The year was 2009, and Colorado Rockies fans were settled deep into one of two metaphorical trenches opposing one another over a vast hellscape cluttered with broken numbers and inundated with a hazy fog of grit. In one trench waved a banner bearing the name “Iannetta,” and in the other: “Torrealba.” These were the Catcher Wars, and there was no end in sight.
Catchers fascinate me. Always have. That was my primary position throughout my young life in organized baseball. When I was 10 years old, it was mostly about getting to put on all that armor – it made me feel like a super hero. As a pre-teen, it was mostly about staving off boredom – catchers get to touch the ball dozens of times an inning. And as a teenager, I went all-in on the more high-minded aspects of the job – deciding when to call for that curveball, managing those poor pitchers’ fragile psyches, positioning my teammates… and even pitch-framing, a practice the adult-version of me has come to disdain.
But however much I’d like to see pitch-framing be displaced by an automated strike zone, I have to admit that pitch-framing intrigues me. Fooling umpires isn’t the sole domain of catchers, of course. Moreover, umpires wholly unaffected by any player’s tomfoolery issue faulty decisions entirely on their own accord. It’s a really hard job, after all. But among all the shouldn’t-be-but-are aspects of baseball – the basket of baseball goodies we often refer to as the “human element” – the thing about pitch-framing in particular that always piqued my interest is how little we knew about its impact on the game. Butterfly effects notwithstanding, we can usually wrap our heads around the immediate impact of a blown call at first base. The effects wrought by a catcher’s frame-job turning a would-be 2-1 count into a 1-2 count – which basically turns Josh Donaldson into a AAA hitter – are much more difficult to appreciate (even if empirically enormous). And most of the time, despite modern technology and on-screen graphics of various sponsorship, we don’t even know when a catcher’s frame-job influenced the umpire in the first place.
Steering every conversation about baseball to yet another pitch-framing rant is the current internal fight of my life, and I won’t indulge myself here. I mention pitch-framing only because it serves as perhaps the best example of an entire suite of ethereal attributes that make catchers so damn interesting. Take pitch-blocking, for example. We have stats that have been around forever, but counting passed balls and wild pitches is woefully inadequate. First, distinguishing between the two is often entirely subjective. Secondly, many balls are thrown in the dirt on purpose, especially in 0-2 counts, and in those cases, the pitcher and catcher are in it together. An unblocked ball thrown in the dirt might be scored a wild pitch, but that doesn’t mean the blame (and bad stat accumulation) should be borne by the pitcher.
Controlling the running game is also oversimplified by the surface-level catcher stats we’ve clung to for so long. Historically, we’ve put stolen bases on the catcher, but insiders have always known that the pitcher has as much to do with stolen bases as the “gun” behind the plate. And modern statistical analysis has proven it. However, as described above with respect to pitch blocking, knowing that the credit or blame implied by a stat like caught stealing is misleading only gets us half way there. We have to realize that the best throwing catches discourage runners from even trying in the first place.
The point is this: catchers, and the impact they have on the game, are quite hard to figure out. Even if we could identify all the ways a good catcher makes a difference, we still haven’t had a very good way to measure them.
Well, that is until Baseball Prospectus saved our baseball souls. A little more than a year ago, they released the best set of catching metrics to date – metrics that got us closer than ever to the “truth” about catching defense. Rather than relying on PitchFX data, the method uses a “Regressed Probabilistic Model.” There are a few different reasons this is a better approach – and I’d encourage you to read in the entire piece linked above if you’d like to dive in – but perhaps the strongest point in favor of BP’s model is to consider Tom Glavine. He got lots of “extra” strikes in his career that had nothing to do with his catchers. If you watched baseball in the 90’s, you know this. And the data confirms it. BP’s model accounts for factors like Tom Glavine when rewarding catchers for framing.
Anyway, tackled first by BP were the areas of pitch-framing and pitch-blocking. In the wake of that shared epiphany, I wrote up an analysis of the Rockies catching options entering the 2015 season. A few months later, as part of Baseball Prospectus’ just-as-revolutionary debut of its new pitching metric, Deserved Run Average, new catching metrics related to the control of the running game became available.
At that point, other than the secret sauce, our catching metrics plate was full. There was much statistical merry-making and all was well. But then, this past offseason, Baseball Prospectus served us dessert – a decadent concoction of modern understanding and historic nostalgia. For the first time ever, we could test our assumptions about the defensive prowess of baseball’s greatest catchers of the past against cold hard data.
