Adieu Ancient Statistics Pt. 2
Time for the second installment of this series, which, in my more sinister moments, I have envisioned as the death rattle of unrefined statistics. If you haven’t yet read it, please catch up with Part One of the series.
I spent the last post detailing what old guard statistics cannot do. But it is important, in discussing the evolution of baseball statistics, to mention what they can do. The Error and the Assist recorded a thing that happened. Truly, they weren’t meant to do anything more. The problem is that they have been used for more.
Let’s put it this way. You don’t dance with a mop. A mop is not a woman. But the men that designed the mop didn’t fail because a mop is not a woman, a mop still has a simple elegance and it accomplishes its simple goal, just as the Error and the Assist accomplished theirs.
What advanced statistics want to give us is not a mop, but a sexy lady. That’s not very clear… Advanced statistics are meant to determine value. To determine value, they must be comprehensive. It should be noted that in many cases one stat alone cannot be comprehensive, though some (like wins above replacement, more commonly referred to as WAR) attempt to.
Sexy Ladies to the Rescue
Let’s return to the questions one should ask to determine the value of a fielder.
- What balls can he get to? (Range)
- Does he make the plays he can get to? (Fielding Ability)
- What throws can he make? (Arm Strength)
- Can he throw it where he’s supposed to? (Arm Accuracy)
- Does he know what to do when he has the ball (Baseball IQ)
One of the most terrific (and terrifically complicated) statistics devised to handle defensive value is Ultimate Zone Rating (UZR). One of it’s components is an extremely intimidating mathematical operation that is not interesting. What is interesting is that UZR takes into account what type of play a defender had to make based on how hard a ball was hit, where it was hit to, and how far the fielder had to go to make the play, as well as how far he had to throw the ball to complete it. Toss all that in the boople bopple machine and the math component calculates whether this was a play that an average defender would have been able to make. Then, over time, a player’s UZR goes up as he makes plays that an average defender can’t, and it goes down as he fails to.
So, where the Error offers only the raw number of instances in which a fielder failed, UZR tells us whether a guy can make play after play that most guys can’t, or whether he is Wilson Betemit. It assesses range, arm strength, fielding ability and arm accuracy all at once. There are stats that are intended to assess each of these skills independently as well. Many are variations of Runs Saved. We can calculate, based on the same factors UZR takes into account, how many runs a fielder saved his team with his arm or with his range. Thanks Sexy Ladies.
The Ethier Situation
Though he is a total Adonis, Andre Ethier did not play Gold Glove defense in 2011. The sole challenge to my argument is that Andre Ethier has a big gold trophy of a glove in his living room. It should be pointed out that by no means was Ethier’s award exceptionally unjustified, but that his story is particularly compelling for reasons pertaining to the Error.
Andre Ethier did not make a single error last season. So the guys who didn’t watch him all year, that somehow get to vote for Gold Gloves, saw a zero on a spreadsheet, and mindlessly obeyed it.
But let’s see how Ethier stacked up against other rightfielders in 2011 using Sexy Ladies:
- Andre Ethier saved 1.7 runs by not making errors, which was tops in the NL and second only to Nick Markakis (1.8).
- He cost his team .2 runs with his arm, landing him 20th overall, 8th in the NL. Jeff Francoeur led in this category saving 9.3 runs, while Jayson Werth led the NL with 5.4 runs saved.
- Ethier saved only 3.8 runs with his range, placing 12th overall and 7th among National Leaguers. Fellow National Leaguer Justin Upton saved 15.4 runs with his range to lead all of baseball.
- His UZR per 150 games (which levels the playing field by illustrating what each player would do given equal opportunity) was 6.8. 7th overall, 4th NL. Rookie Jason Heyward topped the NL rankings here at 12.7, while Josh Reddick paced the Majors at 18.3.
Jason Heyward was probably the National League’s most well rounded defender in right last season, placing first in UZR/150 as mentioned, and also placing 2nd in the NL by saving 10.6 runs with his range, as measured by RngR (range runs).
What we learn from Ethier’s place between old statistics and new is that making all the plays you get to does not prevent as many runs as having the ability to create more plays for yourself. Anybody watching intently would have suspected this, but new statistics crystallized it as the truth that Errors, Putouts and Assists were distorting.
* All stats are from the amazing fangraphs.com and the line about Ethier being an Adonis was provided by artist Kim Alexander at a Dodgers game we attended earlier this season.