Friday, October 16, 2009

Improving the 2009 Red Sox: Left Field

Because they're going to be constantly compared this winter, I think it makes sense to discuss Matt Holliday and Jason Bay together. Although many analysts have said that the Red Sox should offer Holliday and Bay the same contract and go with whomever signs first, I think Holliday is clearly the top position player in this years free agent crop. Holliday has several advantages over Bay.

First off, he is 18 months younger than Bay. Although this won't make much of a difference for 2010 or perhaps even 2011 both players will command long term contracts. Obviously if you have to sign Holliday to a 6 year deal, but Bay is available for only 4 that changes things a bit, but I think a west coast team may dangle 5 years for Bay. Also, Jason Bay has the dreaded "old player" skill set. Essentially, folks much smarter than I am have shown that players who rely on walks and power can decline precipitously. There are always exceptions, but the declines can be massive and abrupt. David Ortiz and Travis Hafner are two examples, and writers at USS Mariner have made comparisons between Bay and Richie Sexson just before Richie signed his massive deal 4 year, $50 million deal with the Mariners. Although the parallels aren't perfect, I think Sexson's size had a lot to do with his injury issues and he was coming off a season where he only managed to play 24 games, it is a bit ominous.

Finally, despite the horrendous play that cost the Cardinals a playoff game, Matt Holliday is a much better defender than Bay. With the exception of 2006, Holliday has been an above average left fielder for his entire career and posted 3.2 UZR/150 this season. Bay, on the other hand, is viewed as a well below average outfielder by UZR (-8.7/150, a substantial improvement from 2008's -18.2/150).

To top it all off, if Bay signs with another team the Red Sox will receive a first round draft pick, as he is a type A free agent. While all the other options I've outlined in left field are also type A free agents, depending on who signs Bay the Sox could actually move up in the first round. Because of the reasons I've outlined above, if the Sox want to spend the money on a top shelf left fielder this offseason, Holliday is the obvious choice. Holliday and Bay have produced similar value with the bat, but thanks to the large discrepancy in defense Holliday has "outearned" Bay $53.4 million to $28.5 million, according to Fangraphs valuations of marginal wins.

If the Red Sox do decide that neither of the top outfielders is worth it, I believe that Bobby Abreu is a solid fall back option. Abreu is a free agent after the season, and is coming off an excellent 2009 where he played for a mere $5 million. If you've watched any Angels playoff baseball, there is no doubt this fact has been beaten into your brain already. Abreu overplayed his hand last offseason, turning down a 2 year, $18 million deal with the Rays early in the offseason before settling with the Angels. Most of the concern was due to his horrid defensive play (-25.3 UZR/150) and declining offensive skills. Although Abreu hasn't been a great defender this year a move from right field to left wouldn't hurt, especially in Fenway. Abreu's biggest problem has always been an inability to move back on a ball; with the Green Monster covering him behind, that shouldn't be nearly as much of an issue. Despite being 35, Abreu also stole 30 bases, drove in over 100 runs, nearly 100 runs scored, all to go with a 115 OPS+ . Abreu recently turned down a two year, $16 million contract extension, which was similar to the offer he refused from Tampa Bay last offseason. Should things not work out with the Angels he could be a cheap, short term solution in left field. If the Sox really think they're not going to contend until 2012 or aren't willing to pony up the years for Bay or Holliday (I don't think the money is the issue) this wouldn't be a shocking move. However, Abreu would also require giving up a first round draft pick. For a Holliday type player that seems reasonable, but I'm not sure Theo is willing to give up a draft pick for a short term player.

Another name that has been frequently mentioned is Brad Hawpe. Although the Rockies recently announced that they don't need or want to trade Hawpe, he clearly was the odd man out in the outfield rotation down the stretch and during the playoffs. Hawpe swings a very good bat (.285/.394/.519) which isn't substantially inflated by Coors Field, as he posted very similar numbers on the road (.890 OPS) and at home (.917 OPS). The problem with Hawpe is that the stat heads and the scouts agree - he's an absolute butcher in the field. He posted a -19.5 UZR/150 this season, which, while terrible, was actually a massive improvement from 2008's -46.6. While a move from right field to left field would help, as would the Green Monster, I'm not sure how Hawpe would fit into the Red Sox' plan to hang on to their upcoming talent and try and improve their below average defense. Unlike Abreu, Hawpe would require giving up prospects to acquire.

Overall, I think the Red Sox should take a run at Matt Holliday. Given his age and all around skill set I don't think a 6 year, $90 million contract is out of the question. With Jason Bay, anything beyond a 4 year deal makes me very nervous. A 2011 Red Sox team centered around a marginally effective David Ortiz, a 33 year old Jason Bay, and a 32 year old Kevin Youkilis scares the hell out of me.


  1. Is UZR a rate measure or a counting stat? It seems to vary way too much to be a reliable rate. I mean if -8 is average, or a little below, how is it even possible to have a -46.6? How can you be 500% worse than below average? Does Hawpe not play with a glove?

  2. UZR can be both a cumulative and a rate stat. Typiclly, people use UZR/150, which normalizes the player's UZR to be as if he played 150 games. I have seen other UZRs that low, but it's likely that Hawpe's true talent on defense lies somewhat above that. The sample size is so small for most any defensive metric, so they can swing quite a bit from one year to the next. That said, his UZR/150 for his career in RF is about 20 runs, which is about two wins.

  3. And I think it's also pretty easy to think how a player could be that much worse than average- according to the UZR breakdwon, Hawpe has negative career values for his arm, his range, and the number of errors he makes compared to the average player. If you are subpar in ALL the important defensive skills, it's not too hard to see how it could happen. Also, as far as 500% worse, I think that is probably the wrong way to look at it. Ideally, we'd look at the SD for the data, and figure it out from there, assuming the data is normally distributed. I have no idea about this, but I'd guess that it is not. I imagine that there are more negative fielders than positive ones, but that is a SWAG. Another option would be to just look at his percentile, which I would have to guess is in the 10-20th percentile range.

  4. The UZRs I am posting are a rate stat. Basically, given 150 games, what is a player's contribution defensively. The actual numbers, for Hawpe -46.6, are runs. So Hawpe was worth -47 runs with his defense. Which is unbelievably abominable. Now, I wouldn't say that I would expect him to repeat that; just like offensive stats, you expect extreme defensive stats to regress to the mean as well. For his career, Hawpe has been -18.6 runs / 150 games relative to an average fielder. -46.6 would probably be worse than 10-20th percentile, in fact, I think it is the worst UZR/150 I've ever seen. Manny's worst UZR/150 was -28.4 in 2007.

    Although that seems like an incredible amount of variability, I think we need to keep in mind how much offensive stats can vary as well. Because those are much better established or are much simpler we can't just say they're mistaken, but unexpected years do happen.

  5. Is there no concern about Holliday's ability to hit in the AL? In 350 AB's he hit .286/.387/.831. That's not bad but it's a far cry from the .318/.387/.933 you'd be paying for.

  6. I would guess that some of that is from playing in Oakland though...

  7. Oakland is a tough park, plus it is small sample size, so it's probably not worrying too much about.