from Amazon
Sunday, April 27, 2008
Rating the pitching coaches (sorta, maybe)
Here is what I did. I compared my pitcher projections, which are pretty good in the aggregate, with actual pitching performance, for each year, 2003-2007. I did this only for pitchers who changed teams from the year before to the year in the study, but stayed in the same league. IOW, for all pitchers in 2003, only those who switched teams from 02 to 03 were included. For 2007, only pitchers who switched teams from 06 to 07 were included. Etc.
The metric I used for my projections and for actual performance was my context neutral normalized component ERA (NERC), where 4.00 is defined as league average.
The presumption is that if the actual NERC for all the pitchers who came to the Braves is less than their collective projected NERC, then the Braves pitching coach (or something else about the Braves team or the move itself) did something to improve these pitchers’ performance. I realize that this is a shaky assumption, but the numbers are quite interesting.
The average NERC for all pitchers who switched teams was 4.29 after the switch (4.32 in the NL and 4.22 in the AL). These are likely older pitchers or those who had a bad prior year (in general, players who switch teams tend to be both older, a little bad, and a little unlucky). Their average projection was 4.16 (4.22 NL and 4.06 AL). The reason their projection was lower than their performance was probably because they are a little old. In fact, the average age of these pitchers was 31.4 (31.5 and 31.3).
The numbers you see in the first column represent the difference between projected and actual plus the .13 runs we expect for every team. A plus number means the move helped the pitcher (the NEW pitching coach was GOOD) and a negative number means the move hurt the pitcher (the NEW pitching coach was BAD). The second column represents the gain or loss from the OLD team. If all the pitchers that LEFT a team lost some NERC when going to their new team, the new team gets credit, and the old team gets docked. Comparing the 2 columns is a little bit (maybe a lot) of a check on the validity of the assumption. In other words, if pitchers going to a team are helped and when they leave the same team, they get worse, maybe there is something to that team’s pitching coach or the team itself. Before you look at the list, think about which pitching coaches/teams you think are good with their pitchers and which you think are bad.
Also, I used the entire 5 year period for all teams. Obviously pitching coaches on some teams have turned over during that time period. With no further ado:
Team/ NERC Before/ NERC After/ Diff + “league adjustment"/ Diff after leaving (Plus favors the old team)
ATL 4.23 3.84 .49 .25
ATL (w/ Mazzone) 4.22 3.79 .53 .44
ATL (w/o Mazzone) 4.28 4.08 .30 -.97
PIT 4.32 4.05 .37 -.06
NYN 4.21 4.05 .25 .04
MIL 4.20 4.13 .17 .04
MON/WAS 4.45 4.40 .16 -.09
COL 4.36 4.40 .06 -.32
HOU 4.29 4.35 .05 .12
SFN 4.45 4.56 -.01 .05
CHN 4.15 4.30 -.05 -.28
PHI 3.74 3.95 -.11 .32
FLO 4.04 4.25 -.12 -.21
SLN 4.21 4.47 -.17 -.14
ARI 4.27 4.61 -.24 .19
LAN 4.13 4.49 -.26 .80
SDN 4.10 4.49 -.29 .07
CIN 4.21 4.95 -.64 -.15
BAL 4.40 3.71 .86 -.05
BAL (w Mazzone) 4.58 3.71 1.03 -1.19
BAL (w/o Mazzone) 3.92 3.68 .39 .45
OAK 3.99 3.81 .34 .28
TBA 4.57 4.49 .24 1.21
DET 4.18 4.22 .13 .21
CHA 3.88 3.94 .10 .00
ANA 3.69 3.80 .05 .29
TEX 4.07 4.19 .04 .08
SEA 4.37 4.59 -.06 -.47
KCA 4.38 4.60 -.06 .10
NYA 3.53 3.78 -.09 .25
MIN 4.42 4.69 -.10 -.03
BOS 3.79 4.12 -.16 -.80
TOR 3.92 4.63 -.55 -.52
CLE 3.93 4.86 -.78 -.11
If you can figure out the charts, in the NL, ATL is by far the best overall (coming and going). Then Pittsburgh and the Mets. The worst in the NL is CIN by a landslide. Then FLO and SLN.
In the AL, BAL (mostly w/o Mazzone), OAK, and TBA are by far and away the best. BOS, CLE, and TOR, are far and away the worst.