The Seventh Coming

New York’s Bard of Grime

Focus on Sport/Getty Images

Frame Jobs

What we can learn about manipulating the strike zone from all-world receiver Brad Ausmus, an elite catching prospect, and Dennis Eckersley, the beneficiary of, ahem, a widened target

On a Wednesday in the first week of May 2009, I passed a senior-year stats final, turned in a paper, and took a train home, finished with school forever.1 The next morning, I walked into Yankee Stadium — not through the gates where fans would file in for that night’s game against the Rays, but through the executive lobby, with the stern-looking statue of a still-living George Steinbrenner. It was my first day as a baseball operations intern, and my last day as a catcher defense doubter.

Sports fans spend a lot of time wondering what their teams aren’t telling them. If a player signs for more than his WAR suggests he’ll be worth, we wonder what WAR might be missing. If a top prospect is traded for a disappointing return, we wonder whether his old team became concerned about his makeup or discovered a hidden hole in his swing. If a team acquires multiple players who have something in common, we wonder whether their shared trait might, unbeknownst to the Internet, be the one weird trick that wins pennants. We imagine market inefficiencies everywhere.

In theory, then, a new hire’s first day in baseball ops should be a time of great revelation, the moment when he or she is allowed to leave the cave and learn what was casting the shadows. Based on conversations with friends in front offices, though, that’s usually not the way it works: New employees pick up information in spurts, as they come across it or as projects require that they be brought up to speed. There’s no day-one briefing, no classified PowerPoint presentation that explains why everything the new hire knows about baseball is wrong. As one baseball ops analyst told me this week, “We’ve made certain insights, but when we hire someone, we wouldn’t have that conversation.”

The Yankees couldn’t help having that conversation with me, because I barged in on the middle of it. The week before my first day, a group lunch at Sheppard’s Place, the cafeteria attached to the press box, had led to an exciting discovery. Half the front office sat together and spitballed: director of pro scouting Billy Eppler, director of quantitative analysis Michael Fishman, pro scouting manager Will Kuntz, baseball operations assistant Steve Martone, and Alex Rubin, an intern who had started the previous season. The night before, backup catcher Jose Molina had guided Phil Hughes through six scoreless innings in Detroit, and the conversation turned to Molina’s defensive edge over regular starter Jorge Posada, who often frustrated observers by catching pitches so awkwardly that he cost his pitchers strikes. Could it be, someone wondered, that the gulf between Molina’s and Posada’s gloves could make up the difference on offense between one of baseball’s worst-hitting catchers and one of its best? The consensus was that it wasn’t possible, and the group tabled the idea.

But Rubin — who would eventually2 be hired as a full-time analyst before leaving to work for the MTA as a self-described “transportation sabermetrician”3 — had gotten curious. He was on Team Posada, and he wanted to be proven right. After lunch, while he was supposed to be doing data cleanup, he started researching the size of the strike zone with Molina and Posada behind the plate.

The pitch-tracking data he needed to settle the dispute was then only in its second full season, and almost no public research on the subject existed, save for one post whose author, Dan Turkenkopf, had been so skeptical about his own results that the Internet had largely discounted them. Rubin found a more sophisticated, probabilistic way to model the strike zone than Turkenkopf had, and his results seemed robust. “I was just trying to settle a lunchtime argument with some coworkers,” Rubin says. “And then I stumbled into a discovery that proved me more wrong than I ever could have believed.” No matter what queries Rubin ran, no matter what adjustments he made, he kept coming up with the same answer: The difference between the two catchers was dramatic. Team Molina always won. And the epiphany had arisen in storybook style out of a conversation between statisticians and scouts, a case where the eye test had inspired a statistical test that reinforced what the scouts saw.

