When Models Look Extra Ugly

Democrats are predictably excoriating Paul Ryan these days for his "attack" on Medicare. His voucher plan for the government-sponsored health care insurance for the elderly (and the not-nearly elderly; you can join at 65) diminishes benefits going forward.

Democrats are comparing Ryan's plan to perfection, or to the current system. Neither, of course, is an apt comparison. Medicare is rushing headlong off the financial cliff without reform, and threatening to take the entire American economy with it. Relative to insolvency, which is the ultimate fate of the program under everyone else's plan (i.e., no plan at all), Ryan's proposal is a revelation.

A similar debate is shaking the Sabermetric community these days in the wake of a pair of disturbing arithmetic developments. But the conclusion is the same: the measure of a new idea isn't in comparison to perfection, but to the old paradigm.

First, some context. Ever since Voros McCracken unveiled the idea in 1999 that pitchers have limited control over what happens once their pitch is put into the field of play -- a conclusion now much more nuanced and qualified -- seamheads have been whipping up concoctions to strip luck and defense out of pitching results to determine how well pitchers are actually pitching.

One of the many formulae brewed by the stat wizards is FIP -- fielding independent pitching. Without excavating spreadsheets filled with higher math, it can be described this way: given the number of batters a pitcher has fanned, walked, hit with pitches and allowed to homer, how many runs should we expect to score against him given average defense and ballpark?

In July, Reds closer Aroldis Chapman broke FIP (in the words of CBS Sports). He faced 52 batters in 14 1/3 innings and whiffed 31 of them. He allowed just six hits, two walks, a HBP and nary a run, nailing down 13 saves. Batters posted a desultory .122/.173/.143 line against him.  In other words, Chapman was Superman for a 30-day period. (He's been no slouch under other moons.)

Putting those numbers through the meat grinder yielded a FIP for Chapman of -0.99. That is, Chapman could be expected, given his performance, to yield minus-one run per nine innings. Certifiably nuts.

Others have noted odd sightings in Baseball-Reference.com's WAR (Wins Against Replacement) calculations. WAR attempts to review all of a player's performance and all of the context and measure him against a Triple-A replacement at his position. In mid-August, Cubs keystoner Darwin Barney had a higher WAR than Brewers slugger Ryan Braun. Just for context, Barney is hitting .268/.309/.386, or about 12% below the MLB average. Braun, at .301/.380/.526, is 53% above average. Braun is also superior in the baserunning department and has grounded into fewer double plays. Moreover, Wrigley Field and Miller Park are about equally kind to hitters, and both batters suffer equally in their inability to bat against their own team's woeful pitching staff.

What WAR sees that we don't is defense. It credits Barney for defense double in value of any other second-sacker's, while Braun's left-field stylings more resemble the staccato flight of a pigeon. Subjectively, there appears to be a grain of truth here, but clearly not enough that anyone in Wisconsin would trade Braun for every Barney ever known, including the purple dinosaur. WAR is suffering convulsions and has been put on bed rest, at the very least.

These two developments have led some to observe FIP and WAR's death throes. The projection models have failed and it's time to put them out of their misery. But in the words of that great philosopher Quick Draw McGraw, "Now hold on there just a doggone minute, Baba Looie."

No model is perfection, not even Adriana Lima. Clearly outlier performances like Chapman's make mincemeat of statistical models. As I've mentioned before, quantitative analysis of defense is just in the bloom of its youth, beholden to bursts of impaired judgment. I rarely rely on WAR or WARP and pay much more attention to offensive valuations, leavened by a general sense of a player's glovework.

Nonetheless, FIP is an extremely useful tool, and though the value it applies to Chapman's performance is nonsense, it's not really wrong. FIP says Chapman was virtually unhittable, and by golly, he was. Only 18 batters out of 52 could even put the ball in play.

The real question then is, are these models less imperfect than the old tools? WAR certainly measures something bigger and more relevant than Triple Crown stats do, but whether it measures relative positional value better than Triple Crown stats measure hitting performance would require some sort of study. There's no doubt that FIP tells us much more about a pitcher than W-L and ERA; it's been shown to be vastly better at projecting performance.

All of which presents the same moral that every other development in quantitative analysis has: the state of the art is improving, but will never be perfect. After all, they measure the performances of people.

The Case of the Fragile Pitcher

Welcome to this week's edition of Mystery Theatre where we delve into the most vexing whodunits of our time.

This week, we examine The Case of the Fragile Pitcher and analyze the theories that abound, ex-post facto, to explain why hurlers in the 40s, 50s, 60s, 70s and 80s took the ball every fourth day, remained on the hill for 300+ innings and wore down no more than their modern counterparts despite advances in conditioning and medical care. 

