Monday, December 17, 2007

FIP Analysis for Tiger Relievers in 2007

In an earlier post, I discussed team run prevention using FIP ERA to measure pitching performance and DER to measure fielding performance. Then I evaluated the performance of individual starting pitchers using FIP ERA. Today, I’ll look at the relievers. There were 82 primary relievers (more relief appearances than starts) with 40 or more innings pitched in 2007. Table 1 below lists Detroit Tiger relievers in 2007. Table 2 lists all 82 qualifiers in the league.

In both tables, the key variables are actual ERA, FIP ERA, FIP-Actual and LOB%. All are defined in the article on starters linked above. Since relievers pitch fewer innings than starters, all of these statistics are less reliable for relievers and should be interpreted with a little more caution. The averages for relievers in 2007 were: FIP (4.09), DER (0.71) and LOB% (0.74).

The following relievers had FIP ERAs which were worse than their actual actual ERAs: Bobby Seay (FIP ERA = 2.86, ERA = 2.33), Tim Byrdak (3.60, 3.20), and Zach Miner (3.85, 3.02). This is an indication that they may have not pitched quite as well as their ERAs indicated. On the other hand, Todd Jones (3.77, 4.26) and Jason Grilli (4.04, 4.74) had FIP ERAs lower than their ERAs which suggests that they may have been somewhat unlucky.

One thing that comes out of this analysis is that all six qualifying relievers had FIP ERAs which were league average or better. Since FIP ERA is a better predictor than ERA, this gives us reason for cautious optimism for 2008. However, keep in mind that sample sizes (of innings) for relievers are very small and that it's very hard to project relievers into the future. Also note that most of these pitchers do not have long track records of success. The bullpen is going to be another key factor to the team's success in 2008 and I'm still hoping they add another pitcher with a better track record.

The raw data for this report were abstracted from The Hardball Times database.


