Overthinking It: What ... (01/10)
28 comments have been left for this article. BP Comment Quick Links
| |
|
|
January 17, 2012 Overthinking ItJorge Posada, the Hall of Fame, and the Fog of WARPIt’s good to be a baseball fan in the 21st century. Not only is it easy to keep up with the action-packed offseason (which for the Yankees, pre-Pineda/Kuroda, amounted to re-signing Freddy Garcia, Andruw Jones, and Rick Down) at sites like this one, but thanks to the wonders of modern-day technology, we no longer have to wheel out a motorcycle or a piece of army ordnance every time we want to find out how hard a young pitcher throws. We also spend a lot less time arguing about things that aren’t subjective. In the 20th century, debates about velocity went something like this:
Yellow journalist 1: Who throws harder: Jack Pronto or Jack Celerity? Now, they go more like this:
Blogger 1: Who throws harder: Justin Verlander or David Price? I prefer concrete answers to pretending we can distinguish between the fastest and second-fastest fastballs with the naked eye, but I can see why some might worry that we’ve lost something in the process, that baseball was more romantic before so much of it could be quantified. If you feel that way, you might be pleased that pitchers who predated PITCHf/x are forever out of reach of Sportvision’s cameras (though you’ll probably want to cover your ears* when FIELDf/x starts telling us which players spit the fastest sunflower seeds). *Because FIELDf/x will speak, like a computer from Star Trek. Even if you enjoy the numbers, though, today’s unprecedented availability of baseball data can be bittersweet. Future generations will be able to compare the pitchers they grew up watching to those of their parents’ and grandparents’ eras, right down to the spin rate, pitch type percentage, and release point. But we’ll never have access to the same rich information on past players, for some of whom we lack basic box scores, let alone play-by-play accounts and motion-tracked movements. Wouldn’t it be nice to know how Barry Zito’s curve compared to Darryl Kile’s, Bert Blyleven’s, or Sandy Koufax’s in a more-than-anecdotal way? Even if it meant supplementing our memories with more mundane, but more reliable, measurements? You might be wondering what all this has to do with Jorge Posada (if you haven’t long since stopped caring). Since Posada played for 17 major-league seasons, hanging on until after his 40th birthday, he was still around as the PITCHf/x system was installed in major-league parks beginning in 2007, giving us partial PITCHf/x coverage of his career. As a result, we know far more about Posada’s play than we do about that of the players who predated him, but we still lack the near-perfect information we’ll have for players who spend their entire careers being watched by the big-league equivalent of Big Brother. As Posada prepares to announce his retirement, recaps of his career and assessments of his Cooperstown candidacy are popping up all over the internet. The information we have is enough to tell us that Posada was an excellent player. But it’s the information we’re missing that might just put him in the Hall of Fame. ***
A few weeks ago, Joe Posnanski wrote this:
According to Posnanski, most of the Hall of Famers we tend to think of as inner-circle, first-ballot members—the men who might’ve been unanimous selections, if ever there were such a thing as a unanimous selection—excelled in all three areas. A tier below the Willie Mays, Hank Aaron, and Rickey Henderson types are the players who had two of the three categories covered. At least half of the Hall’s position players, Posnanski says, fall into one of those groups. The remainder—aside from some questionable members whose only special talent was timing their careers to coincide with that of Frankie Frisch—did only one thing well, but did it so well that they earned admission anyway. Were he to be enshrined, Posada would owe his plaque entirely to his bat, which was every bit as good as you remember it being. If we could make direct comparisons between players in different eras and parks, making a case for Posada might be as simple as observing that he had a higher career OPS than Yogi Berra (848 to 830). Of course, that sort of sloppiness sets off all sorts of alarms at Saber HQ. But even after we make the appropriate adjustments for offensive environment, Posada’s bat comes out looking Cooperstown-caliber. His .290 career True Average—a BP all-in-one offensive statistic that puts all players on an even footing—is a dead ringer for the .292 standard among Cooperstown catchers. Had Posada been anything close to average in the field and on the bases, he’d be an extremely strong candidate. But Posada wasn’t one of those players who excelled in one area and was mostly harmless in the others—outside of the batter’s box, his game was weaker than his chin. For the most part, unless he was sitting on the bench or standing at the plate with a bat in his hands, Posada was actively hurting his teams. It’s not surprising that Posada was a bad baserunner. Catchers, as a group, aren’t fleet of foot, both because guys who can’t run gravitate to a position that doesn’t require much movement, and because countless hours of crouching and wearing heavy gear in the hot summer sun don’t tend to improve foot speed. Since 1950, only one catcher, Jason Kendall, has been at least two wins better than a league-average baserunner, according