February 13, 2007
Lies, Damned Lies
PECOTA Takes on Second Base Prospects
Second basemen are the bastard children of Major League Baseball. In a study we conducted for Baseball Prospectus 2006, we found that free agent second basemen were dramatically underpaid relative to their WARP scores, a pattern that seems to have continued this year with the cheap contracts signed by Ray Durham and Adam Kennedy. Players are almost never selected as second basemen in the amateur draft; of the top 200 picks in last June's draft, only six were listed as second basemen, and none of those went in the first round. Second basemen aren't supposed to be good prospects: if you've been moved to second base before you've been promoted to the majors, the feeling is that you must have something wrong with you. Since 1990, just six second basemen have cracked Baseball America's top 25: Delino DeShields, Michael Tucker, Chad Hermansen, Josh Barfield, Rickie Weeks, and Howie Kendrick; Hermansen and Tucker both made subsequent moves to the outfield.
PECOTA sings a far more optimistic tune when it comes to second basemen. Kendrick was its #1 overall prospect last year, Dustin Pedroia was #4, and Barfield and Ian Kinsler ranked 13th and 14th, respectively. Second basemen don't do quite as well this year, but Pedroia is back, and joined by several other players who are sure to be ranked higher by PECOTA than by most other systems.
So who exactly is getting it right? I've speculated in the past that second basemen might be overrated by conventional replacement-level metrics. The argument goes that if you take two middle-infield prospects, the more talented, athletic player is both more likely to be the shortstop and more likely to develop a good bat. If there is a correlation between offensive and defensive abilities, you run into certain problems. For example, imagine that you have the following six prospects, who are ranked according to runs created or something similar, as well as their prospective contribution as a defensive shortstop.
SS Player Offense Defense --------------------------- Alpha 70 +5 Bravo 60 +5 Charlie 50 +5 Delta 60 -5 Echo 50 -5 Foxtrot 40 -5
Obviously, as shortstops make more plays then second basemen, you'd want to take the three more skilled defenders, Alpha, Bravo and Charlie, and play them at shortstop, leaving Delta, Echo and Foxtrot to play second base. But Alpha, Bravo and Charlie are also better offensive players as a group; they average 60 runs created as opposed to 50 runs created for the second basemen. If these were the only six players in the universe, you'd wind up concluding that replacement level for shortstops was actually higher than it was at second base. But that doesn't make a lot of sense in a world where most shortstops could play a good second base, but most second basemen could not play a good shortstop.
The higher the correlation between offensive and defensive ability, the more skewed this result becomes. This problem isn't unique to shortstops and second basemen, by the way. It probably also exists to some extent between center field and the corner outfield positions, and it certainly exists between starters and relief pitchers. One of my goals this year is to resolve these sorts of ambiguities in replacement level, looking at ways in which actual or prospective position shifts affect replacement level across different positions. In the meantime, until that work is done, we might want to take a little bit off PECOTA's valuations of second basemen.
Still, it's not like second basemen are chopped liver. Second basemen are responsible for making more outs than any defensive position except shortstop; it's a position that requires both quick hands and quick feet, and which carries a fairly material risk of injury. In addition, the distinction between second basemen and shortstops isn't always clear. The Mets rather arbitrarily decided to assign Jose Reyes to second base when they brought Kazuo Matsui over from Japan, a decision that it took only a couple of months to reverse. Pedroia might have been playing shortstop instead of second this year if the Red Sox had signed Ray Durham or Adam Kennedy instead of Julio Lugo. Finally, as James Click's work in Baseball Between the Numbers suggested, an average shortstop will typically be only two or three runs above average at second base, which doesn't imply a huge gulf between the positions.
You can put an asterisk by PECOTA's rankings of second basemen if you want, but it's likely that both sides have a little bit of room for compromise here. Certainly, when the bias against second basemen is so strong that players like Barfield and Kinsler were completely left off the Baseball America Top 100 last year, things have gotten fairly ridiculous.
