Archive for the 'Stats' Category

In Defense of Outliers

Occasionally, baseball players lose ownership of their own names.  Steve Blass, Mario Mendoza and Tommy John have become adjectives, terminology rather than personality, their careers condensed into a single trait.  Such is also the fate of Brady Anderson, who played fifteen seasons in the major league and yet in a very real sense played only one.  In that infamous year of 1996, the reedy Anderson hit fifty home runs, nearly a quarter of his career total.  It’s an accomplishment that only twenty-two players in baseball history can claim, and yet it’s invariably followed by an invisible asterisk.  It’s not that the home runs didn’t happen; it’s that they shouldn’t have.

The value embedded in the phrase “Brady Anderson”, naturally, is its connection to the steroid era.  It’s one of those cumbersome tasks that every discussion like this has to start with, even though author and reader alike already understand the implications.  Amazing feats of baseball abounded in the era directly following Anderson: names like Luis Gonzalez, Ken Caminiti and Bret Boone flung themselves onto the headlines, while Sosa and McGwire smeared their fingerprints ontorecord books, distending the numbers.  The resulting chaos has left fans weary and confused, unable and unwilling to sort through the ashes.  Anderson has firmly denied any steroid use, but such denials are useless; it isn’t Brady Anderson that has become attached to juicing, but greatness itself.

Several months ago, Joe Posnanski of Sports Illustrated wrote an article about Jose Bautista.  Bautista’s career began even more ignominiously than Anderson’s, and has since soared even higher.  And much like Anderson, Bautista has faced a significant amount of scrutiny for his achievements.  Posnanski begins with the simple question: “Do you believe in miracles?”  He then conjures the familiar names of the great and unlikely, Lance Armstrong and Kurt Warner and Dazzy Vance.  We’ve grown skeptical, as a nation and as a sport.

It’s the ultimate condemnation of Anderson and Bautista: what they did doesn’t make sense.  But was Brady Anderson’s 1996 a miracle?  Is Jose Bautista’s ascension?  Voltaire wrote on the subject of miracles in his Dictionnaire Philosophique, defining a miracle as “the violation of those divine and eternal laws.  If there is an eclipse of the sun at full moon, or if a dead man walks two leagues carrying his head in his arms, we call that a miracle.”  This is the ultimate condemnation of Brady Anderson and Jose Bautista: what they did doesn’t make sense.  A man doesn’t go from hitting fifteen home runs to fifty.  He doesn’t go from being cut by losing teams to being an MVP candidate.  These things aren’t independently possible, and so there must be something else causing them, something unnatural.

But though Voltaire’s eclipse and his headless man were both considered miracles at one time, they’re very different.  One violates the natural laws as we know them.  The other violated the natural laws as we knew them at the time, but later came to be understandable.  As we grow more knowledgeable about baseball, and we become increasingly skilled at analysis and projection, we become increasingly resistant to aberration.  The flaw in so much of analysis (baseball and otherwise) is that while we smirk at the ignorance of the past, we neglect to factor the ignorance of the present.  We do not know what we will know, and what fails to make sense now may be perfectly clear tomorrow.

In this sense, miracles are dangerous, revolutionary things.  They challenge the solidity of accepted wisdom.  They force us to question our assumptions about the world.  They challenge the laziness of our thinking.  Steroids have become one example of this laziness: a refusal to examine greatness, to admit the possibility of being impressed. Occam’s razor has gone from being a guideline to a law.

Perhaps most importantly, miracles chip away at our fundamental preference for certainty.  Luck is something we understand, at least when it turns against us.  We want to believe that our successes, however, are the result of nothing except our own pluck and determination.  Anderson seems to agree.  He described 1996 as “just one more home run per week, just one more good swing. That is the data that simultaneously comforted me and haunted me, the small difference between greatness and mediocrity.” The role that luck plays in the success of a baseball player is only an exaggeration of what goes on in our own lives.   How many people who have condemned Anderson’s achievement as impossible have gone home to play the lottery?

Ultimately, I’m not in a position to say whether Brady Anderson used steroids or not.  The possibility exists, as do other possibilities.  What interests me is the potential for greatness, the acceptance of outliers.  Every game, every season, something happens in baseball that defies expectations, and demands that we dig deeper.  Call them miracles, call them flukes, call them statistical deviations.  Regardless of what they are, they bring vitality to the sport, and in some cases, they form the origins to amazing narratives.  It’s a possibility I find infinitely more palatable than the predictable alternative, no matter how much sense it might make.

