June 15, 2006
Diamondbacks, White Sox
Quite a bit of attention is being paid to the first few appearances of young Enrique Gonzalez' career:
Date Opp. IP H R ER BB K HR #PIT ------------------------------------------------------- 5/28 @CIN 6.0 3 1 1 1 4 1 86 6/03 @ATL 6.0 4 2 2 1 8 0 102 6/07* PHI 3.0 2 1 1 0 4 1 46 6/13 SF 7.0 1 0 0 2 4 0 99 * not a start
In BP 2006, we wrote of him (OK, this author wrote) that he was "the closest thing to a major-league ready pitching prospect this organization has." This was certainly true then (and now, perhaps), but spoke more about the paucity of upper-level starting pitching talent than how close to Phoenix he was. But after just four appearances, he's already vaulted to third on the team in VORP, which is more than a little surprising. Competence, yes; dominance, no.
Almost-but-not-quite-ready to be revealed to the public are Clay Davenport's minor league translation pages. Long-time readers will recall them from earlier years, but newer readers may want a bit of an introduction, and so this is as good an excuse as any to delve into the world of translated minor league performances.
Let's start with Gonzalez' Triple-A line from this year. His raw numbers:
W L ERA GS IP H R ER HR BB K ----------------------------------------- 4 3 2.24 10 60.1 61 22 15 2 14 35
Other than ERA and HR, not much is spectacular here, and not much indicates he'd have such a high big league strikeout rate (admittedly skewed by one 8 K game in a rather small sample, and admittedly skewed by striking out four pitchers so far). His performance runs from solid (H/9, BB) to unsettling (K) to curious (he's continuing his career-long habit of allowing a lot of unearned runs).
As Nate Silver reminded us this Spring, we're tempted to read that line as we would a major-league line, but we shouldn't. But without knowing the league average line for a pitcher in the PCL, his raw numbers lack appropriate context; as Edward Tufte says in his seminars, an appropriate question to ask when evaluating statistical information is always "compared to what?" Clay's numbers are the "compared to this" that we need, as he finds the League averages for all performance categories, then normalizes them so that 6 K/9 is average, 3 BB/9 is average, 4.50 ERA is average, etc. Here's what happens to Gonzalez' line when we do this:
G IP H ER HR BB K DERA W L H/9 HR/9 BB/9 K/9 NERA PERA ---------------------------------------------------------------------- 10 59.0 60 26 3 15 25 3.97 4 3 9.2 0.5 2.3 3.8 3.97 4.12
Despite the PCL's reputation as an offensive league, strikeouts are more common among pitchers, and so an already-low K-rate gets discounted even more. In other words, a pitcher who strikes out 7 men per nine innings isn't really doing anything above-and-beyond the league average, just as striking out 6 men in the majors isn't anything out of the ordinary. Since the average PCL pitcher strikes out 7.1 batters per nine innings, Gonzalez was actually quite a bit worse than league average here, as his strikeout rate is approaching the Saarloos Zone--pitchers generally lose something off their strikeout rates as they advance, not gain. There's not much left to lose off 3.8/9. His control has been quite good, though, as has his tendency to keep the ball in the park. Let's call his strengths strengths.
But that doesn't make his weaknesses any less weak. Not being able to consistently strike out PCL hitters isn't a great sign, and strikeout ability is arguably the most important indicator of future success. It should be noted, too, that this season is a bit of an aberration in his career--he's kept his translated strikeout rate between 4.4 and 5.4 for the most recent three years. Still not great, but not as low as our current Triple-A sample. While we shouldn't weigh this season too heavily just because it's convenient to be a skeptic, we also shouldn't apologize for it.
As the fine print at the end of a Get Rich Quick infomercial would read, "Past performance is no guarantee of future results." And so all this doesn't discredit his prospect status, nor does it disqualify enjoyment of what he's already accomplished in his brief big-league exposure. But normalized minor league numbers--which contribute to PERA, a better predictor of future performance than raw ERA--are an invaluable part of evaluating his current performance and future chances. You don't need a particular brand of genius to look at his year-to-date major league performance and expect a dropoff (we can start with his .154 BABIP), but, thanks to some translated minor league numbers, we're able to more precisely understand his narrative.
