I’ve been wrestling with an idea for a while, and I finally sorted something out today. There are a lot of different measurements of defense and defensive capabilities, but they are all inherently flawed. If you look simply at fielding percentage, then you automatically assume that all players have the same ability to get to a ball. If you try to factor in a player’s range, then a lot of subjectivity comes ino play. If you talk about raw tools like speed or arm strength, you can’t take into account baseball intelligence, like the ability to make the right read on a ball or get in the best position to make a play. Then there is the positional adjustment. Justin Morneau is a awesome defensive first baseman, and Brian Dozier is an average second baseman. That said, Dozier is probably a better fielder than Morneau.
Trying to quantitatively evaluate an individual’s defensive ability is hard.
I think, though, you can figure out how good a team defense with a fairly reliable degree of certainty. Most every fielding independent pitching statistic takes into account batting average for balls in play, and dub it a component of luck. I think it can also help explain some things about a team defense.
Now, all batted balls are not the same. Sometimes, the ball is hit harder than other times. Batted ball statistics are out there, telling you what percentage of batted balls are ground balls, fly balls and line drives. A team that has more line drives hit versus them seems like the kind of team that would have a higher BABIP against, but that might not be luck, that could very well be a mediocre pitcher serving up meatballs.
What I wanted to do, therefore, was regress LD% versus BABIP. I used 2012 statistics for all 30 teams and came up with an equation for the best fit line. The line that I came up with had an R^2 of around .36 which suggests the equation isn’t all noise. There is still a bit of luck, but the variation could be attributable to different levels of skill in the teams. I am only using 30 data points, after all.
The equation that resulted was .8677*(LD%) +.1121 = BABIP (predicted). I then found the difference between the predicted BABIP and actual BABIP which, I think, is a huge step towards telling us what teams are helping their staff the most on defense. Really, the only problem is with the numbers it creates. Some of them are negative and almost all of them have an absolute value of less than .02. The first thing I wanted to do was multiply the number I had by 100 so it would read as an actual number, rather than something like “point zero zero” etcetera. This change was to make things less combersome. After that, I added 3 to my new number, which was enough to bring all the scores above 0, which will help with any further math we want to apply to the results. With that in mind, an average PID for a given team is 3.
Last but not least, and since errors don’t go against a teams batting average or BABIP, I wanted to take into account fielding percentage. After all, getting to the ball is only part of the battle. You have to make the play when you get there too. So I took our raw PID and multiplied it by fielding%. In truth, only one team really benefitted from this. The White Sox had one of the best filding percentages in the league last year, and as a result nosed ahead of the Marlins and Reds to have the 9th best PID in the league.
So who were the best teams, defensively, by this metric? The top 5:
Oakland: 4.96
Seattle: 4.10
Chicago Cubs: 4.04
Los Angeles Angels: 3.97
Los Angeles Dodgers 3.90
Interesting to see how much further beyond the other teams the A’s are. Is it possible that the A’s have found another market inefficiency? (Yes) Is this it? (Defense might be it, but they certainly didn’t use my equation). The A’s BABIP was 20 points lower than what was predicted.
Who was at the bottom? The bottom 5, from 30-26
Houston: 0.61
Colorado: 1.21
Kansas City: 1.56
Cleveland: 1.97
Arizona: 2.30
It hurts Colorado that they play in an enormous and there is a lot more ground to cover. The other teams on this list? I don’t think anyone thought they were elite on defense. Certainly, there were good individual defenders on some of these teams, but the teams themselves were necessarily good on defense. For every Justin Upton, there is a Jason Kubel, so to speak.
This is all well and good, but does it mean anything. That’s what we will try to find out in part 2 on Wednesday.
Add The Sports Daily to your Google News Feed!
