February 14, 2010
Give and Take
Discussing Jonathan Papelbon's Worth
Marc Normandin: Jonathan Papelbon once again waited until the last minute to avoid arbitration and signed a one-year deal with the Boston Red Sox. His reasoning for delaying the proceedings has been pointed out in the past, by Papelbon himself-it boils down to the fact that Paps, a closer, feels monetarily slighted after being removed from a career path as a starting pitcher. The following quotes come from early in the 2007 season, his second full campaign as a reliever, and the first in which it was guaranteed by the Red Sox that it was his role for the future:
I'm here to get my fair share of money. My main priority is to stay healthy and be able to make money, not to go out and try and hurry up and win a championship this year [at the risk of injury]. It's not like I'm hurrying up and going back to the closer's role because we have a good team this year and I'm going to blow [my arm] out and try and win as many games as we can [at any cost]. No, it's not going to happen.
Papelbon has been vocal about his desire to climb the salary ladder through one-year deals, either through arbitration or by inking a deal right before he goes to arbitration, as he has these past two seasons. The goal is to take away whatever sting the contracts starting pitchers get in the open market by ratcheting up the value of his relief contracts-it sounds like a chip-on-his-shoulder situation, if you read what Rob Bradford had to say about it prior to this new deal.
Here's one take on it, though: Papelbon is now making and will continue to make more money in the future as a relief pitcher than he would have as a starting pitcher. There are various lines of reasoning that support this stance, from pitcher health to the starter market to the Red Sox and Papelbon themselves, but seeing as this is hypothetical in nature, we're going to flesh out the arguments for and against it.
Tommy Bennett: Let's first figure out the implied free-market value of Papelbon's services from his arbitration deals. Then, we can work backward to calculate the approximate number of wins Papelbon would need to produce as a starter in order to earn the same amount.
A handy rule of thumb for arbitration cases holds that players earn 40 percent of the free-agent value in their first year of eligibility (for non super-two players), 60 percent in their second year, and 80 percent in their third year. We can divide Papelbon's earnings in years one and two of arbitration by the percentage players typically earn of their free-agent value. That will give us the implied free-market value of Papelbon's services. Here's a table:
Salary Implied FA Salary Arb. Year One $6.25 million $15.625 million Arb. Year Two $9.35 million $15.58 million
The Red Sox and Papelbon seem to have a pretty clear agreement on his value somewhere around $15.5 million. Next year, assuming he stays healthy, Papelbon can look forward to earning approximately $12.4 million.
Next, let's see if we can figure out how much value Bizarro Papelbon would have to provide as a starter to earn the equivalent of $15.5 million of value. Last year, A.J. Burnett, CC Sabathia, Tim Hudson, Jason Schmidt, and Derek Lowe all earned approximately the same amount. However, for various reasons (especially that Schmidt did not even pitch 20 innings), this is not a good way of calculating how many wins Papelbon would have to be worth.
One way of doing this would be to find a comparable player. Earlier this offseason, the Red Sox signed John Lackey to a deal that pays him an average of $16.5 million for the life of the deal. PECOTA's weighted mean projection for Lackey's WARP value in 2010 is 4.1. Simply dividing his average salary by his projected 2010 WARP implies the Red Sox are willing to pay approximately $4.1 million per WARP. This is not the most rigorous of calculations, since the Red Sox are actually paying Lackey $18 million in 2010, and a proper analysis would consider all five years of the contract. However, the difficulty of projecting pitcher performance five years into the future make this a reasonable approximation of their willingness to pay.
By analogy, in order to earn $15.5 million, Papelbon would have to be a slightly worse pitcher than Lackey. Using the Red Sox implied willingness to pay, Papelbon would need to be worth about 3.9 wins. Something like 200 innings of 3.80 ERA would get the job done. Does that sound like the kind of projection that would be reasonable for Bizarro Papelbon as a starter?
MN: While it seems easy enough to say yes to the above, given how much Papelbon has thrived in the role of fireman, the relationship between his performance as a starter and a reliever is littered with question marks. Back when Papelbon was first officially made a reliever, Nate Silver wrote that the average difference in ERA between a starter and reliever was 25 percent-you could expect a pitcher to move from relief to starting and have that significant change in his ERA for a variety of reasons, such as the need for fewer pitches and increased velocity.
Papelbon is a special case, though, given his status as a reliever-he's one of the league's elite in that regard, and adding 25 percent to his ERA makes him appear as if he would be one of the league's best starting pitchers as well. It's more likely that Papelbon is on the high end of that scale, as a pitcher who picks up far more than the average in ERA by virtue of closing. In his best days, he used a splitter with fantastic drop alongside a blazing fastball that ticked upwards of triple digits in tandem in order to put hitters away. The splitter's effectiveness has dropped with time, as the opposition is more likely to sit on it and wait for another fastball-Papelbon's mentality is that if the 95-mph heater didn't finish them, the 97-mph one should do the trick. While he has flirted with a changeup that he uses off and on, generally he's a one-trick pony these days, though his record still shows him to be more Seattle Slew than Elmer's Glue.
