>Yesterday, in a post about Jenrry Mejia, I mentioned briefly that Zack Greinke is far, far more valuable than Jonathan Pabelbon. Only I sort of skirted over why that is, saying something vague about innings pitched — partially out laziness, but mostly because the debate about starters vs. relievers really deserves its own post. Like this one. So to really get into why sticking Jenrry Mejia in the bullpen is not as good as letting Jenrry Mejia develop into a starter, let me try to answer this question:
Why is a good starter more valuable than a good reliever?
The first and most basic reason why starters are more valuable than relievers is the lazy one I stated above. Innings pitched. More innings are better than fewer innings. More is more is better.
Imagine you have two pitchers, both putting up 2.55 ERA. One pitcher, George Feaver, is going to start and pitch 200 innings with a 2.55 ERA. The other pitcher, Sidd Finch, is going to work 60 innings out of the bullpen and also post a 2.55 ERA. Even though they allow the same number of earned runs per inning pitched, George Feaver is more valuable because he throws more of those 2.55 ERA innings. Would you rather have $200 dollars, or $60? Would you rather have 200 innings of 2.55 ERA, or 60 innings? So even though Jonathan Pabelbon put up a 1.85 ERA in 68 innings in 2009, Zack Greinke and his 229.1 innings of 2.16 ERA were more valuable. Greinke ate up about 250% more high quality innings than Pabelbon. More innings is more innings is better.
But that’s sort of a lazy and simplistic way of looking at things, because it ignores something really, really important: context. In other words, when are these innings being pitched? In tight games, or in blowouts? Jonathan Pabelbon did throw far less innings, but Tito Francona saved Pabelbon’s 68 innings of work for the most important spots in games. Almost all of Papbelbon’s innings were of the late-and-close variety, when the Red Sox were trying to protect a late lead or preserve a tie. Zack Greinke threw 229.1 innings, but some of them were in blowouts when the Royals were up 10-0*, or in early innings, when the game was still far from being decided. Anyone can see that a shutdown inning when your team is ahead 1-0 in the ninth is more valuable than a shutdown inning when your team is ahead 15-2 in the fifth, or even a shutdown inning in the first of a tied game. All innings are not created equal — the importance of the innings pitched is just as vital as the total number when figuring out how valuable a pitcher is.
*May not have ever happened, because it’s the Royals. Just pretend it did.
In fact, if you want to get into things like Win Probability Added (WPA), which accounts for the importance of situations, Jonathan Pabelbon (5.13 WPA) increased his team’s chances of winning games almost as much as Zack Greinke did (6.07 WPA) and more so than every other starting pitcher in the majors except Chris Carpenter. The 68 innings Jonathan Pabelbon threw, because they were in such important spots, helped his team win more games than every starter in the majors, save the two just mentioned.
Crap. I just accidently disproved what I was trying to say.
Oh, right. It doesn’t quite work like that.
Here’s where it gets a little tricky. Taking the importance of the innings into account doesn’t really make Jonathan Pabelbon as valuable as Zack Greinke. Greinke, as a starter pitcher, is much harder to replace than Pabelbon, and thus is still far, far more valuable. Let’s use the Mets as an example to explain why — actually, let’s use the Phillies, because I’m going to be imagining freak injuries to key players, and I don’t want to add to the hex on this Mets team.
Anyway, pretend that tomorrow, Roy Halladay is tragically mauled by a circus lion. Halladay survives without any crippling or otherwise disfiguring injuries, but the lion did manage to rip the UCL out of Halladay’s right elbow. Oh no! Now Halladay needs Tommy John surgery and will miss all of 2010. Sorry, Phillie fans. There’s always the Eagles, right?
Now Kyle Kendrick gets bumped into the Phillies starting rotation to replace Halladay. Halladay was going to throw 240 innings with a 2.50 ERA. Kendrick is probably going to throw just 160 innings with a 4.60-or-so ERA, so the Phillies will still need to replace another 80 of the innings Halladay was going to throw, probably with some combination of middle relievers. Those middle relievers will also pitch admirably, also to the tune of a 4.60 ERA.
What that all mean is that if Roy Halladay should be mauled by a circus lion tomorrow, 240 innings of 2.50 ERA baseball are going to be replaced with 240 innings of 4.60 ERA baseball — that works out to about 55 runs over those 240 innings. So losing Roy Halladay would cost the Phillies 55 runs over the season, at least in circus lion-infested guesstimation land. That’s a big loss.
Now let’s imagine Brad Lidge was the one who was mauled by the circus lion, and not Roy Halladay. Also tragic. Lidge’s UCL is ripped out of his arm with surgical precision by the lion, so Lidge needs Tommy John too. By the way, for the sake of this thought experiment, we’re also going to pretend it was the 2008 incarnation of Brad Lidge. Pretend Lidge is a good closer and losing him doesn’t actually help the Phillies out.
If Brad Lidge gets attacked by Simba, the Phillies are going to need another arm in their bullpen to soak up the 65-or-so innings Lidge would have thrown. Let’s say they choose Jose Contreras and his 4.60 career ERA. For some reason.
Okay, now because the Phils are replacing their closer Brad Lidge, they’re going to have to use Jose Contreras in the closer’s role, right? (hint: no.)
