Koji Uehara

Mixed Model Pitching Metrics and Signs of Hope

As of Sunday afternoon, the Red Sox had given up 4.84 runs in their first 37 games. If you think that’s bad, that’s only because it is. Only four teams in MLB have given up more runs on a per-game basis. It’s quite hard to be a winning team when you spot your opponents five runs per outing.

We’ve written a lot about pitching here at BP Boston this year, and that’s in part driven by the fact that Boston’s pitching is, uh, let’s go with interesting. It was interesting that the team didn’t get an ace in the offseason. It is interesting that the team chose to invest heavily on offense during an offseason where top-tier pitching seemed to be available everywhere via trade and free agency.

If you’re a fan of the Sox, I think that you’re quickly moving past interesting as a descriptor of the team’s pitching, and more likely to use some combination of four-letter words in it’s place.

Fortunately, I have a little bit of good news for you. Over at the main Baseball Prospectus site, Jonathan Judge, Harry Pavlidis, and Dan Turkenkopf have recently rolled out two new metrics for pitchers: cFIP and DRA. Maybe these two metrics — arguably more advanced than any other pitching metrics that have come before — can help us find areas of hope in the Red Sox pitching staff.


The nut of contextual FIP (cFIP) is that you can use it to measure a pitcher’s true talent level … and nothing else. How it calculates this true talent level is by measuring the events accounted for in the original FIP (walks/HBPs, strikeouts, home runs) and then adjusting these events for a number of mitigating factors like the catcher, umpire, stadium, etc. If you’re hungry for more on cFIP, check out this piece from the Hardball Times.

Deserved Run Average (DRA) is, dare I say it, even cooler than cFIP. Instead of trying to determine a pitcher’s true talent level like cFIP does, DRA’s goal is to use a similar method to tell us how many runs a pitcher should have given up … instead of how many they actually gave up, factoring in a whole host of useful factors. (And for more on that new statistic, go here first.)

First off, I’d like to take a look at a couple of guys in the Red Sox pitching pool who’ve posted a decent DRA. Perhaps some of the pitchers with big gaps between their ERA (what actually happened, more or less) and their DRA (what should have happened) could give us room to think things aren’t quite as bad as they could be.

Clay Buchholz 45.7 4.93 2.89 3.74 77
Koji Uehara 12.0 1.50 2.38 3.76 76
Tommy Layne 11.0 3.27 3.04 3.94 101
Matt Barnes 7.3 1.23 3.26 4.13 108
Anthony Varvaro 11.0 4.09 3.31 4.35 109
Rick Porcello 50.7 4.26 4.33 4.35 96
Wade Miley 35.3 5.60 4.34 4.34 115
Joe Kelly 40.3 5.58 4.37 3.68 105
Junichi Tazawa 17.3 1.56 4.45 3.78 98
Alexi Ogando 16.0 3.38 4.75 4.13 100
Craig Breslow 17.3 3.63 5.03 4.19 102
Robbie Ross 11.7 6.17 5.18 4.89 110
Justin Masterson 35.3 6.37 5.28 5.23 122
Edward Mujica 16.7 3.78 4.99 4.51 108
Steven Wright 15.7 4.02 5.81 4.25 115
Dalier Hinojosa 1.7  0.00 7.93 4.25 103
Heath Hembree 1.3 40.5 10.63 5.61 96

So, there are three guys on the Sox right now who see a substantial difference between their ERA and their DRA: Clay Buchholz, Joe Kelly, and Wade Miley. Each of these pitchers has a terrible ERA (4.93 or higher), paired with a perfectly reasonable DRA. Let’s start with Buchholz.

Clay Buchholz is doing fine, everyone. Though he has, at times, showed his propensity to look like a star one day and a scrub the next, he’s actually been quite good. Sure, the ERA of 4.93 looks bad, but his DRA is a much more manageable 3.74. It’s not quite as good as his FIP (2.91), but that’s okay. If the universe were a fair and just place, Buchholz would’ve likely given up nearly a run fewer per nine, or about five fewer runs over the course of the season.

We have a lot of prior data that Buchholz is a better pitcher than his current ERA reflects, thanks to his previous seasons of success. Now, what DRA is telling us, is that he still is better than that ERA. I think it’s fair to expect that number to drop over the course of the rest of the season, as his true talent level — both by the eye test *and* by his cFIP and other ERA estimators — is higher than his current runs allowed might suggest.