Good News for Eric Wedge

Perhaps the most interesting aspect of the Mariners’ recent fall from grace is the lack of acrimony inspired by it.  There are plenty of stories in the national media breaking down the quantitative futility; everyone, after all, loves an outlier.  The local fan base is mourning the loss of Eric Wedge’s mustache almost as much as the team’s season.  Wedge, although capable of throwing out his share of baffling lineups, is generally respected as a manager.  Jack Zduriencik, unlike his predecessor, has made the kind of mistakes that at least follow some line of logic.  Expectations were reasonably tempered.  Even on the fifth of July, when the team was .500 and two and a half games out of first, everyone secretly knew that this was a roster capable of dropping a dozen games in a row.

Of course, as of July 26, 2011, the Mariners have outdone themselves, accomplishing a feat only twenty teams have done since the American and National Leagues merged in 1903.  And with a truly historical run of failure, Wedge and Zduriencik have been put on the hot seat almost by default.  But as it turns out, losing fifteen or twenty games in a row isn’t the death knell for a career one might think.  The list:


Eric Wedge, as it turns out, has joined some pretty respectable company in the past two and a half weeks.  This isn’t as surprising as it seems; if you stick around the game for thirty or forty years, you’re bound to see some streaks, good and bad.  Still, several of these managers (Herzog, Kuhel, and Mauch) were first-year managers, and were given at least another year to prove themselves.

Many of the teams who fired coaches after losing streaks did so under extenuating circumstances.  Tenney and Collins plied their trade during the player-manager era of baseball; Tenney was traded after his 1907 season, and released at the age of forty after 1911.  Collins, the Hall of Fame third baseman, was stripped of his managerial duties mid-season, a full eighty games after the end of the twenty-game losing streak.

Ted Turner gave Dave Bristol a ten-day leave of absence in 1977 so that he could manage the team himself, until N.L. President Chub Feeney stepped in and slapped the rulebook in his face.  Turner somehow persuaded Bristol to come back as a lame duck.  The world remembers the 1988 Baltimore Orioles for its staggering 0-21 start to the season, but Ripken, Sr. was actually fired after only six games.  Replacement-level manager Frank Robinson lost the other fifteen.

Of the nineteen managers, three of them were fired after and because of their losing streak (Collins, Fohl and Bristol).  Four were enshrined in the Hall of Fame.

Where does this leave Eric Wedge?  Probably in neither category.  Miller Huggins once said that “a manager has his cards dealt to him, and he must play them.”  Nobody envies Wedge’s hand.  He can’t be accused of losing the players, and he’s shown a willingness to be flexible with his roster without making constant, desperate changes.  But for lack of a better alternative, we continue to measure managers by wins and championships.  Gene Mauch might prove a solid comparison: a pretty good manager who led some pretty awful teams.

The Seattle Mariners are a fascinating ballclub right now; rarely has a team lost so much and had so little meaning attached to it.  Usually, this kind of unabated failure beats down even the sensible fan, wears them raw until they need something, anything to be done.  They attach responsibility to whatever they can reach, and usually the field leader is the first in line.

In the case of the Mariners, however, there are no mutterings about intangibles, no hidden knowledge of winning.   They’ve lost sixteen times to teams that are better than they are.  Ordinarily, inferior baseball teams win their share of games against superior opponents; right now it isn’t happening.  It feels like an inevitability, but one of probability rather than fate.  Sooner or later a team is going to lose fifteen or twenty games; why not now?

 

Celebrate Different

Last night I was talking with Eric Freeman about No Hitters. By talking I mean agreeing that it’s pointless and purposefully joyless to whine during and immediately after a no-hitter that the game isn’t as “well-pitched” as other games that aren’t no-hitters. There is inherent value in the sheer improbability of a game like the one Francisco Liriano pitched yesterday. It was a messy, unbecoming, a nerve-wracking, defense-driven spectacle of the highest order.

No-hitters are news because they are rare and dramatic, not merely because they are impressive. As far as I know, there isn’t a contingent of people out there screaming and tweeting that Francisco Liriano pitched the best game of the season last night. There isn’t a contingent yelling about how he’s the league’s best pitcher. That’s besides the point. The point is that there is joy in the string of lucky plays and defense and building tension that defined his performance.

I went to the Mariner game last night. Erik Bedard tossed five of the ugliest no-hit innings in baseball history to open the game. He even allowed a run. Anybody in the park could have told you that it wasn’t a great performance. They would have also told you that they were disappointed — and yet not surprised — when Ian Kinsler broke it up with a double. It has to do with the streak. It has to do with the fact that the baseball fan experience is defined by narrative as much as it is by statistical understanding.