to BP’s Baserunning Runs. Thurman Munson is the only other catcher to top one win. Posada, though, was a bad baserunner even for a backstop. He wasn’t quite Bengie Molina bad, but he played long enough to do more damage. As a result, he cost his teams over five wins with his legs, the highest total among catchers and the second-highest at any position. To accumulate the most of anything in baseball—even the bad things—you have to be a pretty good player. Truly bad players don’t stick around long enough to sink to the bottom of the leaderboard. In fact, the 10 worst baserunners in the BP database include Hall of Famers and aspiring Hall of Famers like Frank Thomas, Harmon Killebrew, Edgar Martinez, Willie McCovey, and Eddie Murray, with Jim Thome, Mike Piazza, Manny Ramirez, and Mark McGwire not far behind. All of those players, Posada included, were good enough to overcome their bad baserunning. The same bulk that slowed them down made their batted balls go farther—often far enough that they could afford to trot, rendering their lack of speed inconsequential. Again, much of what we know about players comes down to the quality of the information we have. If Posada had played before 1950, we would have known he was unsuccessful on the rare occasions when he tried to steal, but not that he hemorrhaged runs on the bases at other times, too. As it happens, he came along after 1950, so we have a good handle on the extent of his baserunning blunders. But that doesn’t mean we’re not missing other information that might paint him in an even less positive light. We’re coming to that. *** If you’ve been reading your Jay Jaffe, you know all about JAWS, the system Jay designed to assess how players measure up to those at the same positions who are currently in Cooperstown. Jay tallies up each player’s career Wins Above Replacement Player (WARP) and peak WARP, then averages them to come up with one number. (WARP, by the way, includes baserunning.) As Jay wrote last week, here’s how Posada compares to the average HoF catcher:
According to JAWS, he's awfully close. He gets even closer when you consider his playoff record. Last season’s ALDS notwithstanding, Posada was not a great post-season performer, but he did play the equivalent of a full regular season in his 15 Octobers, making major contributions to four championship teams. That seems like it would be enough to push him over the edge. Except. Until fairly recently, measuring individual players’ defensive contributions was considered an impenetrable puzzle. Measuring catchers’ defensive contributions was—well, more impenetrable, if possible. But things have changed. No, we haven’t yet solved the puzzle. Our fielding statistics are still less dependable than our offensive ones, and that probably won’t change without FIELDf/x (which may never go public). But we do have a better grasp of defensive performance than we used to, catchers included. *** Currently, Fielding Runs Above Average (FRAA), our proprietary fielding metric, factors in only a catcher’s effect on the running game and his skill at fielding the few batted balls that fall within his reach. At those two aspects of catching, Posada grades out as average. (The average HoF catcher is, not surprisingly, above average, by about nine runs.) But FRAA doesn’t come close to capturing the entirety of a catcher’s work behind the plate. For one thing, it doesn’t account for pitch blocking. Bojan Koprivica’s research using—what else?—PITCHf/x revealed that Posada was the second-worst catcher at blocking pitches from 2008-2011, costing the Yankees over five runs (roughly half a win) per 120 games with him behind the plate. The impact of framing pitches to get favorable calls isn’t yet included in FRAA, but that doesn’t make it any less important. In fact, Mike Fast’s research last year suggested that it’s far more important than most people had ever imagined. According to Fast’s results, the difference between the best and worst catchers at framing pitches in any given season is something on the order of four to five wins. And not surprisingly, Posada has been among the worst, if not the worst, during most of the seasons for which we have data. From 2007-2011, Posada cost the Yankees an average of almost .003 runs per called pitch. That doesn’t sound like a lot, but catchers catch a lot of pitches, and all those fractional losses add up. For Posada, they added up to 50 runs, the third-worst total among catchers, and nearly as much as his baserunning cost the Yankees over the course of his career. Between his poor framing and blocking performances, Posada’s work behind the plate might have meant death by a thousand cuts for lesser teams, at least in his later years.* *Fortunately for Yankees fans, Russell Martin is among the best framing catchers, which goes a long way toward excusing his middling, non-Posada-like bat. As I tweeted last week, between Posada’s forced retirement and Jesus Montero being sent to Seattle, it’s a bad time to be barely a catcher in the Bronx. This is where Posada’s timing may prove to be his best ally. Without factoring in framing or blocking, Posada looks like a borderline candidate for the Hall. Inducting him wouldn’t noticeably shift the Hall’s standards in either direction. Subtract several wins for his dismal defense during the PITCHf/x era, and suddenly he looks like a stretch. But imagine how tenuous his case might appear if we could peer further into the past.