Kelly and Miley are slightly different animals. Both Kelly and Miley have truly horrendous ERAs — the kind that in previous eras might see them demoted to the minors or put out on an ice floe. Enlightened sabermetrics folks can point to their FIP numbers — more than a run lower for each man — and say “hrm, I suppose things should’ve gone better.” And that’s a good start.

But taking things a step further, we can look at their DRA numbers. Miley’s DRA of 4.34 matches his FIP, and shows a pitcher who hasn’t quite been good, but hasn’t been the kind of disaster that gives up five and a half runs per nine. Again, we’re talking about about seven fewer runs that should’ve been allowed, compared to what actually happened. That’s a substantial difference over seven starts.

Joe Kelly’s performance? Even better. His DRA is all the way down at 3.68 these days. So he should have allowed two fewer runs per nine innings, or the same amount of runs allowed over his time pitching as Buchholz should’ve allowed. A DRA of 3.68 puts him at #55 in the major leagues, company more in line with relief aces than with back-end starting pitchers.

Kelly and Miley sporting sharp DRAs doesn’t necessarily mean that they’ll be great pitchers in the future — but what it does tell us is that they’ve been unlucky in the past. When we pair it with things like cFIP or PECOTA, we can get some idea of what we might expect moving forward. With Kelly offering a cFIP of 105, and Miley’s all the way up at 115, —  we perhaps shouldn’t expect them to be great in the future. A mark of 100 is league average after all. But we may not expect them to give up a boatload of runs every time out for the whole season.

See, when looking at cFIP, keep in mind that we’re looking for the lowest score possible. Last season, Koji Uehara was Boston’s cFIP champion, with a score of 63 that rates as “Superb” on the cFIP leaderboard.

Last season, Koji Uehara was Boston’s cFIP champion, with a score of 63 that rates as “Superb” on the cFIP leaderboard.

This year, Koji’s still the team cFIP leader with a score of 77, merely “Great” instead of “Superb”. It’d be a fair guess to assume that Uehara could continue to perform at his current level or thereabouts. So at least one guy on the team can be counted on for great performance.

The next guy who played for the Sox last year to snag a low cFIP for 2015 is Rubby de la Rosa. Rubby has posted an 86 cFIP, firmly in the “Above Average” range. Now that he’s finally able to pitch every fifth day in the Arizona rotation, he has emerged as an … oh no.

Oh no.

Let’s move on.

As scary as it seems, past Uehara and Clay Buchholz (76! He’s great too!) on the cFIP leaderboard, you only start seeing the cFIPs of Red Sox players around the “Average” range. And while, sure, that makes sense given just how bad the team’s been, we’re looking for signs of hope here. Rick Porcello’s hanging out at a cFIP of 96, but we’ll skip him, as that’s not too far out of line from what we might expect.

Instead, let’s look at four relief pitchers who’ve put up average numbers by cFIP this season: Junichi Tazawa (98), Alexi Ogando (100), Tommy Layne (101), and Craig Breslow (102). With the exception of Layne, these other three make up what I’d consider the heart of the Boston bullpen. And all three of these guys are classic cases of ERA outperforming their FIPs.

We hear all the time that pitchers whose ERAs outperform their FIPs seem likely to regress — that their ERAs will swell as luck catches up with the not-so-shiny peripherals. Well, Tazawa, Ogando and Breslow are all out-performing their FIPs, but cFIP seems to think that their true talent level is that of league-average pitchers. They’re not 10-25% worse than league average by FIP … which is what we’ve seen.

Truthfully, given just how bad the team’s been, I was really hoping that these early-season looks at DRA and cFIP might shed some real light on players that just happen to be running into bad luck, or are due for some regression in a positive way. Unfortunately, there’s not a whole lot of silver lining to these dark clouds.

Perhaps the best thing to note is that Justin Masterson really was as bad as he’d been looking, and that the team may be substantially better off in the short term by relying on just about anyone else to fill his cleats. Maybe that’s the note of positivity we’re supposed to take from this exercise: that at least the Sox have some good young lefties in the minors, and that they can’t be much worse than some of the guys on the big league roster.

Photo by Robert Stanton/USA Today Sports Images

Related Articles

Leave a comment

Use your Baseball Prospectus username