The Tigers Defense – What are the Odds?

Over the last month or so, David Pinto has released the majority of his studies using his probabilistic model of range (PMR). Today we’ll delve into the Tigers defense using this advanced metric.

I’ve explained PMR in the past, but a refresher is probably worthwhile. The PMR model uses data play by play data collected by Baseball Info Solutions. Pinto uses 3 years of this data to find out the probability that a batted ball will be converted into an out. In doing this he accounts for the handedness of the batter and pitcher, the type of hit (grounder, fly, etc), how hard the ball was hit, and the direction the ball was hit. The beauty of the system is that it provides context to the data. Players who have harder to field opportunities get credit it for it. It also removes the subjectivity of an official scorers decision.

What the system doesn’t do is account for throwing ability for outfielders. So a Jacque Jones upgrade in range would be lessened by a weak throwing arm.

On to the data. The first table shows how the Tigers fared by position.

Position In play Plays Exp Plays DER Exp DER Rate Runs
Pitcher 4486 167 159.73 0.037 0.036 104.55 5.5
First Base 4486 296 310.16 0.066 0.069 95.44 -10.7
Second Base 4486 505 494.43 0.113 0.11 102.14 8.0
Shortstop 4486 517 536.95 0.115 0.12 96.28 -15.0
Third Base 4486 446 426.09 0.099 0.095 104.67 15.9
Left Field 4486 327 331.6 0.073 0.074 98.61 -3.8
Centerfield 4486 468 445.78 0.104 0.099 104.98 23.0
Right Field 4486 318 319.88 0.071 0.071 99.41 -1.6


The column definitions are as follows:

  • In Play: the total number of balls in play for the 2007 season against the Tigers
  • Plays: the actual number of balls converted to outs by each position
  • Exp Plays: the number of balls that would be expected to be converted to outs based on the aforementioned factors
  • DER: Defensive efficiency ratio. Plays divided by In Play
  • Exp DER: Calculated based on types of balls put in play. Used to produce Exp Plays.
  • Rate: Plays divided by expected plays
  • Runs: The runs saved above or below expected. Calculated using methodology described here and here.

Most of these results shouldn’t be too surprising. The Tigers struggled with defense at shortstop, first base, and left field. All 3 of those positions will have new help coming in 2008. Meanwhile they excelled in centerfield, third base, and were strong at shortstop. Tigers pitchers defended their positions well, but I don’t think anything will put an end to PFP jokes.

Pinto also publishes the results for individual fielders. I have listed the Tigers primary defenders at each position. The runs column I based on 4000 balls in play which is roughly 145 games.

Position In play Plays Exp Plays DER Exp DER Rate Runs/4000
Sean Casey 3100 198 211.63 0.064 0.068 93.56 -13.3
Placido Polanco 3724 420 409.07 0.113 0.11 102.67 8.9
Carlos Guillen 3361 389 408.05 0.116 0.121 95.33 -17.1
Brandon Inge 4062 400 380.28 0.098 0.094 105.18 15.5
Craig Monroe 2512 166 174.76 0.066 0.07 94.99 -11.6
Curtis Granderson 3995 424 402.22 0.106 0.101 105.42 22.5
Magglio Ordonez 3835 261 264.54 0.068 0.069 98.66 -3.1

And the bench…

Position In play Plays Exp Plays DER Exp DER Rate Runs/4000
First Base 1386 98 98.53 0.071 0.071 99.46 -1.2
Second Base 762 85 85.36 0.112 0.112 99.58 -1.4
Shortstop 1125 128 128.9 0.114 0.115 99.30 -2.4
Third Base 424 46 45.81 0.108 0.108 100.41 1.4
Left Field 1974 161 156.84 0.082 0.0795 102.65 7.0
Centerfield 491 44 43.56 0.090 0.089 101.01 3.7
Right Field 651 57 55.34 0.088 0.085 103.00 8.7

On the whole, the Tigers bench defense was quite solid. Across the board they were pretty much right in line with their expected level of production. Of course with bench players you’d expect them to be strong defensively because often times that is what is keeping them in the league (see Ramon Santiago).

Also notice that the Tigers 3 weakest defensive positions received the most support from the bench. The first step is recognizing the problem, which clearly the Tigers did last year and were quick to address this year.

Out of Left field

Left field is a position that really jumps out. Craig Monroe was pretty poor and accounted for 58% of the innings logged by Tigers in left field. Yet the position as a whole was pretty neutral because Marcus Thames (277 inn), Timo Perez (158 inn), Cameron Maybin (80 inn) and Ryan Raburn (58 inn) were above expected.

Thames and Raburn are good candidates to continue to log innings out there, but the lions share will be held by Jacque Jones. Jones was primarily a centerfielder last year, but was mostly a right fielder prior to that. As a right fielder he rated about 7 runs above expected. So the Tigers should expect to see a positive in net contribution in left field – possibly to the tune of a full win just defensively.

6-3

The other positional shake-up of course is Renteria to short and Guillen to first. Guillen had limited time at first base with only 545 balls in play while he was manning the position. However during that time he turned in 4 more plays than expected. Over a full season that would be 22 plays above expected and over 17 runs. That is potentially a 30 run swing in defense at first base based on range.

A couple caveats here. First, Guillen’s data is limited so projecting a full season based on it should be done with a grain of salt. But placing a shortstop’s range at first, even a declining shortstop, should be a substantial upgrade over a statue-esque Casey. Second, these measures don’t account for a first baseman’s ability to field throws – something that Casey excelled at. But without Guillen having to field his own throws, that task should be easier.

As for shortstop, PMR rates Edgar Renteria as a little below expected. He was -4.13 plays over 3067 balls in play. On a run basis per 4000 that is -4.1 runs.

The Tigers stand to see a real upgrade at 2 infield positions as well as in left field. The total defensive improvements could be in the magnitude for 3 to 5 wins.

I drew on many posts at Baseball Musings, but they can all be found on the PMR category page.