"So, wait, you're saying you haven't even been trying "Oh, sure, I guess you think I should've just pretended to get calls?" (Image h/t RotoExperts) those were strikes." (Image h/t Newsday) I won’t go into the math, since this is all too suppositional to honor with actual numbers. But if we extrapolate Posada’s framing and blocking performance from 2007-2011 across his 12,877 career innings caught and 115,467 career called pitches behind the plate, almost all of his offensive advantage is erased. He becomes a catcher who hit like a Hall of Famer but ran and fielded so poorly that the sum of his contributions would be no greater than those of an average catcher over the same period. He has no Cooperstown case.* *In fact, if Posada’s defense was anywhere near as inept in his youth as it was later on, he might have been more valuable to the Yankees as a DH, despite the much higher offensive bar he would have had to clear. That probably wouldn’t have helped his Hall of Fame chances, though, considering the voters’ lukewarm welcome for Edgar Martinez. There’s also the fact that Posada hit .225/.320/.375 in 180 games and 675 plate appearances as a DH, though many of them came when he was old and/or infirm. If he’d been a DH from the start, he might have warmed to the idea, and the reduced wear and tear on his body might have made his bat even more potent and resistant to age. Is it fair to extrapolate a player’s defensive performance during his late-30s across his entire career, to assume he was no better at fielding when he was in his prime? Of course not. Posada never seemed like an above-average catcher, but it’s safe to assume that he was a better blocker, at least, when he was younger and at least marginally more agile. As for his framing, it’s difficult to say; with only a few years of data at our disposal, we have almost no notion of when framing peaks, or even whether it’s a skill that improves or declines with age. Maybe it’s more dependent on reaction time, in which case it’s no wonder he’d lost his touch by the time PITCHf/x appeared. Maybe it’s more dependent on experience and knowing how to manipulate the men in blue, in which case he might’ve been worse when he first came up. (That would explain why he remained a part-time player long after it became apparent that he could outhit Joe Girardi with one urine-stained hand held behind his back, preventing him from recording even more impressive offensive totals.) We don’t know exactly how Posada’s fielding affected his teams prior to 2007, and we probably never will. If his career had begun a few years earlier, his defense might never have been subjected to this sort of scrutiny. If it had begun a few years later, his chances of induction might be even more slim. By the time Posada’s eligibility for induction expires, a full two decades from now (assuming he doesn’t drop off the ballot before then), our understanding of catcher defense will have been further refined, and the voters will be more inclined to trust the statistics. We’ll know more, and we’ll wonder less. Of course, we’ll never know everything. But, whether you like it or not, there might be a little less room for debate.
Ben Lindbergh is an author of Baseball Prospectus. Follow @benlindbergh
|
I am a Yankees fan but think this article is great. Posada's D was pretty bad and, in my mind at least, this stops him getting into the HOF.
I'd second this. His defense tortured me for his entire career. Worst pitch framer I've ever seen.
I find it funny how Posada's defense is "bad" and Bernie Williams defense is "good but with a weak arm" to people who saw them both play. I also find it funny how people, even some Yankees fans, promote Bernie as a Hall of Famer but dismiss Posada when, given the choice, I would've rather had Posada since it's much harder to find switch-hitting catchers with power and high walk rates.
