Sunday, April 11, 2010

Using Relievers Part Duex: Torre Blows It

Several things define the Giants fan.  She enjoys AT&T park, she goes to the stadium whether or not the team is losing or winning, she laments the Giants' continual failure to capture a World Series since moving to SF, and she takes great schadenfreude in seeing the hated Dodgers blunder their way through a season.  He also makes liberal use of pronouns to avoid sexism; afterall, this is San Francisco.

My last post discussed some options and strategies for maximizing the use of a team's bullpen, in particular when to use your elite reliever.  Recently, the Dodgers lost a series to the Florida Marlins, most due to a woefully inefficient use of their bullpen.  Considering I enjoy seeing the Dodgers lose  and that they lost in a way that is directly linked to a post I recently made, I would like to discuss what the Dodgers did, why it was wrong, and what they should have done in order to avoid an embarrassing series-loss to the Florida Marlins.

The Facts

The Dodgers played a weekend series against the Florida Marlins on 9 April, 10 April, and 11 April.  The Dodgers won the first game of the series, coming off a strong performance by LA starter Hideki Kuroda.  The Dodgers would start the next two games off well, only to come up short due to late inning heroics on the part of the Marlins, coupled with late-inning blunders by Dodger manager Joe Torre.

In game 1, the Dodgers entered the ninth inning with a 7-1 lead.  Torre, wisely, initially brought in Russ Ortiz, a weak reliever, to close out the game for LA.  As I discussed in a previous post, there is little statistical difference in having a weak or strong closer end a game where the closing team has a three run lead; the closing team wins at the same rate whether the close is elite, average, or weak.  Surely, such a difference is even more insignificant when a team has a 6-run lead.  However, Ortiz began the inning by giving up 2 runs, closing the Dodgers' lead to 4 runs, which is still a huge lead in the last inning of a ballgame.  Torre apparently choked and made a knee-jerk reaction to save the game by bringing in Jonathan Broxton.  Broxton is a wonderful pitcher, and he may one day be the best closer in the league if his weight does not end his career prematurely.  Broxton is capable of bringing the heat: his fastballs regularly breach 100 mph.  Needless to say, Broxton came in, fanned two batters, and the Dodgers took the day.

Game 2 is a different story.  The Dodgers entered the bottom of the ninth with a 6-4 lead.  Instead of bringing in Broxton, who Torre used the day before to close out a blow-out game, Torre brought in George Sherrill, who, with a 1.36 career WHIP, is a fairly average reliever.  With Sherrill on the mound, the Marlins made a three-run rally and won the game.  Broxton, meanwhile, sat on the bench watching as the game got away from LA.

Game 3 also got away from the Dodgers.  Their starter put on a great performance, striking out 12 Marlins, however the Dodger bullpen collapsed, and gave up enough runs for the Marlins to make a late inning return to take the game and the series.  

Ineffective Assistance of Bullpen 

As I also noted in an earlier post, while the difference is not huge, there is a difference between using a weak and strong reliever when there is a 2-run lead.  Remember: with a 2-run lead an elite closer will lose the game 4.9% of the time, while an average or weak reliever will lose 9.0% of the time (Tango 215).  However, when there is a 3-run lead there is no statistical difference between using a strong or weak reliever.  An ace-reliever will win 97.%5 of the time with a 3-run lead, while a bad reliever will win 95.5% (Tango 213).

Thus, Torre's initial decision to use Russ Ortiz to close out game one, where the Dodgers have a huge 6-run lead, was the correct decision.  However, after Ortiz game up two runs, Torre freaked-out and put in the ace, Broxton, despite the fact that the Dodgers still had a 4-run lead, and had 1 out.  It was absurd to put in Broxton.  If Torre had no confidence in Ortiz's ability to end the game, then he should have taken him out and put in another reliever, but not Broxton.  As it was, it was a waste to use Broxton when he did.

This would lead Torre to not use Broxton the following day, when the situation was appropriate to use an ace-reliever.  Instead of using Broxton, whom Torre must have believed needed rest, Torre put the game in George Sherrill's hands.  Sherrill, being an average reliever, had twice as large a chance of blowing the game than did an elite closer like Broxton.  And, no surprise, Sherrill blew the game.

Thus, Joe Torre made a huge blunder by wasting Broxton's service on a blow-out game, and then not using him the following day when it woudl have been appropriate.  Sherrill should have closed out the blow-out game, while Broxton should have closed out the close game.  All in all, very pathetic decisions from the Dodgers' skipper.

How Often Can a Reliever Pitch?

Torre probably did not use Broxton because there is a belief in present-day baseball that relieving pitchers should not pitch back-to-back.  In his book, Tango addresses this issue and finds that relieving pitchers who have a heavy workload, show no difference in effectiveness from pitchers who have a comparatively light workload (233-6).  In fact, Tango argues that relievers should be used up to 40% more than they already are (236).

Broxton had pitched about 1.2 innings before game 2, when Torre decided not to use the closer, apparently because he was overworked.  This is simply wrong, and is refuted by statistical analysis of pitchers over a many year period, which shows that relievers can handle a much larger workload than current MLB-practice would have.

Simply put, another blunder from ol' Joe.


This past series highlights the reasons why managers must know how to effectively use their bullpen.  Had Torre used Broxton in game 2, and used Sherrill in game 1, LA probably would have taken the series.  Instead, the Dodgers get to fly home knowing they are in last place after a week deep into the 2010 season.

Works Cited

Tango, Lichtman, and Andrew E. Dolphin.  The Book: Playing the Percentages in Baseball. Washington D.C.: Potomac Books, 2007.


  1. Ortiz didn't give up two runs. He did load the bases, but it was Broxton who gave up the 2RBI double.

    And, since when does using your closer one day ruin him for the next? Closers can pitch back to back days - that's their job.

    Analysis of box score statistics is not good analysis.

  2. I agree, you should use your closers back to back. In fact, I made that argument. Re-read what I wrote, particularly the last section. I noted there is a belief in baseball today that you cannot use your closers too often, however you certainly can. For some reason, I don't know why, Torre didn't use Broxton the next day.

    Thank you for correcting the part about Ortiz not being the one to give up the RBI double. That still doesn't change anything: Broxton should not have been in that game.

    Thank you for reading.

  3. This comment has been removed by the author.