I’ve been playing around with the CHONE Projections for the Cubs, in particular the pitching staff. It’s interesting to see how it breaks down. CHONE projects the Cubs’ top 5 pitchers to be Zambrano, Lilly, Harden, Dempster and Marmol, with Marshall coming in sixth, in that order. When I saw that, a couple of things jumped out at me:
1) Harden is less valuable than Lilly? Not that I don’t love Teddy Ballgame, but when he’s healthy Harden is one of the very best pitchers in baseball, and Ted is not that.
2) Marmol is worth more than Marshall? Marmol’s a better pitcher, but I expect Marshall to pitch a lot more innings than Marmol.
Both of these can be explained by playing time estimates. CHONE’s projections only have Harden pitching 94 innings, to Lilly’s 172 projected. Harden’s a great pitcher, but I’m willing to believe he’s not twice as valuable per inning as Lilly is. The projections also project Marshall only pitching 117 innings, versus Marmol’s 80. Once again, I’m completely willing to believe 80 innings of Marmol’s slider is better than 117 innings of Marshall.
But those playing time estimates don’t really sound right to me. It’s an inexact science, of course, but I think the Cubs can get more than 94 innings out of Harden (and for that matter, more than 172 out of Lilly). And since Marshall is only 26 this year, the only thing that would keep him from pitching more than 117 innings would be his performance.
So what I did was first take the CHONE innings pitched numbers and working from them, come up with some numbers I think are a little more likely – although I admit this is just a guessing game, and I only guessed for the players who are likely to be on the opening day roster. In reality, of course, the Cubs are going to use more than these pitchers. I’m just assuming any unaccounted-for innings (about 250 in the CHONE projection, and about 80 in my estimate) are going to be taken up by replacement-level pitchers. So, here are the two sets of innings estimates:
Are these perfect? Probably not. Harden does have a tear in his shoulder after all, maybe he won’t make 120 IP. My estimate for Heilman seems high, but I assume that he’ll be making a few starts. I lowered the Samardzija estimate, but it still seems high to me – I think he might end up spending most of the season at Iowa, with Hart probably next in line to come up.
But now that we’ve got these two sets of playing time numbers, what can we do with them? Well, let’s see where the Cubs’ pitching wins are coming from. CHONE projects, using its playing-time estimates, a total of 16.1 WAR for the pitchers I used. The estimate goes up to 18.8 with my numbers. Here’s how those wins breakdown between the starters (not accounting for spot starts) and the bullpen, in each estimate:
Either way, the starters are doing most of the work, which is about what you’d expect. The bullpen’s weight is really even a little lower in my playing time estimate, because the bullpen estimate includes starts I expect Heilman to make.
So we can get a better estimate by looking at this on a per-pitcher basis. The starters are on the main pie chart, with the relievers on the breakdown. We’ll do the CHONE PT estimates, first.
And finally, using my playing time estimates:
There’s no particular conclusion to be drawn from all of this; my playing time estimates are just guesses, and seem a little optimistic to me. At the same time, I don’t put a ton of stock into CHONE’s playing time estimates, which seem to be based mostly on past playing time. That doesn’t make a ton of sense for players like Marshall or Dempster. But it’s interesting (to me, at least) to see where the wins might be coming from. I’ll have a similar post on position players in a day or two.
Update: For some reason I forgot Luis Vizcaino is on the team and will be on the opening day roster. So I updated, replacing Kevin Hart, who I had on the opening day roster, with Vizcaino. That also means that Hart is probably in line to get some of those unaccounted-for innings, and he’s better than replacement, so the bullpen is really worth a little more than I estimated – maybe half a win.