Also..."According to Posnanski, most of the Hall of Famers we tend to think of as inner-circle, first-ballot members—the men who might’ve been unanimous selections, if ever there were such a thing as a unanimous selection—excelled in all three areas. A tier below the Willie Mays, Hank Aaron, and Rickey Henderson types are the players who had two of the three categories covered"
Last I heard, Henderson wasn't an excellent defender.
Williams won four consecutive Gold Gloves, which probably contributes to that "good but with a weak arm" impression. And while he ranks as one of the worst defensive outfielders ever according to some metrics, FRAA sees him as only slightly below-average (-16.2 career). For what it's worth, I don't think he's a Hall of Famer either, even though he was one of my favorite players growing up and was better in the postseason than Posada.
As for Henderson, I don't know what the prevailing opinion of his defense is, but he finished with 144.5 FRAA, the most of any left fielder since 1950 by over 30 runs.
So you wanna use FRAA?
Jorge's at +12.7 for his career.
That would refute the "Had Posada been anything close to average in the field and on the bases, he’d be an extremely strong candidate." paragraph.
That's one of the many reasons (listed elsewhere) that I hate FRAA because a +12.7 FRAA for his career implies that he was close to average, especially if -16.2 for Williams is considered "slightly below-average".
You can't use FRAA to support Williams yet dismiss FRAA to denigrate Posada.
I think it is fair to cite FRAA in one instance and not the other, since FRAA accounts for a smaller portion of a catcher's performance than it does for center fielders.
The FRAA we have on the site for Posada does show him as slightly above-average, but the catcher FRAA currently displayed basically accounts only for performance in fielding batted balls, which is a pretty minor part of what a catcher does defensively. Colin has worked on a version that incorporates throwing and performance against the running game, which, in Posada's case, erases those positive runs, taking him back down to average. But I don't think we have to pretend that Posada was average defensively just because the current version of FRAA says he was. We know catcher FRAA omits important things, and we know Posada wasn't good at those things.
Um, FRAA has been used to take away and give runs across the board, whether in a JAWS context or a WARP context, so, no, you can't cite it in one instance and not another.
Also, you seem to imply things like performance in the running game are already incorporated "Fielding Runs Above Average (FRAA), our proprietary fielding metric, factors in only a catcher’s effect on the running game and his skill at fielding the few batted balls that fall within his reach. At those two aspects of catching, Posada grades out as average." So, if FRAA is already accounting for Posada's performance in the running game and giving him a +13.7 for his career, then what exactly is erasing those positive runs? Is there another calculation besides FRAA going on behind the scenes?
And if FRAA says one thing but "observation" says something different and you defer to observation, then why use FRAA at all? That's the point of having a metric, right? To provide an unbiased evaluation based on certain criteria and calculations.
Richard, I'm not sure I understand what you're saying in your first paragraph. I still think it's fair to give more credence to FRAA in a center fielder's case than a catcher's, since CF FRAA and C FRAA aren't created equal. Both measure fielding, but FRAA for catchers is a much less comprehensive measure. I don't know if they're apples and oranges, but they're certainly apples and, oh, apple peels.
The FRAA total on Posada's player card does not factor in his throwing. As I said in my comment (and as Jay explained in the article of his that I linked to above), that's an improvement that Colin is working on but hasn't yet made public. Once you factor that in, the +12.7 runs that the public version of FRAA currently credits him for are wiped out.
As for your last paragraph--I don't think I am deferring to observation or contradicting FRAA. FRAA is agnostic on the matter of whether Posada is a good framer. It doesn't even attempt to answer that question. So, yes, providing an unbiased evaluation is the point of having a metric, and we shouldn't necessarily disagree with what a metric says because it doesn't agree with what our eyes say. But I'm not disagreeing, since FRAA doesn't currently evaluate his framing. We plan for it to do that in the pretty near future. But for now, I'm relying on other metrics/unbiased evaluations: the linked research by Bojan Koprivica on pitch-blocking, and the linked research by Mike Fast on framing (which would be the basis of any framing-related change we make to FRAA in the future).