Last week, responding to the struggles of rookie centerfielder Brian Anderson, who is hitting .163/.263/.296 in 157 plate appearances, the ever-unpredictable Ozzie Guillen announced the team was going to stay the course in center. Bolstering Guillen's support was his belief that no one else in the organization "can play center field [well enough] to win a championship."
Usually, such proclamations are based upon a misguided understanding of the dynamics of winning baseball, in that run production is inherently far more valuable than defense, and also easier to accurately evaluate. Last year's White Sox team, however, revealed just how powerful elite defense can be, and thus Guillen's words can't be simply laughed off. Anderson has indeed been horrific at the plate, with a negative 8.8 VORP. Through his first 54 games in center, though, he has saved 10 runs more than the average center fielder by RAA, and 15 more than a replacement glove. Anderson is producing on defense at a RATE of 124, meaning that if he keeps up his current play he'll have saved an astounding 24 runs above the average centerfielder (think the once-spectacular Andruw Jones since 2004) over 100 games. Since 1900, no centerfielder has kept up a 124 RATE for a full season--not Willie Mays, Tris Speaker, Curt Flood, or Jim Edmonds:
Ten Best Defensive Seasons by a Center fielder, by RATE (minimum 100 games)
Player Team Year Games RAA RATE ---------------------------------------------- Curt Welch STL-A 1887 123 34 128 Eric Davis CIN-N 1987 124 26 122 Darin Erstad ANA-A 2002 143 29 121 Eddie Milner CIN-N 1984 108 18 121 Kirby Puckett MIN-A 1984 128 26 121 Thurman Tucker CHI-A 1944 120 25 121 Johnny Mostil CHI-A 1923 135 26 120 Torii Hunter MIN-A 2001 147 28 119 Otis Nixon ATL-N 1992 102 17 119 Paul Blair BAL-A 1973 144 22 118Thanks to Clay Davenport for the defensive figures.
Anderson's work in centerfield has been worth a full win to the White Sox over an average defender (10 runs=roughly 1 win), thanks to catches like the running grab he made last Friday, so that he's managed to compile 1.1 WARP despite the fact he's been worse than a AAA lifer at bat. Sox fans might be shocked to note Anderson has been basically as good as Aaron Rowand (1.2 WARP)--notwithstanding his legendary grab of Xavier Nady's deep fly that earned him a facial fracture and the undying devotion of blue-collar Philadelphia, Rowand has not had nearly the kind of year in center Anderson is enjoying. The comparison accentuates just how much GM Kenny Williams improved his team by dealing Rowand for Jim Thome, whose 29 VORP has already doubled the total offensive output of Chicago's designated hitters last year and helped the club upgrade 2005's middling attack to the one that ranks fifth in runs scored this year.
Before gorging at the feast that is BP's unique offering of defensive statistics, one must remember that Anderson has played less than half a season in center at the major league level, a huge sample size concern. The fact that the observational data coming from Guillen and others (you can also see Anderson's prowess for yourself) agrees so strongly with the statistics is a good indication that Anderson is indeed an elite defensive player, but history and statistical regression tell us he won't continue to win games with his glove at this rate. Defensive performance is notoriously fickle, a fact emphasized by the lack of repeat performers in the above chart, and Anderson's minor league defensive translations were less than outstanding, leading PECOTA to project that he'd be around average with the glove. Then again, Anderson also won't continue to masquerade as former North Side center fielder Timo Perez at the plate, either. Given his strong offense in the minors--he put up a solid .265 EqA last year in Triple-A and garnered a PECOTA weighted mean projection of .269/.329/.468--you'd probably be better off betting on his RATE to continue than his offensive ineptitude, although the extent of his struggles indicate that Anderson's personal development would likely be best advanced with a trip back to Charlotte for a period to get reacquainted with his swing. Strictly in terms of wins and losses, however, without regard to the long term, replacing Anderson in center with Scott Podsednik (owner of a career RATE of 98 in center) or a platoon of the lefty Rob Mackowiak (90 career RATE in center) and righty Pablo Ozuna (6 career games in center), the two in-house options, would be a net loss.
Ozzie Guillen, then, can justifiably call Brian Anderson a championship-caliber defender, and if Anderson starts hitting like he's capable, it could become awfully tough for AL challengers to block Guillen from backing up his statements with a second ring.