In part one of this series, I presented a system for the analysis of NCAA Division I stats for hitters. In that piece, I hinted at but did not actually present a system for analyzing pitchers' statistics. This article will explain the analysis of pitchers a bit more fully and present a top 50 for 2003. A top 50 for 2002 will follow shortly.
Readers should refer to that earlier article to get some background on this project. As I discussed there, I used RSAA as the basis for my analysis of pitchers. I have used runs allowed, rather than earned runs, to do these calculations in particular because of the extremely wide variation in scoring practices at 280 mostly uncoordinated schools.
The data for most schools (well over 90% in 2003) includes the full breakdown of offensive events for all pitchers. This allows us to use xRuns in the pitching context, by calculating xRuns allowed. I prefer this to component ERA, since component ERA formulas aren't generally very portable from context to context. Basically, xRuns allowed (where it's available) does a very good job of predicting runs allowed, but makes allowances for where pitchers have lowered ERAs or RAs due to being relievers, getting lucky, or getting good bullpen support. Or vice versa.
I think the xRuns calculation is slightly more accurate than runs allowed in analyzing a pitcher's true effectiveness, though of course it doesn't take into account clutch pitching ability and adds in a scorer's bias. Some pitchers are just very, very lucky or unlucky and wind up with runs allowed totals that vary widely from how they performed in individual contexts.
Just as I did for hitters, I calculated park adjustments for each pitcher, both to runs allowed and to xRuns allowed. I also calculated competition adjustments, as detailed in my hitters article. I used these figures to compare each pitcher to the NCAA average, both in terms of RSAA, and as RA+ - which is calculated like the familiar ERA+ figures – 100 is average, and a pitcher who allows half the number of runs of the average pitcher is a 200.
The Top 50 pitchers by **RSAA (park-and-competition-adjusted runs saved above average) are listed below. Before signing off, though, I'd like to give a short update on the project. Robert Dudek, analyst extraordinaire as he is, has been contributing some new park factor calculations which I am hoping to include in the final release of the full statistical breakdowns. I still intend to put up 2002 top 50 data, for both hitters and pitchers, in the very near future – within the next week or so. A lot of work needs to be done on cleaning up the database, removing double entries and so forth, as well as some stray team totals that got mixed in, and there are quite a large number of player names that have been somehow corrupted (Harvard's and Illinois' pitching staffs apparently have no first names, St. Louis's pitchers no last names, to name just three!) so I need to chase those down. Once that is done, a more analytical article should follow along with a full listing of every player for 2002 and 2003.
Readers should refer to that earlier article to get some background on this project. As I discussed there, I used RSAA as the basis for my analysis of pitchers. I have used runs allowed, rather than earned runs, to do these calculations in particular because of the extremely wide variation in scoring practices at 280 mostly uncoordinated schools.
The data for most schools (well over 90% in 2003) includes the full breakdown of offensive events for all pitchers. This allows us to use xRuns in the pitching context, by calculating xRuns allowed. I prefer this to component ERA, since component ERA formulas aren't generally very portable from context to context. Basically, xRuns allowed (where it's available) does a very good job of predicting runs allowed, but makes allowances for where pitchers have lowered ERAs or RAs due to being relievers, getting lucky, or getting good bullpen support. Or vice versa.
I think the xRuns calculation is slightly more accurate than runs allowed in analyzing a pitcher's true effectiveness, though of course it doesn't take into account clutch pitching ability and adds in a scorer's bias. Some pitchers are just very, very lucky or unlucky and wind up with runs allowed totals that vary widely from how they performed in individual contexts.
Just as I did for hitters, I calculated park adjustments for each pitcher, both to runs allowed and to xRuns allowed. I also calculated competition adjustments, as detailed in my hitters article. I used these figures to compare each pitcher to the NCAA average, both in terms of RSAA, and as RA+ - which is calculated like the familiar ERA+ figures – 100 is average, and a pitcher who allows half the number of runs of the average pitcher is a 200.
