CRAP: Pitching And A Win Total

How the hell, exactly, do you project the pitching staff? I tried it last year and after making subjective adjustments for what I thought would be defensive improvements over the '06 club, I came up with 810 runs. They ended up with 783 allowed. Close, but not quite.

The problem last year was that I just didn't know who to pencil in in the rotation. I started with the first 5, gave them rough guesses on innings, then built a fudge factor of replacement pitchers, who I assigned a generic 6.5 ERA. As it was, I ended up predicting a 5.21 ERA for the starters; they ended up at 5.11. I missed badly on the relievers, using the same general methodology, guessing 4.43 instead of the actual 3.81, missing mostly on the huge number of mopup innings I threw in in the bullpen.

At any rate, I think the general approach is the right one. I just think there are too many variables to come up with a close match. But we can come up with some scenarios, and get a range. For by however many runs I did miss it last year, it did show that those historically bad projections were poppycock, and that the maligned pitching staff wasn't going to be much different than the terrible one that played a year before it.

With that in mind, I think we need to make a few assumptions.

First, the starters NEED to give more innings. The Nationals were last in the league in starters innings. League average was 928.5, and the Nats were at just 856. For sake of argument, we'll look for 890 innings from the starters next season (which would've bumped them up two spots)

Second, the average team had about 1450 innings pitched. Eight of the league's 16 teams were between 1440 and 1450. If we use 1450 as our guide, that means we need 560 innings from our bullpen.

Let's start with the bullpen. We have fewer headaches there.

We'll pencil Cordero in for 75 innings and a 3.25 ERA. Rauch gives us 80 and a 3.75. Ayala, 75/3.25. Rivera, 75/4.00. King 40/4.50. I'm fairly comfortable that those will be in the rough ballpark. Now we've got Colome/Hanrahan/Schroder, some of the unknowns. For sake of argument, let's give each of them 50 innings with a 4.50 with the understanding that we just don't know!

Add it all up and we're about 65 innings short of our goal. We'll plug that in with some mopup guy TBD. We'll give him a 6.25 ERA just to be conservative. In reality, only two relievers were over 6 for the team last year.

So we've got 261 ER in 560 innings, a 4.19 ERA. Seems reasonable enough. It's up .38 runs, but given the park change, that's a decent enough guess. It drops the Nats about 7 spots in bullpen ERA, but there are a huge number of teams tightly bunched there, so it's not as bad as it initially seems.

Oh, the starters. Where to begin?

At this point, it looks like Patterson, Bergmann, Perez, Redding and Lannan with Hill licking his wounds, waiting in the wings. Do you feel comfortable projecting any of 'em?

I sure as hell don't. So I'm going to let those projection systems I referred to in those prior CRAP posts do the work.

Patterson: They're all over the place with IP totals, as you'd expect. He doesn't seem to be having any arm pain, so let's be optimistic and give him 140. The ERAs seem high to me, if he's healthy, but we'll stick with the lowest one there, 4.25.

Bergmann: Average IP is about 120 with a 4.25-4.50 ERA. I'm not optimistic that his arm will hold up, so I'm going to lean high here, up to a 4.75. (His ERA was 5.60 in the second half).

Perez: They're guessing about 130 innings with an AL projected ERA of about 5.30. We'll tick a few points of that for the NL, and call it 4.80.

Redding: His projections are all over the place. I think last year was a bit of a fluke. His K/BB ratio was pretty poor, and he hasn't improved on it this spring. I think he'll be in the 5.25ish range. If that's the case, how long is his leash? 80 IP?

Lannan: We know he has giant brass balls, but does he have the stuff? For players with little MLB experience, I trust ZIPS and CHONE a bit more. They're ~5.00 ERA and ~120 IP. 5.00 might sound bad, but that's basically what Chico would've put up in a neutral park, and that didn't seem bad for a rookie, did it? My gut keeps screaming 5.25 and 100 though.

Hill: We know what he's capable of. But we also know how often he's injured. And we know he's not 100% healthy now. Tough decisions here. All those systems project about the same, ~4 ERA, ~120 innings. I'll be a pessimist and knock it down to 100 IP.

We're about 220 innings short of our goal.

Chico's going to get a chance at some point. I think he'll have a tough transition to a newer park, so 80, 5.25?

Let's give Tyler Clippard a few innings. 70/5.25? Balester probably gets a chance at some point. 40/5.50?

We're down to about 30 IP now. Let's just assign a general mopup value to those. Whether they go to Bacsik, Speigner or Mock, it doesn't matter. The 6.50 ERA we'll use is ugly though!

Sum that all up and we've got 482 earned runs in 890 innings, a 4.87 ERA, which is actually down from last year's 5.11. That seems somewhat hard to believe at first blush, but the team did have around 200 SP innings from guys with ERAs over 6. Clearing out Speigner and Jerome Williams helps a lot (even if there's a chance that one or more of this year's guys could duplicate that performance!) 4.87 from the starters would've been good for 12th in the league, so it's hardly a good number.

If we add the starters and relievers together, we get a 4.61 ERA in 1450 innings, which is basically the same ERA as last season (4.58).

But we need to take one more step before we can figure out a run estimate. We need to figure out our unearned runs. We'll use the brute force method. Last season, the Nats allowed 47 unearned runs. League average was 58. I think there's a chance that the UER total could decrease with a real 1B to catch throws, but at the same time, a move from an extreme pitcher's park is going to make it easier for teams to score after an error. So we'll just stick with 50 and be done with it.

Add it all up, and we're at 792 runs, up a few from the 783 allowed last season.

The full spreadsheet in all its gory glory is here.

You can tinker around with the numbers for difference scenarios on your own! What if Patterson's a Cy contender? What if Hill stays healthy? OK, I can't resist those last two. Let's give 'em both 180 IP of 3.50 ERA! We'll drop Clippard and Balester to make the total IP come out, and we're at...750. So that's not as good as I figured!

Wrapping up...

If we're in the 790 range with runs allowed, and in the 770 range with runs scored, we're looking at about 79 wins.

Would you take that?

Before beginning this pointless exercise, my gut said 75-77. So we're in the ballpark.

I'll probably take another look at these projections right before the season starts, when we've got more of a sure handle on the pitching staff.