Of course not. They’re just going to shift everyone in the bullpen up a slot. The eighth inning guy, Ryan Madson, is going to become the closer, the seventh inning guys get moved up, the sixth inning guys start pitching in the seventh and so on — Contreras is going to the bottom of the totem pole, not the top. It’s not a straight-up substitution like Kyle Kendrick for Roy Halladay was, because there’s far more picking and choosing in the bullpen. Managers can decide who pitches in what spots, so Ryan Madson is going to pitch Brad Lidge’s important innings now. And remember, we already established that what makes a reliever valuable is the importance of the innings he pitches. Lidge’s “replacement”, Contreras, is only going to throw meaningless innings in blowouts, not the ninth inning in close games, so he’s not really replacing Lidge. It’s not a one-for-one substitution. Contreras may replace the same total number of innings, but not all innings are created equal. This is called bullpen chaining, or something vaguely S&M like that. You can read more about it here, if you so desire. (Read more about bullpen chaining that is, not S&M.)
To summerize, if the Phillies lost Roy Halladay, they would need to replace all of his innings with their sixth starter. There’s not that much picking and choosing — all the starting innings need to be pitched by someone. Every game begins 0-0, so every starter enters the game in the same situation. However, if the Phillies lost their closer Lidge, they don’t necessarily need to stick his crummy replacement in a game to protect a 1-0 lead in the ninth. They can let their second best reliever, Madson, do that, while Joe Replacement soaks up innings during blowouts. Losing Lidge would be a downgrade, but nowhere near as devastating as losing Halladay would be. The Phillies can slot Lidge’s replacement into meaningless spots. You can’t do that with a starter pitcher – no games are rendered meaningless in the first inning.
To belabor the point and make it as simple as possible, think of w
hich scenario you would prefer as GM:
-Mauled Roy Halladay being replaced by Kyle Kendrick and various middle relievers.
-Mauled Brad Lidge being replaced by Ryan Madson.
That oversimplifies everything, but it’s also far, far more true than it is false. For Greinke and Pabelbon:
-Mauled Zack Greinke is replaced by now-starter Kyle Farnsworth and various middle relievers.
-Mauled Jonathan Pabelbon is replaced by Daniel Bard.
This is why losing Joe Nathan should not be an epic disaster for the Twins — their eighth innings guys will have to step in and close, but they’re not exactly a bunch of Replacement Joes* off the street. Now, of course, replacing a closer doesn’t always work out smoothly, and there are plenty of pitchers who can’t handle the mental task of closing games. Kyle Farnsworth, to name one. There is a human element to the game. But then again, every year, some no-name pitcher racks up 30 saves. Leo Nunez saved 24 games in 28 chances for the 2009 Marlins, without having a single save in his previous 106 career games. Closing isn’t for everyone, but it’s probably for more than one might think.
*I should mention that sometimes teams with horrible bullpens will actually have to pick Replacement Joes — Luis Ayala — off the street and have pitch the ninth when their closers go down. But most teams presumably have more relief depth than that.
Back to Flushing. Now imagine that Jenrry Mejia pitches out of the 2010 bullpen for the Mets. And let’s assume he’s awesome — far from a given. But let’s pretend he will be. If he’s the eighth inning guy, all Mejia really serves to do is push someone like Elmer Dessens off the roster — actually, he probably pushes someone better off — and then move everyone else down a spot. Now Sean Green pitches in the sixth, maybe Parnell and Igarashi in the seventh, and Mejia takes over the eighth as a bridge to K-Rod. That’s where most of his value would lie — moving everyone down a slot. How much is that really worth to the Mets?
Now let’s pretend it’s 2011 or 2012. Jenrry Mejia, starting pitcher, is keeping scrap-heap material like Livan Hernandez and the beard of Tim Redding off the roster. The replacement guys like that, who throw 180 innings with 5.50 ERA, are replaced with someone throwing up a 4.00 ERA in 180 innings. That saves 30 runs over the course of a season, more if Mejia pitches more innings. Mejia doesn’t just move Hernandez down a spot — he replaces him completely.
So what the Mejia question really comes down to is whether moving Parnell and Igarashi down an inning today is more important than replacing crud in the fifth starters spot tomorrow. It’s basically deciding between 50 cents today or 3 dollars tomorrow. You can have chance to win a few more games this year, or a chance to win even more games in the future.
You know what? If you’re just 50 cents away from a championship, then you should spend the 50 cents today. But I don’t think the Mets are 50 cents away from a championship. So this doesn’t make any sense, unless Jerry and Omar are just trying to save their jobs.
But, hey, maybe Mejia can push the Mets into the playoffs this year. Maybe a fantastic bullpen allows the Mets win more one-run games than they should and outperform their meager expectations. Maybe one nasty pitch is good enough for Mejia to succeed as a reliever in the major leagues for a long time. And, after all, there is no guarantee that Mejia will ever develop into a quality starter. The bullpen may be his true calling, and the Mets may be wasting his talent in the minor leagues.
But Mejia might turn into a special starting pitcher, and it seems like a chance worth taking for a team with very few starting pitchers. To me, not attempting to maximize Mejia’s value seems horribly short-sighted and impulsive. The Mets might use a little foresight. I doubt having too much was how they got where they are today.