Our sleuths will uncover why today's moundsmen are bigger, stronger, better conditioned, yet throw no harder and dis-assemble faster than their progenitors despite massive attention paid to their workload.

Actually, our sleuths will do no such thing. The point of this installment of Mystery Theatre is that this week's mystery has earned its stripes and avoided explanation like Jimmy Hoffa's killers.

The timing of this discussion coincides with the implosion of Stephen Strasburg's elbow, which would have ended his career had not Tommy John presented the same ulnar collateral ligament injury to Dr. Frank Jobe in 1974.

How, we're all asking, could Strasburg have contracted this quintessential pitching woe when management of the Washington Nationals treated him like a sickly Southern belle in a heat wave? Despite gestating his talent beyond necessity in the minors, truncating his starts, plying him with extra rest and reacting to his every boo-boo with ever more caution, the franchise is nonetheless rewarded with the very injury they most feared.

Listen to sports talk radio and television and you will hear ex-jocks propose theories big and small for this vexing phenomenon. Their explanations range from the unprovably plausible to the patently ridiculous. Mostly they are stand-ins for the far more accurate and credible answer: I don't know. My favorite is the former players who rail against pitch counts. Although limits on pitch counts got a little out of hand for awhile, they are still based on solid research supporting certain limits on some pitchers.

We know that today's pitchers are bigger, stronger, better-conditioned and better-trained than their predecessors. We also know that the highest heat today is no hotter than the highest heat of yesteryear. Nolan Ryan, Bob Feller, Sam McDowell; hell, Walter Johnson would all be considered fireballers today. And we know that there seems to be some risk in increasing the workload of young pitchers dramatically from one year to the next, though the Verducci Effect is no longer the universally accepted metaphysical truth it once was. None of that explains why Cy Young's great-grandchildren can't hold a candle to Cy Young's grandchildren.

I would like to take a stab at explaining part of the problem by noting three simultaneous truths. Before I do, a caveat: this is not a universal theory of pitching durability. I don't really know why throwers of Bowie Kuhn's day outlasted today's crop.

I wonder whether the average pitcher today doesn't throw harder than the average pitcher of 1970, even if the pinnacle of the profession hasn't changed. Someone probably knows the answer to this, but I don't. A major league fastball is at least 88 mph in 2010. Rare is the violator of this rule. Was that the case in 1975, or could you get away with a speedball that peaked at 82 with sufficient craft and a 12-6 curveball?

I also wonder what impact new pitches like sliders, cut fastballs, etc. have  on athletes' shoulders and elbows. Are they even new pitches at all, or just new names for strategies being employed all the time?

Onto the facts: one reason pitchers pitch fewer innings today is that they pitch fewer innings. No, really. Isn't it conceivable that if pitchers took the ball every fourth day while working their way through the minor leagues and then continued in that vein in The Show, they would get acclimated to that workload? Under those conditions, some major league pitchers would compile 250+ innings, just like Tom Seaver and Juan Marichal did.

Fact Two: Bullpens are used differently today than in the past. Before the one-inning closer and the LOOGY (left-handed, one-out guy), managers let their starters pitch until they ran out of steam or had to be replaced by a pinch hitter. Teams had 10-man staffs -- four starters, a closer and a bunch of inning-eaters. Today, perfectly fresh starters get yanked in order to maximize match-ups. We have closers and set-up men and long relievers and situational lefties and righties who together comprise 13-man rotations. No wonder a starter can't get an inning in edgewise.

Fact Three: Pitching today is a full-time job. Pitchers face hitters up and down the lineup who are bigger, stronger and more powerful than in the past. Throw in the DH and today's hurler must make every pitch count and use every pitch to set up every other. Ken Griffey, Sr., Tony Gwynn, Sr. Cecil Fielder, Bob Boone, and others whose loins produced today's breed, competed with middle infielders who were Michael J. Fox-sized and couldn't go yard off a tee. Juan Marichal could cruise through the bottom of the order; a hanging curve might be a hit, but rarely extra bases. He didn't have to futz around with setting up Roy McMillan and his 292 SLG.

In short,  the full story behind the discrepancy between pitchers of bygone eras and today remains a mystery. There seems to be evidence that pitching today is more difficult and that usage patterns have changed. Beyond that: I don't know.

b

Pitcher, Deconstructed

In the first wave of sabermetrics, back in the days when pencil-necks with bad skin and no friends wiled away their solitary existences knee deep in notebooks, calculators and newspaper sports sections, the emphasis was on deconstructing the statistics we were familiar with -- e.g., RBIs, batting average and pitcher wins -- and demonstrating their unworthiness.