There's another, more tangible reason that PECOTA tends to like second basemen better than conventional prospect rankings, which is that PECOTA's Upside rankings are focused on what a player is going to be able to contribute before he files for Free Agency. It's true that second basemen don't tend to age very well, in part because they don't really have anywhere to go once their defensive skills atrophy, but these consequences are more relevant after a player is out of his arbitration window. A player like DeShields, for example, whose career was ultimately considered a bust, averaged a very solid 5.4 WARP during his first six professional seasons. It's also true that second base prospects tend to be a bit older, since they've usually undergone at least one position switch since being drafted, but this too tends to take more value off the back end of a player's career than in his arbitration window.
1. Dustin Pedroia, Red Sox (23) 178.6 2. Eric Patterson, Cubs (24) 147.7 3. Brooks Conrad, Astros (27) 105.3
PECOTA says that Dustin Pedroia is going to be a .300 hitter in Fenway Park, with about 10 homers a season, 35 doubles, 50-60 walks, and defense at second base that's average or slightly above. Does that seem unreasonable to you? It doesn't seem unreasonable to me. His strikeout rates are very, very low--he stuck out just 34 times in 591 plate appearances last year--which allows for plenty of contact. His ability to take walks will get shaved a bit against major league competition, but his good pitch recognition and small strike zone will help hedge against that. A good right-handed line-drive hitter is going to hit a lot of doubles in Fenway Park. Even his skeptics concede that his second base defense should be fine.
Perhaps the better question about Pedroia is: does that kind of performance seem commonplace to you? Because it isn't. Since 1990, only 13 players have hit at least .300 with 50 or more walks and 30 or more doubles while playing at least 100 games at second base:
Craig Biggio (four times)
Ryne Sandberg, Edgardo Alfonzo, Jeff Kent, and Jose Vidro (twice apiece)
Chuck Knoblauch, Fernando Vina, Junior Spivey, Marcus Giles, Mark Loretta, Brian Roberts, Chase Utley (once each)
Although Pedroia doesn't have the upside of someone like Alomar or Biggio, that's a list that includes some of the better players of the past two decades. You do all of those things, and you have a good chance of representing the league in the All-Star game. It isn't a flashy skill set; in fact, it's a skill set that tends to be a little bit underrated, even when it comes to fruition. But we aren't judging this contest on style points.
Eric Patterson presents a slightly different challenge, which is that his numbers at Double-A West Tennessee (.263/.330/.408) just don't look all that impressive. But we need to remember how tough the Southern League played last year--just a 691 collective OPS. Patterson's numbers crudely translated into the Texas League, for example, would read as .283/.351/.458, which suddenly looks like a pretty bright prospect. What's more, his stint in Tennessee was the weak point on Patterson's resume. He went gangbusters in his limited time in Iowa last year, had an outstanding season in his Midwest League debut in 2005, and had a .345/.408/.460 campaign in the Arizona Fall League (not accounted for in his PECOTA). He was once old for his levels, but climbing from the Midwest League to the cusp of the majors in the span of one season has removed that concern. PECOTA doesn't see much growth and there's the risk that the Cubs bury him--I don't like that they're screwing around with him in the outfield when they have plenty of depth at that position--but he's an excellent medium-term prospect.
Speaking of near-term … here's Brooks Conrad, who incidentally went to the same college program as Pedroia (Arizona State). It's clear that the Astros don't take Conrad seriously; they took three years to advance him out of A-ball, even though he was hitting at every level and playing solid defense, and have now left him at Round Rock for three more seasons. But since the Astros are counting on the 41-year-old Craig Biggio and the 35-year-old Mark Loretta to man second base, perhaps Conrad can pull a Dan Uggla on the league this year or next.
Very Good Prospects
4. Elliot Johnson, Devil Rays (23) 85.6 5. Wilmer Pino, Yankees (21) 76.9 6. Tony Abreu, Dodgers (22) 56.1
In spite of being undrafted out of high school, Elliot Johnson rates as more of a conventional prospect, with great speed, good defense, and occasional power. His plate discipline has evaporated a bit, which could indicate that he's pressing, but the D-Rays can't be thrilled with Jorge Cantu and will give him a fair shake.
Every time I see Wilmer Pino's name, I think of Fez from That 70's Show. That's about how seriously I take this projection. The combination of some funky park effects in Staten Island, a limited sample size, and poor range at second is too much to overlook. Tony Abreu had a solid year in Jacksonville that was masked by tough league and park effects. He doesn't get a lot of press, but his skill set--focused around contact-hitting ability with some speed--finds its way to the major leagues fairly often.