Of course, as a survivor of the Rockies Catcher Wars, my thoughts stomped right past Johnny Bench and Mike Piazza and straight on to Chris Iannetta and Yorvit Torrealba. At the time, the opposing sides in that war could be generally described as analytics-lovers who backed Iannetta (“That walk rate! That isolated slugging!”), and the metaphysicals who backed Torrealba (“That presence behind the plate! That clutch!”). I’m generalizing, of course, but it really did seem to at time to boil down to a pretty straight-forward argument between “head” and “heart.” Personally, I myself liked Torrealba better – I rooted for him more instinctually – for the same reasons everyone else liked him: he was a joy to watch, was adored by his teammates, and, dag-nabbit, the dude really did seem to be clutch. I bought a Torrealba shirsy back then that I still wear proudly on occasion. But in the end, it was Iannetta I wanted behind the plate, because I wanted to win. I believed in things that could be measured, and so I was pretty much convinced that Iannetta was better.
From 2007 through 2009, the three-year period during which Iannetta and Torrealba split Rockies catching duties (almost exactly, 991 PAs for the former and 946 for the later) Iannetta accumulated 5.2 WAR to Torrealba’s 0.7 (using Baseball Reference’s measure). The difference was almost entirely made up of Iannetta’s superior offensive production, but even isolating defensive WAR by itself, Iannetta was better: 1.8 dWAR to Torrealba’s 1.5.
But Baseball Reference’s WAR numbers – which use the Fielding Bible’s Defensive Runs Saved (DRS) for defense in years 2004 and beyond – couldn’t really capture the true art of catching. True, DRS is a noble effort to capture what happens in the field, and even as a skeptic of zone-based defensive metrics, I think DRS is better than having nothing at all. DRS even makes an effort to account for framing as part of a larger estimation of a catcher’s impact on his pitcher’s ERA, albeit quite indirectly. But it’s still a defensive metric whose own makers concede there’s a lot it doesn’t capture, at least and especially with regards to catchers.
Many Rockies fans just knew that Torrealba was better behind the plate, and even as an Iannetta supporter, I admit that it really did seem that way. Apart from the notion of “clutchness,” nobody disputed that Iannetta had more to offer with the bat. But when even an Iannetta supporter could look at those defensive numbers and reasonably conclude that they didn’t pass the eye – or smell – test, it’s easy to see how the Catcher Wars erupted.
And by the following offseason, after the Rockies last playoff run in 2009, they were over. Torrealba left as a free agent, rejecting a Rockies offer to take equivalent money from San Diego, where he, ironically, would split duties with a young Nick Hundley. Iannetta lasted a couple more ho-hum seasons in Colorado before being traded for Tyler Chatwood. The Catcher Wars probably don’t matter much anymore.
Except that they probably do. Remember, that 2007-2009 era represents the team’s historical apex and included two trips to the playoffs. At no other time in the team’s existence would a marginal difference have mattered more. So, with the benefit of greater analytical hindsight, can we settle this thing once and for all?
Yes, I think we can. But before I deliver the punchline, I’d like to share the rack-up of all Rockies’ catchers who ever received meaningful playing time for the club. It was fun combing through this data in researching the piece and, since I’ve already built the spreadsheet, I’d like to share it with you. For these purposes, I’ve defined “meaningful” in such a way as to capture all Rockies catchers who have served either as starters or primary backups in any single Rockies season through 2015. These Baseball Prospectus metrics (for framing, blocking, and throwing) are all expressed in runs above or below league average. The last metric, FRAA, stands for Fielding Runs Above Average, which is BP’s cumulative defensive metric and, for catchers, represents the sum total for framing, blocking, and throwing. I’ll omit Iannetta and Torrealba for now to keep you in suspense for one more paragraph.
|Name||Seasons||Innings||Framing Runs||Blocking Runs||Throwing Runs||FRAA|
|Sandy Alomar Jr.||1||269||-6.8||1.3||0.2||-5.3|
A few interesting notes:
- You can see how much pitch-framing matters – both in absolute terms and relative to other catcher skills. Everyone is getting wise on this subject now, and there’s evidence, including some supplied by Chris Iannetta himself, that framing will soon cease to be the differentiating factor it once was. But we’re not there yet, and we certainly weren’t there “back in the day.”