By the time I arrived the following week and took my seat opposite Rubin in the drab “bullpen” where the Yankees’ summer intern armada fought for desk space, Rubin had further refined his model, and a framing report had been built on the team’s internal information hub. He had become a convert, and it took only a short time for him to sway me. Before long, I was delighting in looking up precisely how players rated at a game-changing skill that, hours earlier, I hadn’t believed existed in any significant way. It was no secret that Posada was a below-average defender — it was why he’d split time with light-hitting Joe Girardi4 until his age-28 season — but he was a five-time All-Star and Silver Slugger Award winner and had twice finished in the top six in MVP voting. Molina, I had thought, was lucky to be in the league. In one day on the inside, my beliefs about baseball had been shaken in the most amazing way. This was my hallway walk-and-talk with Agent K, my encounter with aliens in the break room. What would the second day bring? Ramiro Pena, MVP?

Later that year, after the World Series was won, Rubin and fellow Yankees quants Dave Grabiner and Jim Logue produced a report that summed up the statistical evidence in a sentence that would have strained belief if not for the inarguable numbers that went with it. “Jorge Posada could hit like Albert Pujols and Jose Molina could hit like Jose Molina, and Molina would still be better,” it said.5 If the technology had been available, I’m sure they would have neuralyzed me when I left a year later.

The experience taught me four things. First, that while most transactions don’t have a secret sabermetric underpinning, it’s not out of the question that some teams, at some times, are acting on proprietary knowledge that could completely change our perception of a player. Second, that such edges are fleeting: In less than five years, framing went from a secret that might have been worth many millions of dollars to one that I can write about openly without looking over my shoulder to make sure a headbanded Brian Cashman isn’t rappelling down the wall. Third, that discovering something about baseball isn’t the same as getting top-to-bottom buy-in and being able to benefit from it. The Yankees, during the window when they may have understood the value of receiving skills better than anyone else, let Molina leave for a division rival and, later, allowed themselves to be outbid for Russell Martin by a team whose previous biggest free-agent buy was Clint Barmes.

The fourth thing I internalized was a respect for the strike zone. Where else on the field could a massive impact be concealed so casually? Last month, I wrote about the strike zone’s inexorable expansion, a significant factor behind baseball’s decreased scoring. There’s no escaping this expansion: It affects every player, and it’s out of any individual athlete’s control. But the strike zone does have its master manipulators: the umpires, pitchers, batters, and catchers who can pull and prod at its edges, reshaping it slightly in a way that can work wonders over thousands of pitches.

Last week, more of those manipulators6 were unmasked through the work of Jonathan Judge, Dan Brooks, and Harry Pavlidis of Baseball Prospectus, who not only extended our insight into the zone back to 1988 — far before the PITCHf/x era — but supplied a new framework with which to evaluate the framers and nibblers we had already identified: Called Strikes Above Average, a statistic that measures each player’s impact on the probability that a caught pitch will be called a strike.7 Their research quantified the contributions of three players whose influence over the umpire had been suspected but not confirmed: a former catcher and current major league manager, a Double-A prospect, and a Hall of Fame pitcher. I asked each of them to explain the source of his Molina-like powers.

NLCS: Cardinals v Astros Game 3Stephen Dunn/Getty Images

Brad Ausmus

Old editions of the Baseball Prospectus Annual — now 20 volumes deep — are time capsules that tell us what statheads thought about players before BABIP, before PITCHf/x, and before framing. My research reveals that they thought nothing nice about Ausmus. In 2003, BP wrote, “Like an old-time medicine man, Ausmus peddles the elixirs of veteran leadership and game-calling skills to an Astros organization all too willing to buy.” And in 2006: “There are few players in the history of baseball who have been as consistently bad and consistently on the field as Ausmus.” By 2008, the 12th time Ausmus appeared in the book, BP was out of interesting insults. “He’s just so bad,” the comment read. “Ausmus’ defense doesn’t come close to making up for his utter lack of offense. He has been killing the Astros for the better part of five years.”

At the time, this was a defensible statement. Over the previous four seasons, Ausmus had been baseball’s fifth-worst qualified hitter, a replacement-level player or worse even according to today’s WAR figures. He was the embodiment of the Nichols Law of Catcher Defense: The worse a catcher’s bat, the better his defensive reputation. Ausmus’s defense was well regarded — he won three Gold Gloves in his thirties — but no one studying the stats could have suspected just how much he was helping.

