July 23, 2015
Cole Hamels' Hapless Helpers
The White Sox and Mets have had atrocious defensive infields this season. Despite very strong starting rotations that miss a great many bats and induce a good amount of weak contact, both teams have allowed way too many groundballs to turn into hits (or errors, or fielder's choices, but one way or another, into non-outs). According to StatCorner, opponents reach base (including on errors) at a .282 clip when hitting groundballs against the Mets' infield, and at a .281 pace when hitting them against the White Sox. Combined, the two units have cost their clubs 51 runs by not getting as many outs when pitchers get grounders as an average defense would.
This article isn't about the Mets or the White Sox. Rather, it's about the fact that the Mets and White Sox, with their combined 51-run infield defense deficit, still have not combined to cost themselves as many runs on groundballs as have the Philadelphia Phillies. The Phillies' infield defense is so bad (how bad is it?) that they've prevented 60 fewer runs on groundballs than an average defense would. Opponents who hit the ball on the ground against these Phillies are reaching base at a .312 clip. Philadelphia also has one of the 10 or so worst outfields in baseball, costing them some 18 runs. Their park-adjusted Defensive Efficiency (PADE) is -2.73, the third-worst in the league, and their unadjusted Defensive Efficiency is a league-worst .678. (Don't hurt yourself trying to match those figures: PADE is expressed as percentage better or worse than average; DE is simply the rate of batted balls turned into outs.)
That's worth mentioning because Cole Hamels is, ostensibly, having a rough stretch in the run-up to a trade deadline that could see him (finally) leave the Phillies behind. Hamels has a 3.91 ERA this year, and in his last five starts, that figure is 7.46. Yet Hamels has a 3.38 FIP (good), an 87 cFIP (very good; 21st-best out of 110 qualifying pitchers), and a 2.56 Deserved Run Average (elite, seventh-best). We've talked at some length about the virtues of cFIP and DRA over the last few months, after we introduced them as our new suite of advanced pitching-value metrics. They can help capture when extreme opponent sets or bad luck are coloring a pitcher's surface-level stats, and they also take care to account for the extra bases some pitchers give away often, and that some never ever give away, like stolen bases and wild pitches/passed balls. They're the best way to truly evaluate a pitcher, and they really, really like Hamels, even if the surface-level numbers don't.
In Hamels' case, though, the truly intricate externalities these stats measure aren't the major issue. Rather, the biggest problem appears to be his teammates. Consider the facts: Hamels is striking out 24.8 percent of opposing hitters this season, which would be his highest strikeout rate since 2012, and is right in line even with that year's number (25.2 percent). He's unintentionally walking 8 percent of opponents, the highest rate of his career, but only negligibly higher than last season (7.8) and right around the league average (7.6). When batters make contact, they're hitting it on the ground 48 percent of the time, which would mark Hamels' highest groundball rate since 2011 (51 percent). When batters are able to elevate the ball against him, they're hitting only 34 percent of those flies to their pull field (the smallest percentage of Hamels's career, and by no small margin). Here's a chart showing how the league does on batted balls to batters' pull fields, to center, and to the opposite field, to lend perspective to that last bit.
Batter Performance on Batted Balls by Direction, MLB, 2015
They haven't been. Hamels has a .307 BABIP, which would be his highest since 2009. The culprit is that atrocious Phillies defense; in particular, the infield on which he has depended more this year than in either of the last two, the infield that has cost the Phillies six wins and counting.
What about that elevated walk rate, though? Indeed, we should talk about that. One (somewhat tempting) explanation could be that Hamels is aware of the weakness of his supporting cast, and is working to keep as many plate appearances within his control as possible. His elevated strikeout rate supports that theory. So does a comparison of his heat map for all pitches from each of the last two years to this year (through July 22nd):
He's clearly burying the ball more, elevating it all the way out of the zone a little more, and trusting the hitting zone a bit less. He's chasing strikeouts, and willing to limit damage by allowing walks instead of hits in a few cases. Still, it's not as though he has lost the strike zone, or is skirting it terribly often:
Cole Hamels, 2012–15
This time, the guilty party is Carlos Ruiz. No catcher has cost his team more runs with poor pitch-framing this year than has Ruiz, according to our numbers. Backup Cameron Rupp rates poorly on that front, too. Hamels is getting squeezed by umpires, largely because his backstops are brutal at their most important job.
Now, there is one notable element that appears to have worked in Hamels' favor to this point. He's faced the weakest aggregate opponent set of any pitcher with at least 80 innings pitched this season. His opposing batters carry a weighted True Average of .253, and his OppRPA+ (which, as the + indicates, sits on a scale where 100 is average, lower is better for a pitcher, and higher is worse) is a league-low 94. In 120 innings, that gap of 6 percent or so between his actual opponent set and the league is probably not indicative of major softness in his numbers. Still, it would be silly not to mention it, because he sits at such an extreme. No matter where he goes at the trade deadline, Hamels will have a tougher challenge before him than he has faced so far, in terms of the guys in the batter's box.
On the other hand, the lack of support from his defenders or his catchers to this point should far outstrip the small break he's caught in terms of opponents. Hamels has been, more or less, his usual, wonderful self this season, albeit with a few necessary modifications designed to help him survive the lack of help his teammates have provided.
We're still hearing anonymous executives hemming and hawing and fretting over Hamels' struggles, of course. That's no surprise. These guys have a real incentive to feed the narrative that recent performance can trump a track record even as impressive as Hamels', because by doing so, they might get him cheaper. Teams aren't actually making decisions based on bad stats over short time periods, and if scouts are coming back from Hamels' starts reporting that he's damaged in any way, they ought to be fired. The numbers show, and common sense says, that Hamels is more or less what he has always been. He's had the bad fortune of playing for a team utterly unable to help him prevent runs this year, and it's showing. There's no reason to expect anything but dominance from him if and when he lands with a club better equipped for that task.
I wrote recently, and still believe strongly, that Ruben Amaro should move Hamels at the deadline. He simply won't be as exceptional an asset, something the market might irrationally price, in any trading period after this one, because of his age, his cumulative workload, and his contract. This is the sweet spot. Rival GMs are feeding notes of concern to Buster Olney and Jayson Stark because they know what Hamels is really worth, and don't want to pay it. They know this is the time of year at which whispers and opinions on single outings can slip past the journalistic filters, because that's what we do during this time of the year. Amaro should call those executives' bluffs, but when someone blinks, he should be ready to pull the trigger.
FanGraphs keeps league batted-ball statistics based on hit direction. StatCorner tracks balls called as strikes, and vice-versa, in addition to the percentages of flies hit to batters' pull fields. Thanks to both for the use of the data.