It’s usually right around this time of year that baseball fans start to draw some conclusions. Of course, for some fans – often the most, let’s say… “passionate” fans – an 0-for-20 or four-game losing streak in April is enough for him or her to write off a player or team for the season. Others stay agnostic until the bitter end.
But, statistically speaking, by the time we turn the calendar to June, most players have logged enough time for a significant number of their performance indicators to be meaningful. In sabermetrics, we sometimes hear this referred to as the “stabilization point.” Put simply, it means that the sample size has become large enough that majority of the variation of a statistic going forward is predictable based on what’s happened so far. Again: the majority; 50% plus a smidge. That leaves a lot of room for all kinds of weird, unforeseeable things to happen. You know, the fun stuff. And the numbers can’t account for changes in a player’s true talent level. Players do get better and worse, after all, even if less often than we’re often fooled into believing because of an isolated run of success or failure that spikes one number or another on stat sheet.
But we have to start having opinions at some point! I mean, we could wait longer. We could wait until we’re all dead. What’s the fun in that!? At least now we can declare some team or player a bonefide stud or abject failure with some modicum of statistical credibility.
So let’s do that! Specifically, let’s do that with Colorado Rockies Pitchers. But we’re not going to use ERA. If we’re going to jump to conclusions, we ought to at least make sure we’re doing so based on the best data available. So we’re going to use Baseball Prospectus’ fancy new pitching metrics: DRA and cFIP. I conducted a similar analysis around this time last season, in part to challenge the community to question what it thought it knew about one player or another. That’s pretty much what it’s about this time around, too.
Regular readers of mine know how often and how loudly I champion DRA and cFIP (and Baseball Prospectus’ methods of analysis generally). I’d encourage you to click on the links above for the full explanations, but put simply: the “D” in DRA stands for “Deserved” and puts the pitcher’s results into their proper context. Most of us are – Rockies fan especially – are familiar with simple park adjustments; DRA takes this concept to the highest level possible, considering not just a team’s home park, but all of the parks in which a team plays, along with all kinds of other important stuff: the quality of the batters the pitcher faced and defenders on the field behind him, catcher framing, umpiring, base-out states… even the dang weather.
As for cFIP, just like DRA, it puts a pitcher’s FIP into context – that’s what the “c” stands for. FIP, which stands for “Fielding Independent Pitching,” is on the same scale as ERA, but is a function of just strikeouts, walks/bean balls, and home runs allowed. It’s a favorite among lovers of advanced baseball statistics because it purports to more accurately measure a pitcher’s “true talent” by stripping away all of the things that happen on a baseball field that, at least according to DIPS Theory, are beyond the pitcher’s control. Of course, we know that that isn’t entirely true. Pitchers can influence the quality of a batter’s contact. We’re still not entirely sure how it works – it’s the dark matter of baseball – but we know the dark matter is out there and that FIP doesn’t hand out enough credit or blame it. Still, FIP is undoubtedly more predictive of future results than ERA. And cFIP is even better than FIP. In fact, if you clicked the cFIP link above (really, please do!) you’ll know that it’s literally the best.
So, just as with ERA and FIP, it’s useful to look at DRA and cFIP in tandem: the former being the absolute gold standard for interpreting what’s happened in the past, and the later as the absolute gold standard for predicting what’s to come in the future.
First, let’s see what’s happened in 2016 thus far. This is your ERA/DRA comparison. We see Rockies players’ ERAs featured prominently on every webpage and broadcast. Can we trust them? (In this table and the one to follow, statistics are up to date for games through June 6th, and I’ve included all Rockies pitchers who have thrown at least 10 innings – which screens out David Hale, Jason Motte and Jason Gurka.)
|Jon Gray||52 1/3||5.33||3.35||91|
|Christian Bergman||18 1/3||5.89||3.93||98|
|Chris Rusin||48 2/3||4.62||4.71||106|
|Carlos Estevez||17 1/3||4.67||3.84||97|
|Jorge De La Rosa||27 2/3||10.08||4.60||105|
|Tyler Chatwood||77 1/3||2.79||4.73||106|
|Eddie Butler||36 2/3||5.65||4.65||105|
|Chad Bettis||69 1/3||5.58||5.37||112|
|Scott Oberg||11 1/3||4.76||5.04||109|
|Chad Qualls||18 1/3||5.40||5.76||115|
Before getting into the weeds, I should first point out that DRA tends to “squeeze” towards the middle. Very few pitchers deserve extreme ERAs at either end – a lot has to go right (or wrong) beyond the pitcher himself to generate outlier numbers. Case in point: The only two pitchers with DRA’s under 2.00 in baseball are Clayton Kershaw (1.93) and Jose Fernandez (1.77). In other words, unless you’re a baseball God, you need some external factors working in your favor to reach such heights. Also, DRA- is included in the table, which puts DRA on a scale centered around 100 (league average)with lower numbers being better. DRA is already park adjusted and all that, but it can helpful to see DRA presented this way to help compare players to one another and the league generally.