Ausmus’s elixir wasn’t snake oil: It worked. Over the four-season span leading up to BP’s “He’s just so bad” statement, Ausmus’s zone-expanding abilities were worth 120 runs (including a career-high 39.4 in 2007), and he leads Molina by 43 runs on the 1988-2014 leaderboard. That’s partly a product of playing time — Ausmus ranks seventh on the all-time list of games caught — but it’s also a reflection of his almost unblemished CSAA rate (the red line and right-hand axis below).


These numbers aren’t new to Ausmus, who learned about framing stats from Padres assistant GM Josh Stein during Ausmus’s time as a special assistant with the team from 2010 to 2013. “When this was kind of a blossoming area, we actually sat down together and … tried to identify some of the tenets of what makes someone a good framer or a bad framer, looking at video of current-day catchers and even some past catchers,” Ausmus says. “So I’m very familiar with the premise behind it and the potential value in it.”

When I ask him to tell me those tenets, he ticks them off from firsthand knowledge. “It seems like the catcher who can subtly angle the ball toward the strike zone or receive the ball toward the strike zone gets the call more often,” he says. “Let me give you an example — catchers who receive the low pitch moving slightly upward to catch it toward the strike zone as opposed to catching the low pitch down and away from the strike zone. Catchers who let breaking balls that are up in the zone travel deeper and drop down into the strike zone. I think the [angle of the catcher’s] body on the corners makes a difference … The less the catcher moves in terms of sudden, harsh movements, I think it’s more conducive to getting calls.”

As the manager of the Tigers, a team that’s trying to win now, Ausmus consults framing stats to make decisions at the major league level. In San Diego, he did so with a focus on the future. “We were really trying to look at it more from a developmental and a scouting perspective, so that when scouts go out on the field to look at amateurs where you don’t have these types of numbers, you might be able to recognize some things that they do from a receiving standpoint that would coincide with catchers at the major league level that were proven to get strikes called,” he says. “And then also, we’ve drafted catchers or had catchers in the minor league system, and we’d maybe use some of the things we’ve learned to improve the framing of the catchers that the Padres had at the time.” Although San Diego didn’t draft either player, Padres catchers Rene Rivera and Yasmani Grandal were among the top 10 framers in the majors last season.

Ausmus makes his framing success, like his dangerously sharp jawline and perpetually perfect stubble, sound like a happy accident. “I never was taught or trained myself to try and catch in a way that necessarily got strikes other than trying to make the pitch look as good as possible for the umpire,” he says. “It just so happened that the way I caught translated at the major league level to getting strikes called.” Although he believes that inherent talent dictates framing performance, he allows that it can be improved to some extent, particularly with the use of PITCHf/x as a teaching tool. “I think there’s less opportunity to take advantage of the difference between umpires [today], but I think there’s more opportunity for catchers to take advantage of this new metric,” he says.

Ausmus remained an above-average framer over age 40, though he fell from his peak in his last few seasons. I ask him what he thinks the typical aging pattern looks like for the full catcher population. “I think it must be some degradation with age, I think that’d be normal,” he says. “It wouldn’t shock me, though, and I can’t prove this, that the slighter-stature catchers probably degrade at a little bit of a slower rate just because they haven’t put as much stress on their body. Jose Molina would be the argument against that.”

Molina recently underwent knee surgery, which threatens a career that was already limping along, but he’s 39 and cartoonishly catcher-shaped, so that breakdown isn’t surprising. Ausmus is right, though, that Molina’s framing ability hasn’t deserted him. In fact, Ausmus is right in every respect. The following graph shows the aging curve for catcher framing skill. At the extremes, the samples are too small to mean much, as the diverging gold and blue confidence-interval lines indicate — age-45 Carlton Fisk is a sample unto himself. However, there is a clear, if gradual, decline over time.


If we segregate the sample by height, another possible pattern appears. Mouse over the GIF to see the aging curves for catchers 73 inches or shorter (423 players, 1,720 seasons) and catchers taller than 73 inches (265 players, 1,094 seasons). (Ausmus, at 71 inches, and Molina, at 72, belong to the former group.)

