Tuesday, November 11, 2008

Plate Discipline Ratio

A term we hear over and over again in baseball is plate discipline. Plate discipline or strike zone judgment means different things to different people. One simple definition would be the recognition of balls and strikes and the ability to react accordingly. That is, swing at strikes and don't swing at balls.

Another way to define plate discipline would be the ability to balance patience with aggression. It is good for hitters to work the count and make pitchers throw more pitches but it's also important for them to swing when they see a pitch they can hit.

In the past, the best statistic for measuring plate discipline was probably walk/strikeout ratio (BB/K). One of the problems with this measure is that a hitter might strike out a lot not because of poor strike zone judgement but because he has poor pitch recognition (e.g. fastball versus change-up) or because he is physically unable to make consistent contact. Then there are some sluggers who don't necessarily have great discipline but they walk a lot because pitchers won't throw a lot of strikes to them. With those caveats in mind, the 2008 Tigers are ranked (among 126 AL players with 300 or more plate appearances) on BB/K in Table 1 below.

Table 1: Walk/strikeout ratios for Tigers in 2008

NAME

AB

BB/K

Lge Rank

Carlos Guillen

420

0.90

6

Placido Polanco

580

0.81

15

Magglio Ordonez

561

0.70

27

Gary Sheffield

418

0.70

28

Curtis Granderson

553

0.64

42

Edgar Renteria

503

0.58

57

Brandon Inge

347

0.46

85

Miguel Cabrera

616

0.44

89

Marcus Thames

316

0.25

119



Thanks to the Baseball Info Solutions play by play database and researchers at Fangraphs.com, there are now better ways to measure plate discipline which address the problems of BB/K discussed above. The statistic ZSWING shown in Table 2 below is the percentage of pitches in the strike zone at which a hitter swings. OSWING is the percentage of pitches outside the strike zone at which a hitter swings. From those, I developed a new statistic simply by dividing ZSWING by OSWING. I call it Plate Discipline Ratio (PDR).

Here is an example of how PDR works: Magglio Ordonez had a ZSWING of 0.76 meaning that he swung at 76% of pitches in the strike zone. His OSWING was .23 meaning that he swung at 23% of balls outside of the strike zone. PDR = .76/.23 = 3.34. This means that he was 3.34 times more likely to swing at pitches inside the zone than outside the zone. This is relatively good - 16th best in the league. Miguel Cabrera had a PDR of 2.01 which means he was only about twice as likely to swing at at pitches in the zone as was to swing at a pitches outside the zone. This formula goes right in line with definition of plate discipline in the first paragraph. We can say that Ordonez had good plate discpline last year and that Cabrera had poor plate discipline (no pun intended).

Table 2: Plate discipline statistics for Tigers in 2008

NAME

AB

ZSWING

OSWING

plate discipline ratio

Lge Rank

Magglio Ordonez

561

0.76

0.23

3.34

16

Carlos Guillen

420

0.69

0.21

3.24

21

Curtis Granderson

553

0.59

0.20

2.95

36

Brandon Inge

347

0.60

0.20

2.95

37

Gary Sheffield

418

0.59

0.21

2.81

50

Placido Polanco

580

0.61

0.24

2.54

79

Edgar Renteria

503

0.68

0.28

2.46

83

Marcus Thames

316

0.78

0.32

2.43

89

Miguel Cabrera

616

0.69

0.34

2.01

111


Looking at some correlations, I saw that there was no correlation between PDR and either contact percentage or batting average. Thus, it's clear that there is no relationship between the two skills: plate discipline and hitting for average. What PDR is correlated with is BB% (r=.72) and BB/K (r=.50). A good feature of PDR is that it remains pretty consistent from year to year (r= .80 based on two years of data). This makes me think that it's measuring a real skill.

one interesting thing I noticed looking at past years of data is how much Cabrera's PDR has dropped in the past two years. He went from 2.92 in 2006 to 2.46 in 2007 to 2.01 in 2008. That's a 31% drop which is not what you'd expect from a stat that is quite repeatable. Most of it has come from him swinging at more pitches outside the zone. He had an OSWING of .23 in 2006 and .34 in 2008. I'm not sure whether this is a good thing or a bad thing. If he is becoming less disciplined over time, that's not a good thing because his OBP is dropping with it. However, the fact that he did have good discipline in 2005-2006 makes me think he can have it again if he works at it. It will be an interesting thing to watch next year.

4 comments:

  1. Wow, nice work. This is great stuff.

    ReplyDelete
  2. Jeff MorfordNovember 12, 2008

    Interesting stuff as always.

    Somehow OSWING seems more real to me. If batters can really 'wait for their pitch' then the highest ZSWING might not be the best disciplined hitter. (Although clearly a low ZSWING is a real problem!)

    Do you know how OSWING correlates to bb/k and correlate year to year?

    ReplyDelete
  3. Jeff, I agree that OSWING is the more important part of the equation. And statistically, it influences the equation more than ZSWING because players vary more on OSWING than ZSWING. The correlations are at home
    but I remember the year to year correlation for OSWING was even even higher that that for PDR. I'll check them for you later.

    One of the reasons I wanted to bring ZWSING into the equation was that OSWING by itself doesn't necessarily indicate bad discipline. If a batter happens to be a free swinger and is good at making contact, that's not a bad thing. A hitter like that should have a high OSWING but should also have a proportionately high ZSWING.
    If the ZSWING is not proportionately high, then I think the high OSWING becomes a problem.

    It's really just an experimental stat though. I need to look at it more.

    ReplyDelete
  4. Just to follow up on the last comment, the correlation between oswing and bb/k is -.54. The year to year correlation of OSWING is .84.

    ReplyDelete

Sabermetrics Book

Sabermetrics Book
One of Baseball America's top ten books of 2010

Blog Archive

Subscribe

501 Baseball Books

501 Baseball Books
Recommended by Tiger Tales

Stat Counter

Site Meter