See the first part here. Today we’re moving on to the newest and latest of analytic crazes, defense. With the advent and improvement of such statistics as UZR by Mitchel Lichtman, Total Zone Rating by Sean Smith, and +/- by John Dewan, defense has come into the spotlight as what many are calling “the new OBP,” signaling the newest of sabermetric crazes. We’ve now learned that defense can play a huge role in the success of a team. Let’s consider some numbers.
The Effect of Defense
UZR and most other metrics convert plays made into run values for teams. As we know, run values are vitally important to statistical analysis because runs are how games are won and lost. Systems like UZR and TZ use linear weight models. models that have been used extensively in analyzing offense, for balls hit in certain zones and compares all players to the major league average. In doing this it provides a reasonable number of “runs saved” by defensive players. Runs can be simply translated into wins using the standard 10 runs/1 win ratio. Doing this provides some context to what teams like the Marlins are doing to help their pitching staffs, and gives us something quantifiable to use in judging a field that was previously held to only qualitative analyis.
Let’s look at the extremes on either end of the UZR spectrum. Data provided by FanGraphs.
Top 5 Defensive Teams (UZR)
Four of the five teams listed here are contenders in their division currently and were bottom-feeders last year. As the Tampa Bay Rays taught us last year, vastly improving your defense can lead to a significant shift in the standings. Even though it appears as if Seattle is saving 2.7 wins these season attributed to defense, the differences are drastic considering where these clubs were a year ago.
2007 Defensive Numbers (UZR)
When compared to their numbers the previous year, the change is wins saved from year to year is far more dramatic. Even if you expect and end-season line for the Tigers of about 35 runs above average, meaning the team did not change much more compared to the league average defensively in the second half, Detroit would have still netted around 74 runs better than their previous season on defense, good for a whopping 7.4-win bonus to last year’s record. If Seattle ends the season about 35 runs better than average, thanks to their drastically improved outfield defense, they will have saved close to 56 runs, or 5.6 wins more than last season. Often times, ballclubs benefit by just being in the positive defensively, even a more modest 8-10 runs better than average, because a larger number of teams are at the average or below. In this day and age of teams trying to find every edge they can get to succeed, statistical defensive analysis has become a tool that has drastically affected the standings this year and last.
Which takes me to the bottom half of this analysis and our team of interest.
Bottom 5 Defensive Teams (UZR)
I placed the ellipse to show that I was done writing the bottom five defensive squads by UZR, but the Marlins actually aren’t far from that list as the sixth worse squad in the majors. None of these teams should come as any surprise to you. The Nationals contain one of the worst defensive regulars in the game in Adam Dunn, the Mets have struggled all year long in fielding a baseball team that can actually field, thanks to injuries and their desire to find Daniel Murphy somewhere in the lineup, and the Braves have sported one of the worst regular outfields in the game thanks to their lack of range. Three of the worst teams in baseball reside in the bottom five in defense, and if you look in years past this has often been the trend. Check out last year:
Bottom 5 Defensive Teams, 2008 (UZR)
Two of those teams were in contention and one of them won their division (though admittedly that was Dodgers and they won the NL West, the worst division in baseball last year), but the other three teams were among the worst in their respective divisions (OK, the Rangers weren’t, but they were the third worst team in a bad division by Pythagorean expectation).
Now that I’ve had my long diatribe about defense, let’s talk a bit about the Marlins.
Runs Allowed: 399
2009 Team UZR: -20.0
2008 Team UZR: -2.4
The Marlins seemingly made long strides last year to improve a squad that was close to dead last in defense the year before. The team recorded essentially a league average UZR, backed by major improvement by middle infielders Hanley Ramirez and Dan Uggla and a monstrous range season by the surprising Cody Ross. Of course, the team saw their state brethren in St. Pete go to the World Series on the back of the defense and Larry Beinfest came into the offseason looking to improve the team’s biggest weakness. Gone was the statue of Mike Jacobs at first base, and moving into that position was the more nimble Jorge Cantu, who also happened to be the team’s terrible defensive third baseman last year. Replacing Cantu at the hot corner was speedy middle infielder Emilio Bonifacio, acquired from the Nationals in the Josh Willingham-Scott Olsen deal. The Marlins dealt Willingham and his subpar glove in left field and looked to replace it with Cameron Maybin in center.
On the surface these moves seemed to have improved the defense greatly. The truth of the matter was that the Marlins’ improvement the previous year was mostly a mirage. This season the Marlins regressed to what they are defensively and then some. Dan Uggla appeared to have benefited from Mike Jacobs’ ineptitude at first, as his range has dropped significantly from last season. Jeremy Hermida has turned into a boulder defensively, as he’s posted one of the worst outfielder UZR’s to date. It turns out Ross was not in fact a +11 center fielder and instead was merely covering for the ground not made up by Willingham and Hermida.
Since UZR is broken down into many different parts, we can look at individual run measurements for things such as range, outfield arm, double plays, and errors. Most differences between ballclubs involve range, as it is the largest component of UZR and indeed the first that was quantized by MGL in 2003. Two things really stand out for the Marlins this season: 1) Their double play runs, which have always been a team strength, are hurting the squad to the tune of -4.1 runs, and 2) The Marlins outfield arms are costing the team 7.2 runs so far this year.
