Today's pinch hit is courtesy of reader Mike Green, who has some thoughts and analysis on reliever usage. Thanks Mike!
The cost of reliever roulette
by Mike Green
Managers are often criticized, including here at Batter's Box, for serial relief changes, usually to gain the platoon advantage. This raises the question- "what precisely is lost by this strategy?"
I did a small study to try to calculate at least some of the loss. One of the consequences of the strategy is that relievers are more often required to pitch on consecutive days. I calculated the Blue Jays relief staff's performance on 0 and 1-2 days rest in 2003. Here are the results
It is a small sample, but the most notable finding was that the home run rate almost doubled in relievers pitching on consecutive days. The walk rate was up significantly as well. Curiously, relievers going three days in a row did not fare poorly on the third day. In 24 innings, the relievers gave up 25 hits, 2 HR, 12 walks, 22 Ks.
It would be interesting if the results were this extreme over a larger sample, and one that controlled for the quality of the opposition.
It should be noted that Kershner, over a very small sample, possessed the same traits as the group, but to an even greater degree. Finally, here are Valerio de los Santos' splits on 0 and 1-2 days rest:
0 days
17IP, 16H, 2HR, 12W, 13K
1-2 days
54IP, 42H, 5HR 19W, 39K
This limited data would be enough to persuade me that ordinarily only one of the lefties should be used in a game, because the cost of bringing in both in terms of reduced performance the following game outweighs the benefit of the platoon advantage in that game. The exception might be if there is no game the following day.
The cost of reliever roulette
by Mike Green
Managers are often criticized, including here at Batter's Box, for serial relief changes, usually to gain the platoon advantage. This raises the question- "what precisely is lost by this strategy?"
I did a small study to try to calculate at least some of the loss. One of the consequences of the strategy is that relievers are more often required to pitch on consecutive days. I calculated the Blue Jays relief staff's performance on 0 and 1-2 days rest in 2003. Here are the results
0 days rest
IP H $H HR $HR W $W K $K
109 107 .982 21 .198 51 .468 81 .743
1-2 days rest
226.2 229 1.01 24 .106 89 .393 164 .723
It is a small sample, but the most notable finding was that the home run rate almost doubled in relievers pitching on consecutive days. The walk rate was up significantly as well. Curiously, relievers going three days in a row did not fare poorly on the third day. In 24 innings, the relievers gave up 25 hits, 2 HR, 12 walks, 22 Ks.
It would be interesting if the results were this extreme over a larger sample, and one that controlled for the quality of the opposition.
It should be noted that Kershner, over a very small sample, possessed the same traits as the group, but to an even greater degree. Finally, here are Valerio de los Santos' splits on 0 and 1-2 days rest:
0 days
17IP, 16H, 2HR, 12W, 13K
1-2 days
54IP, 42H, 5HR 19W, 39K
This limited data would be enough to persuade me that ordinarily only one of the lefties should be used in a game, because the cost of bringing in both in terms of reduced performance the following game outweighs the benefit of the platoon advantage in that game. The exception might be if there is no game the following day.