- You often hear analytics model-builders say that it’s nice when the numbers represent a fair amount of both surprise and confirmation of what you already thought you knew. The former adds the value, and the later helps ensure you’re on the right track. In this case, Wilin Rosario registering as the worst pitch-blocker in the history of the team, and by an amount that puts him so thoroughly in a category unto himself, falls squarely in the latter category.
- In general, I think the scores line up with each player’s reputation (for those who had reputations at all). Joe Girardi and Charles Johnson might be exceptions, but even in those cases, it’s mostly the pitch framing component that sinks them. We just didn’t pay as much attention to that aspect of the job back then.
- This isn’t listed on the chart, but to put the Rockies into context as a team, Colorado catchers have put up a cumulative 328.4 runs below average in the club’s history, which works out to a little over 14 runs per year. In other words, bad catcher defense has cost the Rockies about a win and a half per season.
But enough of the also-rans. Let’s get back to our controversy. Where do Iannetta and Torrealba slot in on the table? Well, as narrative luck would have it, right at each end:
|Name||Seasons||Innings||Framing Runs||Blocking Runs||Throwing Runs||FRAA|
Torrealba is the best defensive catcher in the team’s history. Iannetta is its worst. Huh.
And the thing is, those FRAA totals actually undersell the difference between the two during the Catcher Wars specifically. Torrealba’s total is deflated by a bad 2006 season when he split time not only with a rookie-year Iannetta, but also with JD Closser and Danny Ardoin; and by a similarly bad 2013 season when Torrealba came back to Colorado for a reunion season before retiring at 34. In the 2007-2009 period, Torrealba racked up 20.4 FRAA, including 19.5 in 2008 and 2009 in particular. Torrealba’s 2008 and 2009 were the two best FRAA seasons any catcher has registered in Rockies history. Turns out those Torrealba supporters really did see something special going on.
Iannetta, on the other hand, cost the team 34.6 FRAA from 2007-2009. The difference over those three years? 55 runs.
But that’s just defense. Remember, Iannetta was a better offensive player. As it turns out, Torrealba was indeed more “clutch” in those years, at least according to Fangraph’s metric, by a Clutch score of 1.49 to Iannetta’s 1.26. But “clutch,” in this sense, means only that Torrealba elevated his game from his baseline performance by a slightly larger amount than did Iannetta. That has as much to do with Torrealba’s lower baseline performance as it does anything else. Iannetta still contributed much more on offense, even in clutch situations, as evidenced by Iannetta’s Win Probably Added (WPA) of 2.90 compared to Torrealba’s WPA of -2.04. WPA puts offensive performance into it’s actual context – consider it to be a much, much, much better version of the RBI, if you’d like. So Iannetta did more to produce runs than Torrealba did, both in absolute terms, and in high leverage. Period. (This difference between being “clutch” and actually being good in clutch situations is subtle but important one – I wrote a big ‘ole piece about it last season if you’re interested.) During the Catcher Wars, Iannetta bested Torrealba in batting/base running runs above average (all players, not just catchers) by 4.5 to -37.1 for Torrealba, a net difference of 41.6.
That’s a lot, but not enough to make up Torrealba’s 55-run advantage on the defensive end of the equation. Iannetta was actually a slight better thrower, and because “gunning down” base-runners is perhaps the sexiest and most visible of catching defensive attributes, this might explain at least part of the reason why the pro-Torrealba arguments rang hollow for so many fans back then. Torrealba, meanwhile, was the better pitch-blocker, but in both cases, the differences between the two were quite small. Over the three year period, Iannetta gains 3.5 runs on Torrealba with the throwing, and Torrealba gains 2.3 of those runs back with the pitch-blocking. The difference between these two fine men was almost entirely due to pitch-framing: a whopping 56-run advantage for Torrealba.
And so, the final Catcher Wars victor, by Baseball Prospectus’ WAR, which has been updated for past seasons to reflect this new insight into catching defense…
Yorvit Torrealba: 4.5 WARP
Chris Iannetta: 3.5 WARP
Yet the Catcher Wars did end, and as many wars do, they ended with a whimper. The Iannettas finally declared a hollow victory as The Torrealbas left the field head held high, and only years later did the world learn that it was no victory at all. For the ultimate irony of the Catcher Wars was that The Iannettas perceived strength was in their numbers, yet it were numbers that ultimately vanquished them.