The arms aren’t surprising. None of the Marlins’ outfielders had strong guns in the past, and this year is no different. However, the Marlins traded last season’s only positively-rated arm, Willingham, away to the Nationals and replaced him with Cameron Maybin, who struggled a bit but was mostly a sample-size victim. Ross in center field has an average arm, but his right field arm is not plus and he may have to remain in center to get the most of his defensive production. The difference between last season and this season has been a total of six runs.
The double play runs are far more interesting figure. Qualitatively, we’ve watched Hanley and Danny botch a lot more double plays this year than in years past. Uggla actually led all second basemen in double play runs the last three season, and Ramirez has been average for most of his career. However, this season Uggla has rated as -2 runs in terms of the double play, with Ramirez at -1 run. That being said, the difference is less than 3 runs over the course of half a season so far, so I would not get wound up about it.
Best Performer: Brett Carroll
I wanted to put a team regular in this spot, but no starter has been above league average this season, with only two players rating as essentially league average so far. Carroll on the other hand has been nothing short of excellent during his limited time in the outfield, mostly in right field. In only 142 innings in right field this season, Carroll has been worht 7.2 runs above average. Overall, Carroll has been worth 9.3 runs better than the average outfielder, an impressive feat in such a small amount of playing time.
Carroll’s best feature is his plus range; he’s been 6.4 runs better than the average outfielder with the same playing time. His arm has been effective as well, worth 1.9 runs above average. There are a multitude of sample size issues regarding this analysis however, and it’s simply not possible for Carroll to be this good over a long stretch of time. It is reasonable, given what we have qualitatively seen from him, that he could be a +10 to 15 run outfielder. What we know of his bat is enough to say it would be a stretch for play long-term, but the Marlins are sacrificing a significant number of runs defensively by playing a lineup involving Hermida, Coghlan, and Ross in the outfield. Carroll’s glove so far has been worth almost a win above replacement. If he brings a replacement bat with +15 run defense, that’s almost a win over replacement for the entire year, passable if guys like Hermida and Coghlan won’t hit well enough in the corners anyway.
Worst Performer: Jeremy Hermida
There were a lot of players worthy of this nod, but Hermida’s amazingly bad defensive year takes the cake. Hermida can only play the corner outfield positions, limiting his options in getting his bat in the lineup. And while his bat has been league average so far this season, he’s erased any benefit of his increased walk rate and patience numbers by being truly atrocious in the outfield. So far this season, Hermida has posted a whopping 12.9 runs below average playing both corner outfield positions. In just the first half of the season, he has eclipsed each of his previous three season totals.
The value is truly staggering. The last week or so since finding out just how bad he truly was, I’ve been referring to Hermida’s defense as Dunn-esque. Adam Dunn of the Nationals is posting similar but more drastic numbers, posting an absurd 16.2 runs below average and fully eliminating his amazing offensive season to date. Hermida has become Adam Dunn without the offensive ability. He is still an average offensive player, or at least has been this season, but he has completely lost his range, currently almost 10 runs below average. This has been clear qualitatively as well and, based on his previously poor track record on defense, can be expected to continue to some degree. Hermida is plodding in the outfield, a surprise given his atheltic build. He also has always lacked the arm for right field, sporting currently 3.2 runs below average in outfield arm runs. Hermida has also only committed one error so far this year, but has been somewhat error-prone throughout his career.
Unlike many of the Marlins players on the field, there isn’t much hope for improvement for Hermida’s terrible glove. His value is severely deflated because he has been so poor in the outfield, and if the Marlins continue to run him out onto the field, he will continue to eat up runs in right field. It’s imperative the Fish rest him and Coghlan a total of two times a week in order to fit Carroll’s plus glove to offset this sort of poor performance.
Key Second-Half Improvement: Emilio Bonifacio
It is rare for me to praise Bonifacio in a season where he has been atrocious with the stick, but I will admit that, since the Marlins are insisting on playing him full-time, he should make the most of the situation and play well. His bat is a completely lost cause and will always be replacement level or worse. However, his fielding at third base has improved over the course of the season, or at the very least has appeared to improve.
In my maddening rage at Bonifacio’s inability at the plate, I failed to recognize the reasoning behind his struggles on the diamond. He was posting an extremely poor UZR, but such small sample sizes were not really effective at telling the whole story of his defensive game. Looking at the numbers now, he’s at a more respectable but still reproachable 3.9 runs below average at third. However, if you look at the breakdown of his numbers, the problems he is having are not in the all-important range category, but in the error mark. Bonifacio leads all third basemen with 13 errors, but it is plausible to expect his error run total right now to improve if he begins to field the ball more cleanly. Third base is a position prone for error, and if Bonifacio is indeed improving in his reflexes at the hot corner, he can perhaps cut down on his errors and allow the rest of the league’s third baseman catch up.
Bonifacio has shown he has league average range at third so far this year. If he can improve as he gets his bearings at the hot corner, he may be able to improve into the positives as he continues playing. I don’t doubt that it would take time, but if he can improve just slightly in his range this season (again, small sample size issues in hoping for this, as the second-half”s innings won’t be enough to be conclusive about anything) and show a little more sure-handedness, he could become a +2 run third baseman this season. Currently he sits at -4 runs, but the majority of those are error runs that ideally are behind him.
That being said, this is an improvement of six runs, almost a win but barely anything to get excited about. Still, anything to improve Bonifacio’s value can help at this point, as he continues his season-long ineptitude at the plate and Fredi continues to give him chances.
Next up, we’ll look into the team’s pitching staff and how the Marlins may have to ride two aces to the finish line if they want to compete in the NL East.