Sunday, October 30, 2005

Component ERA Analysis

Component ERA (ERC) estimates what a pitcher’s ERA “should be” based on a pitcher’s performance. It is the companion of runs created for batters and is computed similarly using pitching stats instead of hitting stats. There are many versions of the formula. The specific formula for this analysis can be found at this site (author unknown). It’s the same formula that was printed in the 2005 Bill James Handbook.



Why use ERC when we already have ERA? ERA is based on timing of hits, walks and homeruns allowed. I’ll give an extreme example to illustrate the point. If a pitcher pitches 9 innings and gives up 9 hits with each hit coming in a different inning, he’ll give up fewer runs than if he gave up all the runs in one inning. On the other hand, ERC is just based on how many hits (and other things) they gave up and is not based on timing.



If you believe that timing of hits is characteristic of a pitcher, then ERA is a better measure. If you believe that timing of hits is random, then ERC is a better measure. Evidence shows that the hit timing of most pitchers varies greatly from year to year which indicates that hit timing is more a matter of random events than a characteristic of a pitcher. So ERC is generally a better measure of performance. In fact, it has been shown that ERC tends to be a better predictor of future performance than ERA.



The tables below display the component ERAs of the Detroit Tiger pitchers in 2005. The numbers were taken from the ESPN.COM statistics section. ERC% is the ratio of ERC to ERA. A pitcher with an ERC% significantly above 1 pitched worse than his ERA indicates and a pitcher with an ERC% below 1 pitched better than his ERA indicates.


Starters


Pitcher

IP

ERA

ERC

ERC%

Johnson

210

4.54

4.27

0.94

Maroth

209

4.74

4.74

1.00

Robertson

196.7

4.48

4.40

0.98

Bonderman

189

4.57

4.20

0.92

Douglass

87.3

5.56

4.83

0.87

Ledezma

49.7

7.07

6.66

0.94


Relievers

Pitcher

IP

ERA

ERC

ERC%

Urbina*

79

3.62

3.20

0.88

Spurling

70.7

3.44

3.04

0.88

Farnsworth*

70

2.19

2.12

0.97

Colon*

69.3

5.58

5.77

1.03

German

59

3.66

5.93

1.62

Walker

48.7

3.7

3.73

1.01

Rodney

44

2.86

3.70

1.29

Ginter

35

6.17

6.86

1.11

Dingman

32

3.66

3.95

1.08

Percival

25

5.76

4.35

0.76


*Includes combined statistics for both teams on which they pitched.


Based on their ERC%, Sean Douglass (0.87), Jeremy Bonderman (0.92), Jason Johnson (0.94) and Wilfredo Ledezma (0.94) pitched significantly better than their ERAs and would be expected to achieve lower ERAs next year if they pitched as well as this year.


Among relief pitchers, Troy Percival (0.76), Ugueth Urbina (0.88) and Chris Spurling (0.88) pitched significantly better than their ERAs indicated. Franklin German (1.62), Fernando Rodney (1.29), Matt Ginter (1.11) and Craig Dingman pitched worse than their ERAs indicated.

No comments:

Post a Comment

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