Souza Caps Breakout Year With Callup
BEST PLAYER: Triple-A Syracuse outfielder Steven Souza earned both the MVP and rookie of the year awards in the International League, which he led in average (.350), on-base percentage (.432) […]
Statistics only tell part of the story
by Alan Schwarz
NEW YORK--Archie Bunker looked straight into Meatheadís bushy eyes. ďDonít feed me no statistics!Ē he barked. ďLook at the facts!Ē
It was 1973 and the subject was gun control, but olí Arch might as well have been talking about baseball broadcasters, many of whom take to statistics like donkeys to hay. They babble on about how so-and-so has "12 hits in his 17 games" and "holds lefties to a .251 batting average," with the clear implication that since the stats sound good, they must be good. Rarely, if ever, do they explain what the average player does in those situations. Is 13 high? Is .251 low? Is a 75-6 record when leading after eight innings good, or just a smokescreen for a bad bullpen?
With all data graciously donated by Baseball Info Solutions, the relatively new statistics company founded by former STATS Inc. boss John Dewan, we present here a listenerís guide to statistical prattle. Feel free to tape it to the back of your remote; I know I will.Closer Relief
"Has 30 saves in 38 opportunities": This statistic is spouted for closers left and right, with no appreciation of what expectations are. Now, all save opportunities are converted an average of 68 percent of the time, but thatís including those of setup men who arenít expected to complete the game. Filtering out pre-ninth inning opportunities from pitchers who are not their team's closer, saves are generally converted 85 percent of the time--or 32 times for every 38 chances. Thirty doesn't look so good anymore.
"Has hit in 12 of his last 17 games": This is meant to imply that some batter might not be on a technical hitting streak, but is hot. But if you look only at every game's starting non-pitchers, they get at least one hit about 67 percent of the time. That means that in any 17-game stretch, he'll usually have 11 or 12. Big deal.
"Has an .834 OPS": We all know how on-base plus slugging percentage has become en vogue these days, for good reason. But unlike batting average, RBIs and the like, few are aware of what to want from each position. Here are the averages for each spot on the diamond last year, among players with at least 200 plate appearances: first base (.840), second base (.754), shortstop (.731), third base (.769), catcher (.745), left field (.843), center field (.784), right field (.838), DH (.825).
"Has a 75-6 record when leading after eight innings": Sounds wonderful, but almost every team has a great record when leading late in the game. Even greater than most people think: Teams leading after eight innings the last two years held on to win 95.5 percent of the time (or a 77-4 record in 81 games). Even after seven innings, with two whole innings left, the conversion rate was 91.4 percent (74-7).
"Holds lefties to a .249 average": It's easy to be impressed by the statistics posted by lefty relievers who are inserted into games when they're most likely to succeed. This one's a little tricky to define, but looking at lefty relievers facing lefty batters, it's clear they should be held to a higher standard. The typical batting average allowed was .245 last year, 20 points less than the MLB-wide .265 in all other situations.
"Has stolen 21 bases in 29 attempts": Whether you're a fan of the stolen base or not, it would be nice to know what average is among the players who generally are expected to be good at it. The success rate among all players last year was 69.4 percent. But among leadoff hitters, it was 74.2 percent (suggesting 22 per 29 attempts), No. 2 hitters 70.6 percent (20 per 29).
"Leads all NL rookies with a .287 average": Interesting, I guess, but how many rookies are actually in the mix? Consider this: In the past two years combined, just 11 total rookies qualified for the batting title in the first place, and four of those were National Leaguers: Brad Wilkerson, Scott Podsednik, Marlon Byrd and Ty Wigginton. Not exactly a large group, is it?
"Hits .342 on the first pitch": This stat is often used by broadcasters and writers to demonstrate how a player doesn't need plate discipline. (He mashes when he jumps on the first pitch.) But the stat line they're reading--"on 0-0 counts"--actually includes only at-bats that end on the first pitch. Well, naturally, if you take out times when the guy swings and either misses or fouls the ball off, everyone will look good! Keep this in mind: For the standard stat that most people use, the typical batting average jumps from .264 to .336, OPS from .755 to .888. For the stat that most people mean--when someone offers at the first pitch--average goes up only to .278, and OPS, because they guy will rarely walk, decreases slightly to .754.