Perhaps the first order of business is to correct an injustice. Jorge De La Rosa has not been as bad that nasty ERA suggests. Yeah, he’s given up a lot of homers – that’s on him. But a peak at his game log provides some clues as to why DRA cuts him so much slack – along with insight into how DRA works: every single one of his eight appearances has come in a hitters park (5 in Coors, 1 each in Arizona, Cincinnati, and Boston). There are a ton of quality offenses in that list, as well: Red Sox (1st in MLB by wRC+), Pirates (2nd), and the Giants twice (10th). There are likely other, less easily discernible factors, as well – maybe an usually high number of lefty-killing hitters, an umpire squeezing the zone, 15-mph winds to left-center… who knows? What we DO know is that Jorge’s been a lot closer to his standard self than it might appear.
A couple other pitchers on whom DRA sheds some positive light are Jon Gray and Christian Bergman. I think most attentive Rockies fans know that Gray has pitched much better than his ERA suggests – it’s been mentioned repeatedly on-air by the ROOT Sports crew, and deservedly so. But DRA suggests he’s not only been undeservedly punished by ERA, but that he’s been downright terrific: the 21th best starter in baseball (one slot behind Max Scherzer). Maybe the difference stems from bad luck with sequencing – Gray’s 62.7 LOB% (the percentage of runners stranded on base) is way below league average and is generally thought of as something pitchers can’t control… except for high strikeout pitchers who tend to strand more runners. Folks, that LOB% is going to improve, and Gray’s ERA along with it. As for Christian Bergman, he doesn’t get the same benefit of the doubt during the broadcast – which makes sense since long relievers don’t get much airtime or analysis in general – but he probably should. Dude’s been good.
Speaking of relievers, Carlos Estevez and Justin Miller show well by DRA – much better than high-profile acquisitions Jake McGee and Chad Qualls. But it’s Boone Logan who reigns supreme. His DRA doesn’t tell us anything we didn’t already know by looking at his even better ERA, but it’s still nice to see that a more sophisticated metric believes in what we’ve witnessed thus far. Better late than never, I suppose. If nothing else, even if the Rockies fall well out of contention over the next month, he’s pitched himself into a quality trade chip.
By far the biggest red flag in this table is Tyler Chatwood. He’s managed to keep his ERA in check despite quite a few baseballs bouncing around the yard (12th highest contract rate in the entire league among qualified starting pitchers). Ground ball pitchers like Chatwood can survive that way better than most, but it also means he’s more reliant than most on those ethereal qualities that we believe some pitchers possess (to induce soft contact, avoid optimal launch angles off the bat, etc.) even if we don’t know how or why. Maybe Chatwood has the magic and he’s earned all those good outcomes on balls in play. But maybe he’s just been lucky. His start on Monday night against the Dodgers offers a good case in point. He lasted eight innings, giving up just one run on one hit. That’s awesome. But Dodgers Stadium is a very forgiving place for pitchers. (Remember those two long fly ball outs Chatwood got in the 8th? Those are homers most other places.) And the Dodgers’ offense has been really bad this season (8th worst by wRC+). Anyway, for these and other reasons like them, DRA isn’t buying into Tyler Chatwood’s apparent run prevention thus far.
But enough with the past! We’re forward looking folks here in Rockies Nation – perhaps in part because the Rockies’ baseball past is so ugly, but, well… fart noise. The next table shows the aforementioned cFIP, along with the (slightly) more well-known FIP, along with FIP’s scaled-to-100 version, FIP- (again, 100 is league average and lower numbers are better). This gives us a better idea of what we might reasonably expect going forward, and how traditional FIP might lead us astray.
|Jon Gray||52 1/3||3.45||78||85|
|Christian Bergman||18 1/3||3.85||88||97|
|Chris Rusin||48 2/3||3.25||74||101|
|Carlos Estevez||17 1/3||4.18||95||101|
|Jorge De La Rosa||27 2/3||6.40||145||104|
|Tyler Chatwood||77 1/3||3.84||88||108|
|Eddie Butler||36 2/3||4.62||105||108|
|Chad Bettis||69 1/3||4.70||107||111|
|Scott Oberg||11 1/3||5.97||136||116|
|Chad Qualls||18 1/3||5.00||114||118|
Let’s stay with Chatwood to start off. Turns out that cFIP is throwing up the same red flag as DRA. Traditional FIP likes Chatwood primarily due to his home run suppression, which has been great – so great, in fact, that it makes up for the very low strikeout rate (10th lowest in the league among 107 qualified starters), and the roughly league-average walk rate (52nd of 107). Chatwood has succeed using this model before – check out his numbers in 20 starts in 2013 – but it relies not only a high ground ball rate (which is known repeatable skill that Chatwood definitely possesses), but also a low HR/fly ball ratio on the fly balls he does allow (which falls into that “ethereal” category). Again, maybe Chatwood has that magic, but cFIP doesn’t believe in such things.