Table 1: FIP ERA Ranks for Tiger Relievers in 2007

FIP ERA Rank

Name

IP

ERA

FIP ERA

DER

FIP-Actual

LOB %

12

Seay

46.3

2.33

2.86

.722

0.53

.80

26

Byrdak

45.0

3.20

3.60

.708

0.40

.69

30

Jones

61.3

4.26

3.77

.704

-0.49

.70

32

Miner

53.7

3.02

3.85

.694

0.83

.76

36

Rodney

50.7

4.26

3.95

.707

-0.31

.68

39

Grilli

79.7

4.74

4.00

.694

-0.74

.65





Table 2: FIP ERA Ranks for AL Relievers in 2007

FIP ERA Rank

Name

Team

IP

ERA

FIP ERA

DER

FIP-Actual

LOB %

1

Betancourt

CLE

79.3

1.47

2.25

.760

0.78

.86

2

Soria

KC

69.0

2.48

2.51

.750

0.03

.74

3

Jenks

CHA

65.0

2.77

2.52

.757

-0.25

.69

4

Papelbon

BOS

58.3

1.85

2.59

.785

0.74

.88

5

Street

OAK

50.0

2.88

2.70

.748

-0.18

.68

6

Nathan

MIN

71.7

1.88

2.71

.724

0.83

.86

7

Rivera

NYA

71.3

3.15

2.71

.678

-0.44

.76

8

Bale

KC

40.0

4.05

2.81

.627

-1.24

.73

9

Putz

SEA

71.7

1.38

2.81

.803

1.43

.94

10

Brown

OAK

41.7

4.54

2.85

.675

-1.69

.65

11

Rodriguez

LAA

67.3

2.81

2.85

.701

0.04

.78

12

Seay

DET

46.3

2.33

2.86

.722

0.53

.80

13

Perez

CLE

60.7

1.78

3.05

.766

1.27

.84

14

Tallet

TOR

62.3

3.47

3.16

.730

-0.31

.70

15

Sherrill

SEA

45.7

2.36

3.18

.769

0.82

.84

16

Benoit

TEX

82.0

2.85

3.23

.710

0.38

.78

17

Downs

TOR

58.0

2.17

3.23

.714

1.06

.84

18

Frasor

TOR

57.0

4.58

3.26

.716

-1.32

.63

19

Okajima

BOS

69.0

2.22

3.38

.762

1.16

.86

20

Embree

OAK

68.0

3.97

3.45

.703

-0.52

.71

21

Thornton

CHA

56.3

4.79

3.52

.661

-1.27

.69

22

Accardo

TOR

67.3

2.14

3.53

.750

1.39

.81

23

Bradford

BAL

64.7

3.34

3.57

.679

0.23

.73

24

Green

SEA

68.0

3.84

3.57

.646

-0.27

.75

25

Neshek

MIN

70.3

2.94

3.59

.780

0.65

.76

26

Byrdak

DET

45.0

3.20

3.60

.708

0.40

.69

27

O'Flaherty

SEA

52.3

4.47

3.63

.723

-0.84

.64

28

Greinke

KC

122.0

3.69

3.76

.686

0.07

.76

29

Peralta

KC

87.7

3.80

3.76

.689

-0.04

.74

30

Jones

DET

61.3

4.26

3.77

.704

-0.49

.70

31

Oliver

LAA

64.3

3.78

3.83

.725

0.05

.68

32

Miner

DET

53.7

3.02

3.85

.694

0.83

.76

33

Janssen

TOR

72.7

2.35

3.89

.727

1.54

.81

34

Casilla

OAK

50.7

4.44

3.93

.730

-0.51

.72

35

Delcarmen

BOS

44.0

2.05

3.93

.786

1.88

.87

36

Rodney

DET

50.7

4.26

3.95

.707

-0.31

.68

37

Speier

LAA

50.0

2.88

3.96

.767

1.08

.80

38

Wilson

TEX

68.3

3.03

3.97

.743

0.94

.77

39

Grilli

DET

79.7

4.74

4.00

.694

-0.74

.65

40

Shields

LAA

77.0

3.86

4.00

.724

0.14

.71

41

Guerrier

MIN

88.0

2.35

4.02

.750

1.67

.88

42

Gobble

KC

53.7

3.02

4.03

.671

1.01

.80

43

Walker

BAL

61.3

3.23

4.05

.734

0.82

.75

44

Borowski

CLE

65.7

5.07

4.08

.670

-0.99

.68

45

Morrow

SEA

63.3

4.12

4.09

.686

-0.03

.76

46

Ray

BAL

42.7

4.43

4.11

.727

-0.32

.69

47

Bootcheck

LAA

77.3

4.77

4.12

.690

-0.65

.67

48

Vizcaino

NYA

75.3

4.30

4.12

.729

-0.18

.72

49

Moseley

LAA

92.0

4.40

4.16

.696

-0.24

.70

50

Francisco

TEX

59.3

4.55

4.21

.693

-0.34

.69

51

Lopez

BOS

40.7

3.10

4.22

.726

1.12

.76

52

Mastny

CLE

57.7

4.68

4.23

.665

-0.45

.76

53

Calero

OAK

40.7

5.75

4.36

.664

-1.39

.66

54

Riske

KC

69.7

2.45

4.41

.736

1.96

.90

55

Wolfe

TOR

45.3

2.98

4.44

.772

1.46

.75

56

Parrish

BAL

41.7

5.40

4.51

.691

-0.89

.68

57

Myers

NYA

40.7

2.66

4.63

.737

1.97

.81

58

Timlin

BOS

55.3

3.42

4.66

.767

1.24

.75

59

Duckworth

KC

46.7

4.63

4.73

.706

0.10

.64

60

MacDougal

CHA

42.3

6.80

4.73

.641

-2.07

.59

61

Nunez

KC

43.7

3.92

4.75

.717

0.83

.77

62

Ortiz

MIN

91.0

5.14

4.75

.691

-0.39

.68

63

Logan

CHA

50.7

4.97

4.80

.683

-0.17

.71

64

Marshall

OAK

42.0

6.43

4.88

.689

-1.55

.60

65

Eyre

TEX

68.0

5.16

4.95

.688

-0.21

.69

66

Reyes

TB

60.7

4.90

4.95

.757

0.05

.69

67

Camp

TB

40.0

7.20

4.98

.582

-2.22

.69

68

Villone

NYA

42.3

4.25

5.01

.752

0.76

.74

69

Glover

TB

77.3

4.89

5.05

.691

0.16

.72

70

Farnsworth

NYA

60.0

4.80

5.08

.717

0.28

.71

71

Wright

TEX

77.0

3.62

5.09

.724

1.47

.76

72

Rincon

MIN

59.7

5.13

5.11

.694

-0.02

.70

73

Bell

BAL

53.0

5.94

5.12

.656

-0.82

.69

74

Littleton

TEX

48.0

4.31

5.13

.731

0.82

.75

75

Bruney

NYA

50.0

4.68

5.40

.729

0.72

.73

76

Proctor

NYA

54.3

3.81

5.53

.732

1.72

.79

77

Snyder

BOS

54.3

3.81

5.53

.756

1.72

.74

78

Wood

TEX

50.7

5.33

5.77

.674

0.44

.69

79

Rheinecker

TEX

50.3

5.36

5.78

.675

0.42

.68

80

Fossum

TB

76.0

7.70

5.81

.643

-1.89

.59

81

Stokes

TB

62.3

7.07

5.85

.641

-1.22

.67

82

Baez

BAL

50.3

6.44

6.14

.738

-0.30

.67

6 comments:

  1. I'm not surprised to see that FIP is higher then ERA for that bunch of pitchers. Notice that this group are the ones that often get called in to face one or two batters instead of to work an entire inning.

    As you know, ERA is a terrible stat for relievers because they often come in with 1 or 2 outs and thus have lower chance of having their mistakes result in runs (at least runs that count against "their" ERA). FIP is a great alternative.

    I'm guessing that FIP doesn't correlate as well to ERA in the reliever group as compared to the starters due to the above phenomenon.

    ReplyDelete
  2. Relief pitchers are definitely a difficult bunch to evaluate. I need to read up some more on leverage index and see what I can do with that.

    Another thought I have is to develop a statistic for quality relief appearances (QRA). That is, what % of a reliever's appearances are effective ones? For example, if Todd Jones enters the 9th inning with a 3 run lead and gives up 2 runs, he would get a save but probably not a QRA. If Bobby Seay enters a tie game, pitches a scoreless 8th and Rodney loses it in the 9th, Seay would get a QRA even though he had nothing else to show for it.

    ReplyDelete
  3. Another thought... pickoffs by pitchers should be directly subtracted from walks. This may improve FIP slightly.

    FIP also doesn't take into account a pitchers ability to hold runners. Some pitchers are much better than others. Unfortunately, this burden is shared with the catcher so it would be difficult to incorporate this into a FIP.

    Wild pitchers could also be considered but once again, this is a very subjective stat (live errors).

    ReplyDelete
  4. I'm not sure whether pickoffs should be categorized under pitcher defense or pitcher pitching. There's a fine line there.

    ReplyDelete
  5. This comment has been removed by a blog administrator.

    ReplyDelete

Twitter

Blog Archive

Subscribe

My Sabermetrics Book

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

Other Sabermetrics Books

Stat Counter