November 16, 2009
Voting for Real
AL Rookie of the Year
A vote. A vote? A vote.
Being invited to help select this year's American League Rookie of the Year as a new member of the Baseball Writers' Association of America was an unexpected honor, and one I took seriously. By way of process, I started off with a day spent digging up data to inform my sense of who to rank, and where, and why. Then, I spent a day or two caucusing with a quartet of colleagues, inside Baseball Prospectus and out, and putting various arguments through the ringer, and using a variety of tacks, from devil's advocate to fully faithful, and everywhere in between.
In the end, I wound up finding myself in a somewhat complicated position, ironically thinking back on a Rookie of the Year Award that coincided with my arrival in Chicago 24 years ago, sorting through my own predispositions against the relative value of a certain type of player, assessing a crowded field of starting pitcher candidates, and thinking even further back to a rookie who won despite not playing anything close to a full season. Finally, I was guided by a critical criterion: electors are supposed to vote on present-season success, not on anticipated greatness.
It was relatively easy for me to pare down the field to a final five: the Tigers' Rick Porcello, the Rays' Jeff Niemann, the Rangers' Elvis Andrus, the White Sox's Gordon Beckham, and the Athletics' Andrew Bailey.
In getting down to that quintet, I somewhat quickly dispensed with the trio of notable Orioles-outfielder Nolan Reimold, starting pitcher Brad Bergesen, and catcher Matt Wieters. In each case, the fact that they hadn't played full seasons told heavily against them, and the question of the impact of their seasons made it difficult for me to sign off on including any of them. Wieters was a relative disappointment-given the build-up we helped give him, of course-and his finishing kick was merely decent, so he was the easiest of the three to set aside. Reimold's year was thoroughly sound, with a .289 EqA and a league-leading total for a position player in VORP (20.8), plus decent reviews in terms of his play afield, mostly crediting him with a strong arm. Maybe it was the desire to find exceptional performance to single out, and not seeing it in what comes out as a good year.
Bergesen, Blue Jays starter Ricky Romero, and the A's Brett Anderson were also not going to wind up on my final ballot, but I ended up using them for comparative purposes to the two starting pitchers who were among my five finalists: the Tigers' Rick Porcello and the Rays' Jeff Niemann. Let's go to a chart:
QS+ Pitcher GS BQS IP WHIP K/9 RA9 ERA+ RS/IP SNLVAR SNWP Porcello 31 11 170.2 1.34 4.7 4.27 116 5.8 4.2 .526 Niemann 30 14 180.2 1.35 6.2 4.13 115 5.7 5.0 .559 Bergesen 19 13 123.1 1.28 4.7 3.79 131 4.9 3.3 .563 Romero 29 16 178.0 1.52 7.1 4.45 101 5.4 4.0 .518 Anderson 30 13 175.1 1.28 7.7 4.77 108 4.5 3.6 .508
The second column, QS+BQS, is just a tally of how many quality starts using runs allowed (not earned runs) the starter generated through six innings; a "blown quality start" is as much a matter of a manager with a slow hook or a bad bullpen as any failing of the starter, so I just like to look at this kind of count to get an idea of a raw number of winnable ballgames a starting pitcher gave his team while generating two-thirds of the outs. Understandably, this punishes Porcello quite a bit, but that reflects the amount of care with which he was handled; as smart as that might be to preserve Porcello's future value, in the same way that I'm not voting for future value, it doesn't help make Porcello's case as far as the right-now selection of a Rookie of the Year. It just reflects the remarkable fact that he was already here, and as interesting and impressive as Porcello's back story is-his youth, or that his previous pro experience was limited to just a year in High-A before the big leagues, and that was a season in which his breaking stuff was kept on a tight rein by the organization-it doesn't really mean anything in terms of his actual performance. It's good stuff to know and worth noting when we talk about Porcello during the course of his career, but as far as the final outcome of Porcello's 2009 season, a year that winds up as a credit to his moxie and to his organization's care in handling him, his performance didn't add up to being enough for me to name him the Rookie of the Year.
