Monday, January 08, 2007

Pitching Change Platoon Advantages

As some readers may be aware the Cubs had a really really bad season on their way to losing 96 games. They gave up 834 runs, good for second worst in the NL in no small part as a result of 60 starts being given to pitchers who had never seen big league time before. By comparison Houston gave 44 starts to newbies, Florida 40, and Tampa Bay 34 while Cincinnati had none at all.

While that may bode well for the future as the Cubs will now have some experience at the AAA level (Sean Marshall pictured on the left and Carlos Marmol particularly) they can draw from, it resulted in 2006 in Dusty Baker making 542 pitching changes, easily outpacing the previous record set by the Giants in 2004.

All of those pitching changes got me to wondering how frequently a manager tries to maintain the platoon advantage when making a pitching change. In other words, while The Bill James Handbook publishes the numbers for how often a manager maintains the platoon advantage when making out his lineups, I've never seen the numbers for pitching changes. I wrote and ran a simple script to count each pitching change and determine when the defense had the platoon advantage. The results are shown in the table below and as you can see the percentage varies from the mid 50s to the low 70s. Obviously, these numbers are heavily influenced by roster construction and the effectiveness of the pitchers the manager has to work with at any given time. It's not surprising to me to see the Cardinals near the top although I would have expected the Rockies to be up there as well as Clint Hurdle seems to like using LOOGYs even when they are manifestly ineffective (Ray King).





Team     Changes PlatoonAdv    Pct
SEA          429        309   72.0%
CHA          398        282   70.9%
DET          390        261   66.9%
CIN          475        317   66.7%
SLN          468        312   66.7%
CHN          542        360   66.4%
NYA          488        314   64.3%
TEX          489        313   64.0%
KCA          473        302   63.8%
BAL          472        300   63.6%
OAK          444        282   63.5%
MIL          427        271   63.5%
PIT          504        319   63.3%
MIN          421        265   62.9%
HOU          497        312   62.8%
SFN          438        271   61.9%
ANA          380        235   61.8%
CLE          376        232   61.7%
ATL          522        321   61.5%
TOR          481        292   60.7%
COL          498        302   60.6%
WAS          515        306   59.4%
SDN          475        276   58.1%
BOS          454        262   57.7%
TBA          444        256   57.7%
FLO          435        249   57.2%
PHI          500        286   57.2%
ARI          461        256   55.5%
LAN          454        252   55.5%
NYN          473        252   53.3%
at

1 comment:

  1. Tangotiger2:50 PM

    Great stuff Dan. I always like straight-forward ideas but that takes alot of work to compile, and gives you a great insight.

    In a similar vein, Andy looked at when PH are brought in:
    http://www.baseballprospectus.com/article.php?articleid=5404

    From Andy's sample, 78% of PH had the platoon advantage. Your list for pitchers is 62%. Of course, when a closer is brought in, the manager is not looking at the platoon advantage. As well, in blowouts, the manager is not looking for platoon advantages.

    Can you break up your list based on whether the pitcher is the "ace" or not, and whether the score is within 4 runs or not?

    ReplyDelete