The shorter catchers have a higher average CSAA and an almost imperceptible decline through the mid-thirties. The taller catchers peak lower and tail off earlier. Shocking, I know: Ausmus is instinctively good at intuiting aging curves, also.

In the spring of 1994, Ausmus’s first full season, Padres manager Jim Riggleman said, “Brad can hit .220 and have a long career in the major leagues.” Pitcher Doug Brocail thought Riggleman had been too conservative, adding, “I don’t care if [Ausmus] hits .110, he can catch for me any day of the week.” Hyperbolic as at least one of them was trying to be, those comments, in retrospect, are more justifiable than the ones that would soon be published by BP. “There’s no question my defense is what kept me in the game,” Ausmus says. “In terms of having this catching metric, obviously it wasn’t around for the vast majority of my career. Back then, it was more scouts and baseball people saying, ‘Oh, he receives the ball well, not a lot of movement when he catches.’” And, of course, statheads mistakenly scoffing in response.

San Diego Padres v Milwaukee BrewersChristian Petersen/Getty Images

Austin Hedges

Of all the players on this year’s Baseball Prospectus Top 101 Prospects list, Padres catcher Austin Hedges had the worst offensive season in 2014, hitting .225/.268/.321 for Double-A San Antonio. And yet Hedges fell only five spots in the rankings since last year’s list, from 18 to 23. That’s a reflection of scouts’ regard for his defense, which is fully supported by the statistics. Minor league pitch-by-pitch data extends to 2006 for Triple-A and 2008 for Double-A, so we don’t have a huge CSAA sample, but Hedges’s 37 runs saved last season are tied for the most on record below the big leagues, nearly doubling the second-highest sum in 2014.

Name Year Level(s) Runs
Austin Hedges 2014 Double-A 37
Jose Lobaton 2008 Double-A 37
Jonathan Lucroy 2009 Double-A 33
Erik Kratz 2011 Triple-A 28
Derek Norris 2012 Triple-A 26

Hedges, who says he admires and tries to mimic the receiving of Yadier Molina, Jonathan Lucroy, and Rivera (the first-, second-, and fifth-ranked catchers in career CSAA rate), pays close attention to major league framing stats, but before we spoke, he wasn’t aware that any attempt had been made to measure minor leaguers. Unlike Ausmus, he gives credit for his framing skill to his coaches — including Ausmus, who worked with Hedges as a roving instructor. “I think he had such great forearm strength and he had such a good idea of how to really lead the ball to the spot that he could just stick it without any movement at all in his glove,” Hedges says.

“Probably the most important thing that [the Padres] preach from the catching side … they want you to catch the ball the right way and steal strikes for your pitchers,” Hedges says of the message he’s heard from on high. “As long as I’m doing my job with the pitching staff, then that’s really the most important thing they care about.” No wonder A.J. Preller has traded his best big league framers this winter: He has a better one on the way.

Dennis Eckersley

“If Dennis Eckersley was on the mound, a fastball four or five inches off the plate was a strike,” Gary DiSarcina told the Los Angeles Times in 2011. I read that quote to Eckersley, and he bristles a bit, muttering a few unprintable words about DiSarcina’s offensive abilities. “I’m the one who gets to say it, but I don’t want somebody else to say it,” he says.

Evidently it’s OK if an algorithm says it, because when I tell him about his high CSAA, he owns up to his supersize strike zone immediately. “I know the zone is the zone, but I remember when I was pitching the ninth inning, if I hit the target, it was a strike,” he admits. “It’s almost like the umpire knows you’re going to do it, and maybe you get a little bit more.” He pauses, and I can picture him replaying decades-old images of outside strikes in his mind’s eye. “I’d be the first to not want to say that, but I would say that,” he says. “You want to take the credit for throwing strikes all the time, and that’s a big part of it. But I think I had a bigger strike zone.”

Those extra strikes weren’t a coincidence. By 1988, Eckersley was exclusively a reliever, and once he’d given up starting, he was almost always around the plate. Since ’88, only two pitchers with at least 500 innings pitched have walked fewer than 1.5 batters per nine innings: Eckersley, at 1.30/9, and Bob Tewksbury, at 1.33/9. Both appear on the list of highest career CSAA rates (minimum 4,000 pitches); Eckersley’s CSAA is so high — a 10 percent swing in strike probability — that he’d take the top spot by a wide margin even with all pitch minimums removed.