So what about Niemann? Here, I guess I was particularly influenced by the fact that while he was good, Bergesen was better, and while Bergesen's season was shortened by injury, Niemann's final counting stats get a big boost from his two complete-game shutouts-against the Royals and the Athletics. Otherwise you've got a guy who got fine run support and provided 12 quality starts through six innings in 28 turns, meaning that he was getting taken out early a lot-again, that's a credit to his organization, but not something that helps him as far as piling up counting stats, and for a full-season regular, counting stats should make a difference.
What we're left with as far as trying to get Porcello and Niemann to stand out is two things: the number of their starts, and their raw wins totals. As far as the workload, the fact that neither made a fifth of their teams' starts while being in the rotations all year is a sign of the times. The latter factor was provided in part by their teammates' run support (and fielding). As seems clear from their run support via RS/IP from when they were in the game (from the always fabulous Baseball-Reference.com), Porcello and Niemann got a heaping dose of help from their friends. Nobody was a workhorse, and none of them ranked among the 10 best starters in the league.
So, what about the everyday players, Andrus and Beckham as far as position players, and Bailey as the odd relief candidate? Beckham's automatically handicapped by his playing less than a full season. Like Porcello, that contributes to a great back story-Beckham's the top Sox pick from 2008 who accelerated through the minors fast-and while he was forced to play third base because of the team's need after Josh Fields flopped, whatever defensive metric you use (RZR, UZR, Clay Davenport's Fielding Runs, Plus/Minus), none of them have complimentary things to say about his partial season playing out of position. Again, we know his future's in the middle infield, but like the rosy expectations I think many of us already hold for him, that can't be a factor in picking him. I'm intrigued by a couple of suggestions as far as Beckham's value, however. A .275 EqA (adjusted for all time) is noteworthy for a rookie infielder, and perhaps there's a Willie McCovey '59 sort of candidacy here, in that Beckham stepped into a contending team's lineup and was arguably its best everyday player down the stretch, hitting .269/.352/.481 after the All-Star break. Beckham also managed 19.2 VORP, not far from Reimold's tally, but by WARP3, he rated higher, 2.1 to the Oriole's 1.9, which allows for the relative value of their defensive contributions. As an additional noteworthy minor fact, Beckham's performance with runners on base was outstanding, as he plated a team-leading 19.8 percent of them.
So what about Andrus? The contributions of the Rangers' shortstop weren't overwhelming as far as his batting: a .255 Equivalent Average (all-time flavor) is merely solid from a historical perspective. You can note certain other additional things to his credit-he wound up with a VORP of 17.0 (lower than Reimold or Beckham), and he chipped in 4.3 EqBRR running the bases while swiping 33 bags in 39 attempts, but a big chunk of his offensive value was tied to his hitting .298/.366/.425 in his hitter-friendly home park. However, he did hit .280/.342/.395 in the second half, and while the future isn't supposed to be a factor, in-season adaptability makes for a nice additional consideration. Not that it added greatly to his case, but Andrus also drove in 11.9 percent of runners on base-not a great mark, but a tick better than Reimold's (11.8 percent).
Taken as a whole, however, his offensive credentials obviously aren't great. Essentially, the linchpin of Andrus' value is dependent on the variety of crude tools we have at our disposal to suggest defensive virtue: he had an 11.7 UZR per 150 games via Fangraphs, finished fifth in MLB according to John Dewan's Plus/Minus, and was second in the majors via RZR, with a more modest 102 Rate2 via Clay Davenport's Fielding Runs. Put all of that together as a broad suggestion that Andrus is anywhere from a very good to an excellent shortstop, not unlike the virtue suggested by his scouting reports. Ratchet up the review to team-wide defensive statistics, and the Rangers were second in the AL in Park-Adjusted Defensive Efficiency, BABIP, and NetDP, meaning that not only were the Rangers doing a great job of converting balls in play into outs, they were also excellent at erasing baserunners with twin killings-something shortstops tend to have an impact upon. Finally, the entire Rangers rotation managed a .507 SNWP, better than the Yankees or Rays, finishing sixth in the AL in 2009, and the team's best rotation-wide performance since 1996. Contrasting that sort of unit-wide improvement to the individual performances of the various candidates among the rookie starting pitchers very much redounded to Andrus' credit, at least to my way of thinking, because his contributions elevated an entire rotation, as opposed to individually improving its quality every fifth day or so.