7. Alberto Callaspo, Diamondbacks (24) 46.9 8. Andrew Thompson, Brewers (20) 44.8 9. Joshua Johnson, Royals (21) 35.3 10. Emilio Bonifacio, Diamondbacks (22) 35.2 11. Ryan Raburn, Tigers (26) 32.9 12. Drew Sutton, Astros (24) 32.8 13. Jeff Natale, Red Sox (24) 32.8 14. Eric Young Jr., Rockies (22) 32.3 15. Hernan Iribarren, Brewers (23) 31.0 16. Ryan Roberts, Blue Jays (26) 30.4 17. Luis Valbuena, Mariners (21) 28.2 18. Travis Denker, Dodgers (21) 28.0
As you can see, PECOTA doesn't like all second basemen. Only six second basemen crossed the 50-point Upside threshold (one of whom probably shouldn't have), which is the lowest total at any position this year.
The biggest name here is assuredly Alberto Callaspo, who had some superficially good numbers in Tucson (.337/.404/.478), but that was a batting-average heavy campaign for an oldish player with a pretty marginal track record. He's going to hit for contact because he never strikes out and has an excellent chance to be a good utility player, but it's probably too late for him to develop enough secondary skills to become an impact guy. Travis Denker drew some press early in the season, but his performance cratered at Vero Beach, and PECOTA seems to have trouble with the notion of a second baseman who can't really run.
(Players Ranked in Kevin Goldstein's Positional Top Ten or other noteworthy names with Upside scores below 25)
Yung Chi Chen, Mariners (23) 22.7 Kevin Melillo, A's (25) 14.0 Blake DeWitt, Dodgers (21) 9.9
Blake DeWitt generates a fair amount of scouting buzz for a second base prospect as he was a first-round draft pick, but there isn't a lot to like once you start parsing his numbers. He's been helped greatly by his parks throughout his minor league career, and he's been neither rangy nor sure-handed despite being tried at both second base and third. As a 24-year-old in Double-A, Kevin Melillo needed to do more in hitter-friendly Midland to maintain his prospect status; there's also more and more evidence that his defense won't hack it in the majors. I do think that PECOTA underrates Chen a little bit. His 2006 numbers were genuinely pretty good, and I think you can probably give more credence to a late-development story when it's coupled with a cultural transition.
The Big Picture: Rankings Combined With Non-Rookies 25 Years Old Or Younger
1. Howie Kendrick, Angels (23) 282.9 2. Robinson Cano, Yankees (24) 262.4 3. Dustin Pedroia, Red Sox (23) 178.6 4. Ian Kinsler, Rangers (25) 154.6 5. Eric Patterson, Cubs (24) 147.7 6. Josh Barfield, Indians (24) 122.6 7. Brooks Conrad, Astros (27) 105.3 8. Rickie Weeks, Brewers (24) 90.6 9. Elliot Johnson, Devil Rays (23) 85.6 10. Wilmer Pino, Yankees (21) 76.9
I'm sure that everyone is expecting me to talk about Howie Kendrick, but Robinson Cano might be the more fascinating and overlooked player. The comparison to Carlos Baerga works almost perfectly here. Cano might be an early peak guy because the lack of plate discipline could work against him once he loses a few ticks of bat and foot speed, but he should be really darn good in the meantime. Baerga's career is by no means the only scenario here; Cano's #2 comparable is Rod Carew.
Rickie Weeks' tepid rating might surprise some people, but it probably shouldn't; he just hasn't been a terribly effective major league hitter to date, and his defensive ratings have never been good at any level. PECOTA isn't writing him off completely; in fact, it thinks he'll add 70 points of slugging average between now and his peak. But there's a good likelihood that he ends up as sort of a supertweener, a solid regular who could be a star at second base if he could field the position, or at third base or the outfield with about five more home runs. It says something that of Weeks' top five comparables, none played second base regularly in the big leagues after the age of 24, save for a one-off experiment that Alex Grammas conducted with Don Money in the 1977 season.