Monday, January 27, 2014

Statistics ARE the game

"I’m not a huge stat guy. I’ve always been a believer that if you took stats out of the game, you’d still know who the good players are, just by watching."
From Fred Heumann, the Lansing-area CBS sports anchor:
*

In a sense, a rebellion against stats is a rebellion against intellectual, analytical culture.  We see this in politics all the time, where nuanced opinions and terrific intellect and higher education are seen as negative qualities.  It's far better to be "relatable."  People who rail against statistics do not understand why, after decades of having perfectly good stats like batting average and RBIs, we need to suddenly rethink the way we evaluate players.

But baseball statistics are not Moneyball.  They are not created by nerds, sitting a far ways from any diamond, typing madly upon a computer.  They are not WAR, they are not OPS, they are not PECOTA, and they are not Bert Blyleven's Hall of Fame induction.

Baseball statistics are the game.  Every pitch is a statistic.  Every hit, walk, and out is a statistic.

Mr. Glavine, we cannot take the stats out of the game.  If you pitch seven innings and give up two runs, those are stats.  But maybe you're referring to the circumstance in which a player's performance was not reflected in his box score numbers.

* Let's say hypothetically that you pitched wonderfully, but your defense was poor behind you.  Ah ha:  the introduction of earned runs and unearned runs.  We also can tell which teams' defenses are the worst at turning batted balls into outs.

* What if you pitched poorly, but were bailed out by run support?  We have statistics to indicate this, too.

* What if you pitched wonderfully, but broken-bat base hits and bloopers did you in?  That's where a limited sample size is taken into consideration. The more innings you pitch, the more likely your performance will accurately reflect your true ability.

Really, the more you pitch (or hit), the more your stats will bear out your measure.  No batter will consistently hit hard lineouts his whole career.  One season might be lucky or unlucky, another season might look oddly coincidental, but everything evens out over time.

A hitter who gives up his at-bat in order to move up a teammate does get an 0-for-1.  He also will have plenty more at-bats to balance matters out.  Over the course of a season and a career, if that player is worth his salt, he'll be doing far more than providing situational groundouts in order to help his team win.

This is why the idea of the "Eye Test" is ridiculous.  It's an admission of "I don't care about anything other than what I see."  It's the equivalent of a 6-year-old going to a game, seeing Joe Shlabotnik homer, and deciding Shlabotnik was the best player ever.  For a 6-year-old, that's fine.  For you, a limited scope of observation shouldn't be enough to come to a fair assessment.  Anyone can have a great day.  Pat Seerey hit four homers in a game once.  Jose Jimenez tossed a no-hitter.  Armando Galarraga should have been perfect.  (Galarraga went 4-9 with a 4.49 ERA and a 93 ERA+ in 2010, the year of his should've-been-perfection.)

The charismatic Steve Garvey may have been a Hall of Famer to Mr. Heumann -- and in fact, he was a fine player, a perennial All-Star, a solid defensive first baseman, an MVP, an iron man, a postseason standout, and a similar offensive player to Garret Anderson.

But do you know how you make a case for him?  Numbers.

Okay, Steve Garvey didn't hit 500 home runs or collect 3,000 hits... so what did he do?  How did he help?  Was it defensive?  We have stats for that.  Was it in "clutch" ways?  We can research his situational performance.  The iron man streak?  Heck, this might sound weird, but we keep track of his games played.

(Me, I wouldn't induct him.  There've been a number of these sorts of first basemen in recent baseball history:  John Olerud, Keith Hernandez, Don Mattingly, Mark Grace, and Will Clark.  All helped their teams win, all could hit, all could field, but probably none of them is worthy of Cooperstown.)

Really, anyone who says that they'd rather not go by the numbers is either 1) too lazy to look up what the numbers show, 2) displeased with what the numbers show, 3) confused by how to use the numbers properly, or 4) overwhelmed.

That last one I understand well.  Check out the Baseball Prospectus glossary; how well do you understand the metrics that produced BRR, FIP, FRAA, PADE, PECOTA, TAv, and VORP?  For that matter, how well do you know what those acronyms stand for?  I sure as heck don't.  If ERA and batting average is the 101 level of baseball statistics, those are the 301 and 401 level courses.  But we shouldn't run from them.  Instead, we should choose one, sit down with it, and see what we think.

Here:  BRR are Base Running Runs, and they indicate just how well a player runs the bases.  That's a pretty darn valuable stat.  Is there a player claiming that we can't measure his base running acumen in the box score?  Boom.  Now we can.

*

Returning to Tom Glavine:  It's true, more often than not, that we could figure out who the best players in a game are.  We can see their speed, their power, and their effectiveness.  But, barring an invitation to the MLB Fan Cave, we can't watch every game.  Should that stop us from properly evaluating players whom we don't see?  Why, that might lead to some sort of coastal bias...

No comments: