Over the past few days, statistical reports have been rolling in as people celebrate the final days of the regular season. One of the most jarring to me, was some discussion about the outfield of the Atlanta Braves. The Braves featured three excellent players in their outfield this season: Jason Heyward, Michael Bourn and Martin Prado. All three are good all-around players, capable of producing value with the bat, while also playing solid defense in the outfield. I don't think anyone questions this.
However, the wins above replacement statistics (rWAR and fWAR) tell a different story. According to fWAR, these three Braves outfielders aren't just good, they were in the top-15 most valuable position players in the major leagues. According to rWAR, they were in the top-17. With a combined fWAR of 18.9, and a combined 16.9 rWAR, they have a claim to be one of the best outfield trios in major-league history. And this all comes despite the fact that none of these players hit more than 21% better than the league average!
A huge amount of each player's above-replacement value this season, by the stats, comes from their defensive contributions. Take a look at this table below:
Note: Yes, I did factor out Prado's time at other positions here.
Both major sides of the WAR debate -- FanGraphs and Baseball-Reference -- agree, the Braves might have one of, if not THE greatest three-player outfield(s) in recent history.
I don't think so, guys.
I love defense, and defensive metrics. LOVE them. I just bought two new books on them: Wizardry and The Fielding Bible Vol. III. And I plan on writing many, many more articles on fielding, and how to quantify it, before my time on this earth is done. But the metrics used to identify just how good this outfield was in 2012 -- well, it doesn't sit right with me. If these players were all so individually phenomenal as defenders, how did this happen, and why isn't the team's overall defense better?
Think about this for a moment. UZR (the defensive component of FanGraphs's fWAR) says that between these three players, their defensive contributions were worth five and a half wins above replacement. That's enormous. I mean, really big. It's actually about equivalent to the weight of Austin Jackson's ENTIRE 2012 output, including both offense and defense. DRS is a component of rWAR, and gives these three guys credit for about five wins there also, or about another Austin Jackson-worth of value.
Perhaps the Braves went out and acquired three of the best defensive outfielders in the game today? Perhaps these three players all have stellar defensive records that prove that this season isn't just an outlier in their fielding records? Let's look at the past three years and see.
Note: Again, only OF for Prado
Ok, some of this is weird, and some of it buoys the argument that they really are otherworldly-caliber fielders. Bourn was actually quite awful by both UZR and DRS metrics last season, but was very good again in 2010. Heyward has posted great DRS numbers in every season, but this season's UZR looks like an outlier. And Prado only has two seasons with significant outfield experience, but he took a big jump this year (in a roughly-similar amount of innings).
The numbers show that there's definitely some volatility at play here. The same exact configuration last season was worth a combined 15 runs via DRS (compared to 56 in 2012), and 7.5 runs via UZR (compared to 54.4 in 2012). This happens sometimes with fielding stats, so let's look at some other ways to measure their efficacy.
If the outfielders on this team were so valuable on defense, a logical corollary might be that the team, as a whole, would post excellent defensive numbers by another metric. A metric like, say, PADE. PADE is short for park-adjusted defensive efficiency, and can be found at Baseball Prospectus. According to BP, the Braves have a PADE of 1.21, which puts them at 10th in the major leagues. This score is between "great' and "average", but closer to "great". One might think that a world-class defensive outfield might move the needle a bit more.
In 2011, the Tampa Bay Rays had a legendary season by PADE standards. Their rating of 4.30 dwarfed all other teams, no squad had anywhere near that number, and all sat below 2.00 for the season. One would assume that a team in that position would be worth more fielding runs than the 2012 Braves, right? Well, according to UZR, that Rays squad was worth 55.2 fielding runs above replacement. The 2012 Braves? 57.2 fielding runs above replacement. Again, odd. Why would a Braves squad that's equivalent based on UZR be so much inferior when it comes to PADE? UZR is park-adjusted as well.
The Other Guys
Okay, so given THAT, perhaps the rest of the Braves defenders were just so awful, that the Braves outfield makes up the difference? I mean, never mind that Andrelton Simmons played some of the most statistically-superb defense at shortstop in his short time with the team in ML history (seriously, guys -- he racked up 19 DRS and 9.7 UZR in just 426 innings!), maybe the rest of the team was awful? I mean, I know Tyler Pastornicky was bad at shortstop, right?
It turns out that most Braves defenders put up ok, but not good numbers when it came to UZR and DRS, especially among the ones who got substantive playing time. Pastornicky (-12 runs) was awful, but Simmons (9.7 runs) and Paul Janish (2.5 runs) made up for his failings over time. Freddie Freeman (-4.8 runs) was no great shakes at first, but even less-heralded fielders Brian McCann (2.6 runs), Dan Uggla (2.3 runs) and Chipper Jones (0 runs) didn't tip the scale in a negative way.
So from the numbers I see at this point, the rest of the team played replacement-level defense, but the outfield must have been completely phenomenal. Both UZR and DRS have the Braves as the best defense in the majors by a landslide, with a UZR of 44.5 and a DRS of 55. The Los Angeles Angels, another team with a stellar defensive outfield, not to mention Albert Pujols at first base, only managed 35.4 UZR and 46 DRS. No other team topped 28 in either metric, and only six other teams cracked double-digit scores for either at all. Oh, and the Angels were also second in baseball with a 2.32 PADE. Again, the Braves excellence doesn't exactly sound right.
Batted Ball Data
Maybe we could try checking the batted-ball data? Perhaps if the Braves pitchers gave up a large number of fly-ball outs, we could attribute more value to the Atlanta outfielders, no? Well, it turns out that, according the batted-ball data at FanGraphs, the Braves' pitching staff was a ground-ball-heavy staff. In fact, this team had one of the lowest GB/FB rates (1.46) in all of baseball. Only the Cardinals, Rays, Astros and Padres gave up fewer ground balls.
I'll admit, this could mean that the Braves infielders (who were, ostensibly, inferior defenders) had more chances to drag down overall team defensive value, but it also limits how much combined value the Braves outfielders likely had. And for the record, the Braves' park factors at Turner Field, according to FanGraphs, make for slightly more singles, and slightly less extra-base hits that aren't homers. So take that for what it's worth.
Oh, and also, when it comes to BIP-Wins for pitchers, the Braves are good (4.3 BIP-Wins for 2012), but not the top of the heap. The Angels, Rays, Nationals, Dodgers and Athletics were all better. I'm not sure how much that says about team defense, but it's another data point to do with what you will.
Look, I'm not here to say that the Braves outfielders weren't excellent this season. I'm not even trying to say that they're not all excellent fielders. What I am trying to say, is that the data paints a conflicting picture here. UZR and DRS say that the Atlanta fielders were all among the elite in defensive performance last season. PADE seems to say that the overall team defense wasn't elite, despite these performances, and the batted ball data makes me question how much possible value this elite defensive performance would actually have for the team.
When Michael Lichtman wrote up an primer for UZR over at FanGraphs, he even warned us about the perils of using a single season (or more) of UZR data.
"So, even after regression, there is no guarantee that our UZR number reflects what the player actually did or his true defensive talent over that time period. But, it is the best we can do (not knowing anything else about that player)!"
Ascribing massive amounts of value to a player's single-season defensive performance, when based even on the best stats, is a mistake. We should value the defensive contributions that Bourn, Prado and Heyward bring to the table. But given that their bats aren't exactly world-class, and their actual defensive value might be a little hard to accurately describe, don't you think that maybe we should hold off on calling them one of the all-time great outfields? It'd be my opinion that a world-class hitting outfield -- such as the '64 Twins or the '66 Braves or even the '01 Cardinals -- have a much better claim to that throne, without relying on so much value coming from defensive metrics that may not be accurate.