Dennis Eckersley .102
Cliff Lee .078
Steve Farr .076
Rick Reed .076
Doug Jones .071
Scott Sanderson .071
Bert Blyleven .070
Doug Fister .068
Paul Quantrill .067
Bob Tewksbury .067
David Robertson .064

Lee, Reed, Jones, Blyleven, and Fister all appear in the top 20 on the post-’88 walk-rate leaderboard; Sanderson ranks 30th. Clearly, larger strike zones are the privilege of the command pitcher.

“It’s not like, ‘Oh, wherever you throw the ball, it’s going to be a strike,’” Eckersley says. “It’s not like it’s a gift. I could paint. If you could paint, you’re going to get an inch or two.” Eck did most of his painting low and away — an almost unhittable location if a pitcher can hit it consistently, and he could. “I could do that all day long,” Eckersley remembers. “If you dot that thing, it’s hard to tell whether it’s on the corner or not … You hit that dot, it’s like, man. Next thing you know, you have a reputation.”

Ausmus confirms the command theory: “For a catcher, familiarity with a pitcher becomes extremely important, and obviously if they’re locating, it becomes much easier to frame and receive the ball.” Hedges is looking forward to worrying less about wildness. “I think as you climb the ladder and you get to Double-A, Triple-A, the big leagues, every pitcher has at least some sense of control,” he says. “Even though their stuff might be better, the fact that they have at least somewhat of an idea of where it’s going does make my job a lot easier. The erratic guys are the ones where you really have to work hard back there.”

Eckersley doesn’t think he could have had such unerring aim without sacrificing velocity — a worthwhile trade-off, if his CSAA is any indication. “I threw hard when I was young,” he says. “But when I was relieving, I was already in my mid-thirties. To have that kind of control, you can’t throw ultra-fast … If you’re off a tick, man, you’re in trouble.” He likens the smaller margin for error to pitching before his reputation took root.

“It’s like being a rookie; you’ve got to be better,” Eckersley says. “I remember as a rookie I threw it right down the middle and he called it a ball. I was like, ‘Are they testing me? What are they doing?’ That plays to being around the block, being a veteran, having a bigger strike zone.”

Ausmus backs up Eck’s contention. “The pitcher’s résumé — how well-regarded or talented the pitcher is — can often play into whether strikes are called or not,” he says. “If you have Roger Clemens on the mound, you have to get strikes called that might be further from the center of the strike zone than if you have some rookie call-up on the mound … I think, as a general rule, umpires respect guys who have been around a long time, because they’ve proven that they can pitch at the major league level.”

Turkenkopf saw this in the stats several years ago, concluding that “in general, the older you are, the more love you get from the umpires.” The effect of experience was even stronger.

More recent studies offer additional evidence. It’s possible that survival bias explains some of the effect that Turkenkopf found — it might not be that umps grow fonder of players over time, but that players who know how to wheedle extra strikes tend to stick around longer, skewing the sample. And given the changes in the strike zone since the early days of PITCHf/x, the umpires’ preference for veterans might be fading in the face of stricter standards. “The whole ‘earning your stripes’ thing has diminished somewhat with the more strictly enforced strike zone,” Ausmus notes.

If there’s a lesson to be learned from the uncannily accurate masters of the strike zone, it’s that the players and their employers often have something to teach us. Sometimes they spout nonsense, and it’s appropriate to respond with silence or snark. But once you’ve been burned by one Ausmus or Molina, it’s easier to keep an open mind about the next perplexingly long-lasting player whose value might manifest in mysterious ways. 

Thanks to Harry Pavlidis of Baseball Prospectus for research assistance and to Grantland intern Alex Shultz for transcription help.

Filed Under: Houston Astros, MLB, San Diego Padres, Pitchers, Brad Ausmus, Pitch Framing, Austin hedges, Dennis Eckersley


Ben Lindbergh is a staff writer at Grantland.

Archive @ BenLindbergh