Will Aubrey Huff Improve the Giants?
The Giants have signed first baseman Aubrey Huff to a one-year, $3 million contract. There is no real consensus on the signing, as most people are simply stepping back, shrugging and saying… “meh.” After being spurned by Adam LaRoche, the Giants quickly turned to a 33-year old coming off of a below replacement level season. Huff provides the Giants with a known run producer, having been a well-respected middle of the lineup hitter during his career. And he’s a left-handed hitter, which the Giants wanted.
The Huff signing does a few things—some of which are good, others not so much:
- It ensures that Travis Ishikawa is relegated to a bench role.
- It ensures that Fred Lewis is relegated to a bench role.
- It ensures that the Velez/Torres platoon never sees the light of day.
- It ensures that Juan Uribe will not be a starter.
- It moves Mark DeRosa to left field.
- It moves Pablo Sandoval back to third base.
I think #1 is inconsequential. The Giants made it explicitly clear that Travis would be a backup player from the very beginning of the off-season, as it was widely known that they were pursuing corner infielders such as Nick Johnson, Adrian Beltre, Mark DeRosa and Adam LaRoche. As we all know, Pablo Sandoval ain’t goin’ nowheres. #2 is a bit more bothersome to me, as Lewis is one of the few above-average players on the Giants. He walks, he gets on base, he has marginal pop, he’s a good baserunner, and he plays a fine left field.* We need a leadoff hitter? Fred is our best option at that slot in the lineup. Unfortunately, he’s “uncomfortable” hitting in that slot and our manager is uncomfortable asserting his authority. #3 and 4 are a relief more than anything else. Juan Uribe should not be considered a starter under any circumstance, nor should the Velez/Torres platoon even be considered in the first place. Numbers five and six are just fine by me, although I was hoping that the player to man the position would be Adam LaRoche.
At first glance, the Huff signing looks like a good one to me. He’s a good left-handed hitter with a fine track record (career 112 wRC+) and a bit of pop in his bat. The Giants have solidified their heart of the order—it appears the lineup will read Sandoval, Huff, Derosa—but does bringing Huff into the lineup give us a substantial improvement in the quality of our offense, a marginal one, or does he actually subtract from the team?
I have decided to use four projection systems—James, CHONE, ZiPS and Marcel—to get a consensus as to how a number of players project to perform in 2010. I’d use PECOTA over James, but Baseball Prospectus has yet to release their figures and the James projections seem surprisingly reasonable this year. In fact, the systems were all in wide agreement with one another for the players that I’m testing here. So, here’s Huff’s projection:
All of the systems see Huff hitting around 34 doubles and triples while coming close to 20 big flies. The systems also see Huff hitting between .260-.268 with an OBP between .323-.334 and a SLG between .438-.445. For the OPS lovers, that’s a worst-case scenario of a .761 OPS and a best case of .779. And that’s not taking specific park factors into account. Left-handed hitters tend to see an increase of doubles by around 6%, triples by about 20%, and a decrease of home runs by about 9%. This varies, of course, by the hitter—a left-handed pull hitter won’t be affected as much than by a left-handed gap hitter, for example. So let’s take a look at Huff’s home run distribution the past three years within the confines of AT&T Park:
That’s at least three home runs lost, with one more possibly caught at the wall or falling in for extra bases.
This was Huff’s monster 2008 year in which he blasted 32 home runs. I can count at least 7 home runs that would have fallen short.
And another three falling short. So we’re looking at an average of about 4 home runs lost to AT&T. So if his true talent home run rate is 19, as the projection systems suggest, then we can assume that AT&T would decrease his totals by at least a few. Instead of being a 19 home run guy, 15-17 sounds a bit more reasonable given his HR distribution. Conversely, a decrease in home runs could very well result in an increase of doubles and triples—but there’s really no way to be sure. And to be perfectly honest, there’s really no telling whether or not Huff would really have lost all of those home runs at AT&T, since these figures assume neutral wind conditions. Since there’s so much uncertainty in using these “true landing spot” estimates (not to mention some long flies at other parks that didn’t go out and aren’t included), I won’t be altering Huff’s aggregate projection.
Now, on to the projected additions or subtractions Huff’s presence brings to the team. First, I’m estimating runs created through linear weights in a 4.5 runs per game environment. The formula**:
.468*1B + .753*2B + 1.03*3B + 1.395*HR + .311*NIBB + .186*IBB + .337*HBP – 0.09*(AB – H – K) – 0.113*K – 0.293*GIDP
This is then converted into runs above average. Baserunning is measured through Dan Fox’s Equivalent Baserunning Runs (the player’s “true talent,” prorated to projected times on base in 2010) and UZR projections based on expected playing time in 2010.