While having Michael Young at third and a more stable defensive unit overall contributes to these team-wide metrics, we can and should broadly credit Andrus' availability for having a major impact on the team-wide improvement, which made a huge difference for the club's overall fortunes in the standings as well as compensating for a relatively disappointing offense. His total value relative to Beckham and Reimold as an everyday player at a key defensive position is suggested by his winding up with 2.7 WARP3, easily besting them.
Which leaves us with Bailey, whose value is sort of in the ridiculously good category. Nevertheless, a reliever's playing time and value is heavily impacted by his manager's usage pattern for him. Bailey finished eighth in the AL in WXRL, ranking only that high because he spent the first third of the season as a set-up man, which sort of makes an argument for him somewhat like the other shortened-season candidates, like Beckham or Reimold or Bergesen. However, it's worth noting he wasn't just getting easy save opportunities, because between Oakland's feeble offense and Bob Geren's willingness to use him as a multi-inning closer, Bailey got seven saves pitching more than a lone frame (and also blowing a couple of those opportunities). Because of his quality contributions as a set-up man before being entrusted with the closer's job, he led the AL in ARP. To change gears from the value of WXRL or ARP as counting stats and to get a better rate-based sense of his value, Bailey wound up ranking fifth in the AL in Fair Run Average among relievers with 60 or more IP, as his 1.99 relief FRA trailed just Joe Nathan (1.38), Mariano Rivera (1.80), teammate Michael Wuertz (1.81), and Jonathan Papelbon (1.89). Rating behind three great closers is to his credit, obviously; ranking behind one of his own set-up men suggested to me that his manager deserves some credit for doing a great job of picking his spots with his relievers. However, he finished third in the league in relief innings pitched, so he was clearly being asked to do more than your average closer over the span of a full season.
So, how did I end up resolving this dilemma, and casting a ballot? I voted for Elvis Andrus as the American League Rookie of the Year. Maybe it's because of the limitations of defensive metrics, which are more suggestive than absolutely descriptive, but defense was a key factor for the Rangers' performance as a team, and Andrus' addition was central to that. Andrus was the one player who had his job throughout the entire season, who wasn't skipped or saw his workload moderated or modulated to good effect. He improved during the course of the season, and while his offensive contributions weren't outstanding, we aren't talking about the new Mark Belanger either. And perhaps finally I couldn't resist the irony, that in making this selection I was echoing a past, similar pick by the BBWAA, Ozzie Guillen in 1985, one that coincided with my arrival in Chicago. Back then, I felt as a fan that Teddy Higuera had been jobbed, and his .520 SNWP suggests I was perhaps overfond of Teddy's 15 wins, but it's worth remembering that Guillen's case then was weaker than Andrus' is now: the Ozzeroo had a .224 all-time EqA in '85 at the plate, balanced against a 103 Rate2 in the field.
For second place, I voted for Andrew Bailey. His balance of winding up among the top relievers in the game while taking on multi-inning chores was compelling enough for me to credit him for being more than just a creation of managerial selection. For third place, I struggled for quite a while before going with Gordon Beckham. Because of Andrus' impact on an entire rotation and how that compared to the performances of the starters on the slate, I guess I couldn't really steer my way to picking any one of the starting pitchers instead of Beckham, and since the distinctions between the rotation regulars wound up being slender enough, I felt comfortable in selecting for my last player one whose impact on the White Sox lineup was significant.
Thanks to Dan Malkiel for research assistance.