The point is that he never fully developed a consistent, valuable third pitch, and because of the way he uses the ones he does possess-his splitter is predictable in that it's always down in the zone, falling out of it-it's hard to believe he would have breezed through starts as well as he has relief outings. Let's assume he did work on that change, though, giving him the repertoire he needs to get by. Without the extra velocity on his fastball, you still lose something in the performance.
My own thinking is that Papelbon would have a ceiling somewhere around a 4.00 ERA as a starter, which is not bad by any means, but as Tommy pointed out, is probably not going to get him the kind of money he has found as a relief pitcher. That's just my informed opinion, though-luckily, we can step into the way-back machine and have a look at what his PECOTA forecast as a starter looked like in 2005 and 2006, courtesy of Clay Davenport:
Year Age GS IP H BB SO HR GB% BABIP WHIP RA ERA PERA WARP 2005 24 24 122.2 124 49 109 16 44% .309 1.41 4.65 4.42 4.50 1.5 2006 25 24 126.1 125 48 108 17 44% .301 1.38 4.52 4.30 4.41 1.8
Those projections are a long way off from where he needs to be in order to get the same kind of money, and his comparables don't do much to bolster confidence in his getting significantly better: For 2005, Carlos Fisher, Ben Shaffar (never made it out of Triple-A), Boof Bonser, and James Shields show up, with Shields being the one optimistic case in the bunch due to his late bloom. For 2006, we see Gavin Floyd, Bonser once again, Brett Tomko, and Roy Smith. These are, for the most part, average to below-average starters who didn't have very long or productive careers. Floyd and Shields are the two most interesting names from an optimism standpoint, but it's hard to envision either of them pulling in the kind of value Papelbon has in his time as a reliever.
TB: One valid gripe that Papelbon might have is that he was never given a chance to succeed as a starter. This can work against him in two ways. First, we are assuming that Papelbon wouldn't be able to match his value as a reliever if he were pitching in the rotation. Some of the reasons for this, like health concerns, are separate from the statistical expectations. But some of it also is based on our backward-looking understanding of just how great Papelbon has been as a reliever. In 298 innings since coming to the major leagues, Papelbon has been good for a 2.11 RA. Now, our estimation of his true talent level has to be a little lower than that (as it would be for any pitcher who put up such a good RA in his first five seasons), but it seems silly to suggest that Papelbon hasn't performed better than anyone expected when he made his debut. How can we be sure he wouldn't have outperformed expectations at least as much as a starter? The evidence about his pitches seems compelling, but I worry that it may just be a way to justify our observations.
Another way Papelbon is hurt by never getting a chance to start in the majors is that he was doomed by his own success. I don't just mean that being so good in the bullpen meant he made himself irreplaceable there (although I think that's a factor at work). A pattern that we've seen with nearly every power pitcher who makes his debut in the bullpen is that commentators immediately discount his ability to make it as a starter. Very few observers questioned Joba Chamberlain's ability to survive as a starter until he was dominant in the pen. The same thing is currently happening with Neftali Feliz, whose dominance as a starter in the minor leagues was basically unparalleled last year, and yet many now say doesn't have the stamina to survive in the rotation. I think it's similar by analogy to Nichols' Law of Catcher Defense, in that the more dominant a young pitcher is in the bullpen, the more people wonder about his ability to survive under a starter's workload.
These two separate reasons combined mean that Papelbon may have missed out on some of his upside potential. Since he has already been almost as good as a reliever can possibly be, he can't really have realized much extra upside from here. Can you really imagine many closers with a 10-percent better RA in his first five years? (Keep in mind, Mariano Rivera's RA for his first five full bullpen seasons was 2.24.) However, if Papelbon was a starter, he would have more room to succeed-say as a 3.50 ERA guy-since that's a relatively common level of success.
MN: I think you're right on the nose when it comes to pitchers in these roles like Feliz, but Papelbon is once again a special case. He's a college closer converted to starting in the minors, who was converted back to relief for two reasons: bullpen stabilization and health. The Red Sox were loathe to convert him back to starting once again after watching Papelbon essentially work like a Phoenix Down for the beleaguered pen, and based on the fact that Paps is on a throwing program that, in his own words, is meant to extend his career past 10 years, we can safely assume that the cries for his lack of stamina in the rotation are warranted.
Could the Red Sox have given him a throwing program that extended his career as a starter? I don't see why not, but why waste him in that role if he's as good as he is in the bullpen? I concede that he could have been a bit better than we expected in a starting role, but given what we know about his shoulder and his repertoire, I find it hard to believe he would be much better than that 4.00 ERA or so ceiling mentioned earlier.
Part of the fun of this piece was the back-and-forth between Marc and me. We'd love to have more voices join in the conversation. So, readers, what is your take? How good would Bizarro Papelbon the Starter be? How durable?