That might be the whole point of this blog. I don’t think we’re working against the grain here. I don’t think we’re anti-sabermetrics. We love and embrace them. They are useful. They make us smarter. But from my angle, a central element of that world is this ill-defined quest to seek out the ‘objective best’ of everything. I think that comes at the expense of the ‘subjective best’ — the most interesting, the most dynamic, the most grok.

The Francisco Liriano no-hitter was an anomaly. Jeff Sullivan from Lookout Landing tweeted that “Since the beginning of the 2010 season, 86 starts have a higher game score than Liriano’s yesterday.” If anything that makes the performance all the more compelling. Sullivan later tweeted that the start was “impressive in a different way than usual.” Damn right it was different. Let’s celebrate different. Francisco Liriano is a pitcher on the verge of total collapse. He — with the help of his team and yes, precarious luck — held it together for something magnificent. Let’s sing about it.

On Sabermetric Transparency

Over at Walkoff Walk, 310toJoba (somebody get this guy a first name), writes about the mega-awesome-super news that Bill Simmons, the internet voice of the Sports Media Industrial Complex has officially embraced sabermetrics. This is a major (if inevitable coup) for the stats-y baseball blogosphere. If no longer the Voice of the American Sports Fan, Simmons remains influential. He is also useful as a bell weather.  As Simmons goes, so goes the sports fan.

Anyway, 310toJoba asks many great questions of the article, and hits Simmons for his navel-gazing and the back-handedness of his compliments. It seems futile to point out that a Simmons column without navel-gazing has yet to be written. And as to the back-handedness, I didn’t really read the article as pejorative. But perhaps that’s because I’m not a numbers guy myself and this is not a numbers blog.

But once again, that’s not what I’m here to write about. 310toJoba says the following of The Sports Guy’s desire to understand what goes into making these statistics:

On the one hand, I appreciate his efforts to attain a better grasp on the stats as a whole; he consistently tries to find out how they’re calculated. Good on him. On the other hand, perhaps Simmons is getting a little too overzealous and missing the point.

And later:

Again, it’s admirable that he wants to go all the way with his newfound obsession, but he comes off as being condescending and too in depth when there’s no need to be.

310toJoba then honorably admits that he has no idea how many of these stats are calculated and questions whether actually understanding the formula would make him a better or better-informed baseball fan. All this amounts to the typical argument “there are smarter, better suited people to do this, I’ll just trust them.” (Not an actual quote).

And here is where I find myself disagreeing with Mr. 310. I think Simmons’ desire to understand the formulas is entirely reasonable. And I don’t see how it is in any way condescending. Here he is admitting to the great wide world that sabermetrics are better than traditional numbers at measuring baseball performance. That’s still a pretty big deal, and for people to embrace that notion, they have to understand why these numbers are better.

There is a tendency among people at the forefront of change and new ideas to assume that the masses will somehow intuit why their proposed changes and new ideas are better. This assumption is why Americans were so vehemently opposed to Health Care Reform – they just saw it as an amorphous blob set forth by people unwilling to explain it in a palatable manner. So when guys like Joe Morgan (or Lindsey Graham), say that these ideas are wrong, or un-American, or will have horrible consequences, the urge is to recoil from them. The remedy to all this is spelling out exactly what these new ideas amount to, and doing so in simple and tangible terms. Just saying “trust us, it will be better,” is not enough.

The baseball stats we grew up with are very easy to calculate. If they aren’t counting stats like runs or runs batted in, they are equations with few inputs requiring basic arithmetic. Walks Plus Hits Divided By Innings Pitched. Okay Simple. We trust those stats because we have a good grasp on what exactly they are telling us. And we know that although not perfect, they are not necessarily bad. Baseball was just fine without sabermetrics. So who are you to tell me that this newfangled stuff can make it better?

I’m not sure it’s enough to just have some smart person tell you “OPS+ is a great metric for offensive performance!” and just believe them on blind faith. I’ve grappled with this myself. I am a pretty sabermetrically literate guy. But I hate relying on statistics I do not fully understand. It often feels like I am arguing on a foundation of quicksand; like somebody could open the curtain and reveal that Bill James is as phony as the Wizard of Oz, and because I don’t fully understand how to calculate UZR, I too will be revealed as a phony.

Obviously, I know this is not the case. I know that smart and well-intentioned people are doing this research to help our understanding of the game. But I know this because I write a baseball blog, and because I’m a curious guy who has tried to learn the formulas. I am not inclined to take it on blind faith that new stats are better stats, and neither are most other baseball fans. It might take, as Simmons says, only ten minutes to be a better informed fan. But it takes more than ten minutes to figure out how VORP is calculated. Does being able to rattle off advanced stats really make one a better informed fan? Or is there some obligation to learn how the gears grind beneath the sheen of the number itself?