88
K.C. Royals 69.7 71.3 72
Minnesota Twins 69.0 68.9 66
Texas Rangers 94.9 97.4 93
Anaheim Angels 90.1 89.2 89
Oakland A's 73.8 78.3 94
Seattle Mariners 68.8 68.5 75
2012 National League
Team Opening Day Projection July 4th Projection Final Win Total
Washington Nats 80.5 90.5 98
Atlanta Braves 85.6 85.3 94
New York Mets 79.1 84.2 74
Miami Marlins 84.3 78.9 69
Philly Phillies 86.0 77.6 81
Cincinnati Reds 86.7 88.4 97
St. Louis Cards 87.4 87.8 88
Pittsburgh 74.7 84.3 79
Milwaukee 90.4 80.4 83
Chicago Cubs 73.3 67.7 61
Houston Astros 61.9 63.4 55
San Fran Giants 86.6 88.7 94
L.A. Dodgers 80.0 84.4 86
Arizona D'backs 83.9 82.2 81
San Diego Padres 78.6 69.9 76
Colorado Rockies 76.8 66.3 64
That’s because first half results aren’t destiny, and neither is some statistical formula. Teams are fluid, and even when their lineups don’t vary much, player performance does. Some players improve while others wear out. Teams rise and fall over the course of a 162-game grind. Making personnel decisions based solely on statistical formulas is nonsense; it’s why general managers are people, not robots. They need to evaluate not just the numbers but the people who comprise their rosters.
Sometimes the statheads need to take a humility pill and recognize that they rarely have definitive answers about how humans will act.
And then there's Harold Reynolds
On the flip side, you’ve probably seen Harold Reynolds and Dan Plesac on MLB Network doing their Johnny Appleseed imitations, except what they’re spreading all over the land is wanton ignorance.
In a recent discussion of top rookie hitters and pitchers, they made their picks based largely on reputation and hype, after which Matt Vasgersian offered WAR statistics for top rookies. In each case, the duo had failed to even mention the player with the top WAR number. They’re response was to dismiss WAR with essentially these words:
I don’t understand WAR, therefore it has no value.
The details are irrelevant, though Plesac and Reynolds appear never to have heard of Taylor Jungmann, a first-round draft pick now spinning a 2.26 ERA for the woebegone Brewers (and hitting .316 to boot.)
There are plenty of reasons to be skeptical about WAR, particularly for pitchers. WAR accounts for defense with tools that are still as inexact as a Yasiel Puig throw. When applying it to pitchers, WAR attempts to parse the defense of every fielder behind the pitcher and then adjust the pitcher's performance based on how much help he's getting from the defense. Tiny misjudgments in the assumptions can ripple out to very large errors in the final numbers and so should be taken with a sea's worth of salt.
But that doesn't exonerate the two ex-ballplayers. If they are going to comment on player performance, they ought to learn something about the new tools available to them. Announcing their professional ignorance, indeed reveling in it, is akin to Bill Gates refusing to acknowledge that people are accessing the Internet on phones now. Except Bill Gates would be bankrupt while Plesac and Reynolds somehow keep their jobs.
I don't know what the average baseball fan thinks when they see this pair unknowingly admit that they are unintelligent, incompetent and unprepared, but my reaction was to write them off with extreme prejudice. That can't possibly be the reputation they're going for.
No comments:
Post a Comment