Ouch. Not quite what I was hoping for. I think there’ll be a bit of confusion as to how Huff’s hitting could be rated so low compared to the average hitter among the less stat-savvy crowd. So, let me explain: through the runs created formula above, we project Huff to create about 67.3 runs with the bat. That’s good, but not great. We then convert Huff’s runs created into a runs scored per game format—imagine an entire lineup of Huff’s. How many runs would they score per game? This is done by dividing runs created by outs and then multiplying by 26.25.*** This gives us an estimated 4.6 runs scored per game. Then we finish it off by using (RpG – LgRpG) / 26.25 * Outs. Through this process, Huff is exactly +1.5 runs above the average hitter. The most optimistic of the projections, James, has him as being +3.6 RAA. That’s good, but again, it’s not great. When all is said and done, Huff comes in well below an average Major Leaguer when factoring in the difficulty of his position, his poor baserunning and below average defense. My estimates have him as being a 0.4 WAR player. Given that a marginal win will cost somewhere around $4 million in 2010, Huff’s projected to be worth somewhere around $1.6 million. Ouch. Huff’s most optimistic projection, James, would have him being 0.7 WAR—and this would have him being worth near $2.6 million, closer to what he’s being paid. Unless he outperforms his projections—which could very well happen (although I have my doubts because of his age), it looks like the Giants overpaid by about $400 K to $1.4 million.
But as sad as it may sound, 0.4 WAR could very well represent an upgrade to the Giants’ lineup. Ignoring positional runs and replacement runs, Huff comes in at -1.8 runs below average (offense + baserunning + defense). By using the same process, we can look at the other options available:
There’s no fun in presenting some work you’ve done unless there’s a surprise, right? It turns out that #1 might not be insignificant after all! There’s a bit of playing time discrepancy with all players, but if we standardize them to a full season, this is what we get:
Naturally, there are a number of issues with standardizing these rates. But, it raises an interesting point—is it possible that Ishikawa outperforms Huff in 2010? The discrepancy between their bats is apparent—Huff is projected to hit about +3-4 runs better than Ishikawa, but the gap isn’t as big as one would like to imagine. There’s also an issue with the baserunning projection, as I am quite skeptical that Travis’ true talent is a plus base runner. Additionally, there’s a source for error within the projections—the 2009 season is, of course, weighted heavily—but Huff’s batted ball data suggests that his .263 BABIP was too low and that a .265/.332/.416 line makes more sense (once factoring in his tendencies to over or under perform his expected BABIP). It’s possible that his 2009 was an outlier of sorts and was due to some poor luck, in which case the projections are being a bit more harsh than they ought to be. Huff is, after all, only one year removed from a 4.2 WAR season (albeit after three years of near replacement-level status). Then again, it’s always wise to be conservative while projecting.
Despite what the projections say, I cannot entirely condemn the signing. I’d bet good money that Huff’s 2009 was an outlier, and the switch from the AL to the NL should be favorable to him. Additionally, there’s also the chance that Ishikawa fails to meet his projections, and the discrepancy between the two favors Huff—not the other way around. Finally, when all is said and done, this is an upgrade from the current rotation of Uribe, DeRosa and Sandoval.
That being said, it’s becoming painfully obvious that Fred Lewis is being mishandled by management. He is one of our better hitters, and at this point in time appears to be capable of outperforming Huff as well. Having Lewis in left, DeRosa at third and Sandoval at first may very well be our best chance of maximizing production—but, of course, the Giants would never do such a thing. For the time being, they’ll be happy with their “proven run producer” while we’re left wincing and praying that Huff isn’t as bad as he projects to be.
*I can hear a universal moan coming from Giants fans as soon as I begin to think of writing that sentence. It’s true; he’s got brick hands and his routes are downright atrocious. But this does not negate the fact that he is a good defender relative to other left fielders. When the pool of players in your field includes Manny Ramirez, Jason Bay and Adam Dunn, an average or slightly below average defender immediately becomes “good.” I would never argue that Fred would make an adequate center fielder or that he could handle the dimensions in right field, but he’s a good left fielder.
**I should note that there’s a potential source for error in the run values for outs, strikeouts and ground into double plays. Generally speaking, 0.09 is the conversion of linear weights runs above average to an absolute runs scale, but that’s not always the case. Additionally, the value of a strikeout (-0.113) and a ground into double play (-0.293) are borrowed from empirical linear weights from recent years. The other values are derived through Base Runs.
***There are some variations on this number—the original figure is 27 (because there’s 27 outs to a game), but there have been modifications to it that vary from 25.5-26.5.