That last start against the Dodgers again offers a useful case in point. Chatwood struck out just five over eight innings while walking almost as many (4) in the process. That’s a really bad ratio made even worse in consideration of how bad the Dodger’s offense is. And while Chatwood kept the ball in the park, there were at least those two deep fly balls that would have been homers almost anywhere else, and also more fly balls in general than an elite homerun-suppressor can usually get away with (50% GB rate for the game, which is fine, but not good enough for pitcher who essentially lives or dies on GB% rate alone). cFIP knows all of this.
Other pitchers with whom cFIP is unimpressed include Jake McGee and Chris Rusin. In both cases, the stories are similar to Chatwood’s: too few strikeouts and home run prevention cFIP doesn’t believe they can keep up. cFIP still likes Rusin overall – a better version of Chatwood, basically – and a starting pitcher with an essentially league average FIP/cFIP is actually quite valuable. It’s also worth pointing out that Rusin produced a downright excellent cFIP of 92 last season (though a not-so-great DRA- of 118). So either Rusin lacks the magic that Chatwood does or doesn’t possess, or he’s eventually going to start preventing the runs his underlying talents according to cFIP says he should. McGee’s case is more troubling. We really don’t even need these advanced statistics to know that the ERA is fools gold. He’s striking out almost half as many batters as he used to, while simultaneously upping his fly ball rate. Not good.
Will Jake McGee’s top keep spinning?
On the plus side, just as with DRA, cFIP doesn’t buy the reports of Jorge De La Rosa’s demise. cFIP doesn’t see an all-star, but it does see a roughly league average starter who’s just run into some bad luck thus far. As a Jorge De La Rosa superfan who doesn’t want to see his favorite Rockies pitcher of all time go out this way, I’m clutching to that 104 cFIP will all my digital might. In other news, Boone Logan continues to hold up to scrutiny (which is nice, because the latest free-agent reliever de jour, Chad Qualls, certainly does not). Christian Bergman holds up, too. By the way, Bergman, just like Rusin, also posted a terrific cFIP in 2015 (87) along with a pedestrian DRA- (108). So his results going forward are worth watching for the same reason.
Speaking of 2015 numbers, one notable player who isn’t replicating them – either at the surface level or at the DRA/cFIP level – is Chad Bettis. After a terrific break-out season in 2015, Chad Bettis entered 2106 as the top pitcher on the staff. De La Rosa continued to hold the ceremonial title of Staff Ace given his body of work and history with the team, and (rightly) received the honor of starting on Opening Day. But Bettis got Game 2, and all the buzz was with him. I wrote about him myself over the off-season, wherein I cited these same DRA and cFIP metrics; in 2015 he posted a 95 DRA- and a 92 cFIP. This year: a 112 DRA- and a 111 cFIP. Something is different – something more than just a few bad bounces or tricky match-ups. That’s not to say he can’t “make some adjustments,” as we often hear in relation to players struggling, but given the fact that there are as many bad innings in Bettis’ past as good ones, it’s fair to wonder which version of Bettis is the “real” one.
But look, I’m totally burying the lead here. The best and most important takeaway to be found in either of these tables is that downright gorgeous 85 cFIP of Jon Gray’s. It confirms the truth we think we’ve seen via the eye test, and the legitimacy of what we want to infer from those awesome strikeout rates and other peripherals. By cFIP, Gray is the 18th best starter in baseball, better than guys like Jon Lester and the aforementioned Scherzer. It’s also better, by the way, than peak Ubaldo Jimenez (86 cFIP in 2010). What’s more: this follows a 2015 season in which he posted a similarly awesome 88 cFIP as a rookie.
And so, despite the 5.33 ERA, there isn’t much doubt in my mind that we’re on the verge of a whole lot of run prevention from young Mr. Gray. In fact, since it’s June, and in the spirit of that hollowed time on the baseball statistician’s calendar we call Stabilization Season, I’ll go ahead a draw a conclusion that should bring joy to all the good boys and girls huddled around the diamond at 20th and Blake:
Jon Gray is an Ace.