Regarding the first paragraph, when BP staff evaluate players, they use FRAA to add or subtract runs (and similarly, wins) to a player's value. In the comments of this article, you used FRAA to support the idea that Williams is an average fielder while ignore it when evaluating Posada. From my aspect, I don't think that is fair. If you choose to FRAA to evaluate Williams, why can't I use Posada's FRAA as a counter argument?
You say that in some instances, you use FRAA and in others, you don't. You also are saying you have some outside calculation from Colin showing that Posada is in fact a horrible fielder. How am I, an outsider, supposed to know when FRAA is "good" and when FRAA is "not good"?
It seems that FRAA, even after its recent published revision, is still unreliable. Again, I ask why use a metric at all if it is completely unreliable? Just because it's some BP proprietary thing?
Again, I don't think FRAA gives as accurate an assessment of Posada's fielding as it does of Williams'. And that's not because I'm applying the statistic selectively to support an argument, or because I watched both players and decided that FRAA was wrong in one case and not the other based on my own observations. I don't have a horse in this race, so I'm not trying to make Bernie look good or Posada look bad.
I'm also not saying that in "some" instances I use FRAA and in "others" I don't. I'm saying that I don't use it--or more accurately, I don't assign as much weight to it--for catchers. FRAA primarily measures performance at fielding batted balls. That's a really, really big part of what center fielders do. It's a really small part of what catchers do. So, when I'm evaluating the relative defensive performances of center fielders and catchers, I value FRAA's input accordingly.
I also never said I had an outside calculation from Colin that showed Posada is a horrible fielder. Colin's calculations show that Posada was about a win below-average at controlling the running game. He was about a win above-average at fielding batted balls, as the value on the website says, so those things cancel out. If that were all we knew about Posada, then you could say that we have no statistical evidence that he was a below-average defender. But that's not all we know--the recent advances in evaluating blocking and framing have revealed his struggles in those areas, and it would be silly to ignore that just because FRAA doesn't factor in those aspects of his fielding yet.
So, I wouldn't say that FRAA in "unreliable" in large samples, except to the extent that any fielding metric is unreliable (their limitations have been discussed many times, both at BP and elsewhere). It's just incomplete for catchers, for reasons that I explained in the article. I didn't cite it just because it's a BP stat. I cited it because it tells us something. it just doesn't tell us everything.
Catcher defense via FRAA, in this current form, is far more incomplete than it is for an outfielder. Ergo, using it as a crutch for Posada's case is probably not a good idea, especially given any scout, front office member, or analyst who has seen Posada defensively would choke and die on their own laughter if you suggested he was above-average defensively over his career.
As for using a "BP proprietary thing" when it's "unreliable"... dude, c'mon. FRAA isn't finished at catcher, just like it isn't finished anywhere else on the internet. Catcher defense is even further behind than at other positions. And Colin Wyers has done plenty of explaining while you've had a subscription to read it as to why the back-to-basics FRAA is a good idea in this age of uncertainty regarding defensive stats. Sam Miller wrote just last year about the uncertainty of catcher defense specifically. It's not like this is a topic that's been ignored here.
Lifelong Yanks fan here. Never thought Bernie was a great fielder. He rarely got a good jump, took strange routes, rag-arm, but had good hands when he got near a ball, so had some nice highlight catches and his speed saved some of those bad jumps at times. Never thought Jeter was a good fielder. Thought Posada was awful.
I too am a big Yankees fan; and have been for 20+ years. I agree with tommybones' assessment, and am willing to say that many of the second basemen were pretty bad, too.
But damn, they won a whole lot of games with that miserable up-the-middle defense.
Funny, isn't it? The Yankees fielded a dynasty, and one of the best teams of all time, with a shortstop, catcher, and (depending on your source) center fielder who were among the worst fielders at those positions in their era, if not in the history of the game. Of course, those players were also among the best hitters at those positions in their era, if not in the history of the game. So there's that.
Defense might win championships, but you know what else wins championships? Offense. Runs are runs.
The idea that defense wins champions was based on older calculations of FRAA and Defensive Efficiency though. I distinctly remember a few team essays based on this idea. Do the Rays and Rockies teams of yesteryear still hold up under recalculation?