In the next wave, a growing cadre of socially-adjusting statheads examined context, recognizing that a home run hit in Dodger Stadium, or in 1968, or in a tie game, is worth more than a home run hit in Denver, or in 2002, or in a six-run game.

The third sabermetric epoch, which persists, is focused on the deconstruction of each individual play, which is now recorded for posterity in atomized detail for everyone to see instantly and simultaneously, in order to assess credit and blame for every result. How much responsibility do the batter, the pitcher, the shortstop, the centerfielder and dumb luck have for a given play. Clearly, the two fielders bear no responsibility for a home run, walk, strikeout or hit by pitch. But even those bear scrutiny between the hurler and hitter.

For years, we assigned blame to pitchers for walks and offered little, if any, kudos to the batter, who after all, simply lallygagged at the plate with the bat on his shoulders. We have since discovered that batters have very different abilities in that area and deserve vastly more credit for "earning" a free pass. (We've also come to realize that walks are enormously important.)

What about other kinds of plays that involve base runners and multiple fielders? Except for recording errors, the box score gives little notice to the defensive unit, and makes no effort at all to acknowledge luck. But recent research has demonstrated that defense and luck are big parts of the game. In fact, this generation of statheads -- Ivy League literature majors who run marathons and gather around the nation by the hundreds -- has found methods of analyzing every kind of batted ball (e.g., fly balls, line drives, grounders, pop-ups, etc.) and applying 100 years of past performance to determine how many of them should, on average and adjusted for ballpark and era, turn into hits. They then compare defensive units to the average and compute their defensive efficiency. 

Here's where it gets fun. Consider a ground ball hit with medium velocity between first and second. If the second-sacker is a slacker, the ball trickles into right field and the batter gets a hit that the pitcher surrenders. Suppose instead that the keystoner reads the swing and breaks to the ball, snares it, and flips to first for the out. Lousy at-bat, good pitching, even though the interaction was exactly the same.

Over the course of a game, and certainly a season, some infields will turn more of those grounders into outs than other will. Some outfields will run down more apparent gappers than others will. And some pitchers will be the beneficiaries while others will watch their ERAs balloon. Hence a recent development: defense-independent pitching statistics. Sabermetricians have attempted to glean the pitchers' responsibility from things like defense.

We still haven't accounted for luck. If the ground ball is six feet to the left, any second baseman makes the play. Three feet to the right, none does. Batters generally can't (absent the occasional Ichiro) so finely deliver their hits and pitchers are simply passive actors in those sorts of plays.

The recently discovered truth is that pitchers have a limited amount of control over what happens to balls put into play (that does not include fly balls that clear the fences), which is another way of saying that defense and luck are vastly under-appreciated (by the fans, but no longer by front offices, which have teams of SABR members crunching these sorts of numbers into nubby granules.) They've also discovered that since pitchers control home runs, walks and strikeouts (and HBP, for what little it's worth), that measuring those occurrences only is often more descriptive of a pitcher's performance. The math is stultifying, particularly if you're flummoxed by the profound complexities of single-digit addition, but suffice to say that if you want to predict which starter will pitch well this year, find a guy with a high strikeout-walk ratio and few home runs allowed last year, and ignore his W-L record and even his ERA.

Let's take the example of Tim Hudson's five and two-third inning outing against San Diego for the victorious Braves last night. The veteran righty recorded the win, surrendering just two runs, for an ERA of a 3.18. Bobby Cox is happy, no? Well, Hudson walked five batters, fanned no one and served up a dinger. He also got relief help from Kris Medlen when he ran into trouble in the sixth inning, and more bullpen support over the last three innings to preserve the win. He certainly wasn't fooling anyone, and Padre batters managed to hit safely at just a .261 rate on balls in play, much lower than the league average of .300 or so.

It's likely that Hudson benefited from some combination of strong defense behind him and good luck. He allowed 11 baserunners, including a double and home run, and probably should have relinquished another safety or two. The five free passes and the inability to strike anyone out has got to agitate Cox's digestive juices; he certainly can't be counting on a season of .261 BABIP against Hudson.

So? We still haven't completely divorced pitching results from all the extraneous factors at work in a game or a season, and much of the defensive measuring is still guesswork's second cousin. But scientific gains are made slowly and incrementally, and the information we have now about pitcher value is more illuminating than wins and ERA ever were.

b