The Top 50 pitchers by **RSAA (park-and-competition-adjusted runs saved above average) are listed below. Before signing off, though, I'd like to give a short update on the project. Robert Dudek, analyst extraordinaire as he is, has been contributing some new park factor calculations which I am hoping to include in the final release of the full statistical breakdowns. I still intend to put up 2002 top 50 data, for both hitters and pitchers, in the very near future – within the next week or so. A lot of work needs to be done on cleaning up the database, removing double entries and so forth, as well as some stray team totals that got mixed in, and there are quite a large number of player names that have been somehow corrupted (Harvard's and Illinois' pitching staffs apparently have no first names, St. Louis's pitchers no last names, to name just three!) so I need to chase those down. Once that is done, a more analytical article should follow along with a full listing of every player for 2002 and 2003.
Top 50 pitchers by **RSAA, 2003
**xRSAA **RSAA IP **RA+
1 Jeff Niemann Rice 77.6 80.6 137.3 370
2 John Hudgins Stanford 79.7 77.7 165.3 226
3 Jered Weaver Long Beach State 79.6 68.4 133.3 268
4 Tom Mastny Furman unav 63.2 124.0 358
5 Wade Townsend Rice 65.1 61.8 118.7 284
6 Abe Alvarez Long Beach State 44.6 59.7 122.7 247
7 J. P Howell Texas 52.9 54.7 114.3 229
8 Jason Windsor Cal State Fullerto 49.1 54.7 95.3 334
9 Philip Humber Rice 54.8 52.5 128.0 204
10 Jeremy Sowers Vanderbilt 51.6 50.7 115.0 225
11 Huston Street Texas 50.3 49.7 74.3 471
12 Brad Sullivan Houston 57.0 48.8 123.7 190
13 Mike Pelfrey Wichita State 45.4 48.5 104.7 268
14 James Vermilyea New Mexico 48.6 47.8 119.3 216
15 Tim Stauffer Richmond 55.9 47.5 114.0 227
16 Aaron Marsden Nebraska 45.9 46.9 115.0 207
17 Ryan Wagner Houston 52.6 45.9 79.3 327
18 David Marchbanks South Carolina 29.9 45.9 135.3 175
19 Chad Mulholland Southwest Missouri 44.3 45.1 119.7 196
20 Clayton Jerome Texas Christian 44.7 44.7 111.7 213
21 Glen Perkins Minnesota 42.8 44.4 105.3 217
22 Randy Beam Florida Atlantic 50.1 44.2 101.7 239
23 Brandon Hankins Southwest Texas St 37.2 43.8 94.3 252
24 Greg Ramirez Pepperdine 37.9 43.7 122.0 196
25 Steven White Baylor 44.6 43.1 129.0 166
26 Josh Baker Rice 38.1 42.8 95.0 227
27 Justin Verlander Old Dominion 46.6 42.7 116.3 194
28 Matt Durkin San Jose State 42.6 42.6 100.3 235
29 R. J. Swindle Charleston Souther 51.3 41.3 119.3 189
30 Jessie Corn Jacksonville State 52.9 41.2 115.0 194
31 Mark Romanczuk Stanford 40.9 41.1 112.3 177
32 Jeremy Plexico Winthrop 41.1 40.5 117.7 194
33 Justin Meier Louisiana State 30.0 40.5 95.3 200
34 Danny Zell Houston 41.9 40.2 113.3 174
35 Brent Carter Alabama 51.9 39.7 122.7 165
36 Justin Orenduff Virginia Commonwea 36.0 39.5 95.0 234
37 Scott Lewis Ohio State 45.8 39.2 83.7 270
38 Dustin Miller Cal State Fullerto 34.1 38.9 100.0 190
39 Jaymie Torres UC Riverside 36.9 37.9 110.3 185
40 Chris Niesel Notre Dame 41.1 37.4 98.3 209
41 Paul Jacinto Brigham Young 43.0 37.4 113.0 184
42 Paul Maholm Mississippi State 28.4 37.3 107.7 178
43 Chad Cordero Cal State Fullerto 35.0 37.3 57.0 499
44 Aj Shappi UC Riverside 46.3 37.1 113.0 179
45 Tim Alvarez Southeast Missouri 35.2 36.2 108.3 193
46 Daniel Davidson Florida State 37.7 36.0 85.7 207
47 Trent Peterson Florida State 45.0 35.9 116.7 161
48 Cla Meredith Virginia Commonwea 28.2 35.4 60.3 520
49 Zac Cline West Virginia 33.4 35.2 125.7 160
50 Ben Thurmond Arizona State 26.1 34.6 79.0 235
