Tuesday, October 10, 2006

Jim Leyland may win manager of the year, but he isn't getting my vote. Despite being credited with the renaissance of a franchise long ago left for dead, he's made some basic mistakes that may take center stage in games 2-7 of the ALCS:

1) Leaving Chris Shelton off the playoff roster to keep three utility infielders
2) Refusing to trust the back of his bullpen to hold any lead

Now, with Sean Casey hurt and Shelton unavailable, the Tigers' contingency plan is to move Carlos Guillen to first and give the shortstop job to Omar Infante or Ramon Santiago. Essentially, they're using the first-base slot in their lineup for Infante or Santiago, both of whom hit, well, like backup middle infielders. Meanwhile, Neifi Perez, who took Shelton's postseason roster spot, will stay on the bench without anything to do.

Shelton was left off the roster, in essence, because he hasn't hit well since April. His stats for this year, while not as good as his breakout 2005, are still above-average at first base once you adjust for park factors.

Ah, but Shelton hasn't hit well lately. I've got news for you, Jim: Sean Casey has been terrible since donning a Tiger uniform. In fact, with Detroit, he's hit worse than post-April Shelton. So we can't chalk this up to being fooled by a small sample size. Clearly, Leyland is only willing to trust grizzled veterans with clutch situations.

There's a problem with that conclusion, however: it directly contradicts his philosophy of bullpen usage, which is that no lead is ever too big to bring in his 21-year old relief ace. Consider how he has "protected" his big leads in the playoffs thusfar:

ALDS Game 3: 6-0 lead through 7.2 innings. Brings in Joel Zumaya, who threw 2.1 innings yesterday, to get out of the eighth, then closer Todd Jones, who also worked the day before, to pitch the ninth. Never mind that this is the second of four consecutive days with games scheduled.

ALDS Game 4: 8-1 lead through 8.1 innings. Brings in Jamie Walker, his best left-handed reliever, to face the lefty-heavy Yankees. This one isn't really a big deal, because it was an elimination game and they had two days off afterward.

ALCS Game 1: 5-0 lead through 5 innings. Brings in his second-best reliever, Fernando Rodney, to pitch the sixth and seventh, Zumaya for the eighth, and Jones for the ninth.

To give you an idea of how this type of decision-making should be made, we'll use win probabilities from walkoffbalk.com. The Tigers' chances of winning entering the bottom of the eighth inning of Tuesday's game were 98.4%. By bringing Zumaya into the game, they perhaps increased their chances of winning to 99%, while they would have been lowered to 97.5% with Jason Grilli on the mound. So the decision added 1.5% to Detroit's chances of winning. (In reality, it didn't even do that, as Zumaya gave up a run.)

It seems like smart managing, until you realize that this inning restricted either Zumaya's availability or his effectiveness in future games, because there is a diminishing return on each additional inning you ask your relievers to pitch. If Zumaya tires later in the playoffs during a close game, it will largely be blamed on his youth and heavy workload during the season, but it may actually happen because Leyland won't save his arm for when it is really needed.

The whole idea behind optimal reliever usage is to save your best pitchers for when you really need them. In other words, you want them in there when one run will often change the result of the game. These situations occur only when the game is close, and more often later in the game. So far, in five games, only one of which was anything resembling close beyond the sixth inning, Leyland has used the back end of his bullpen to throw exactly two pitches, both by Jason Grilli in ALDS Game 1 to retire one Yankee batter.

If the game is 4-3 Tigers in the seventh Wednesday, will Zumaya have enough left in the tank to go for two innings? If he does, will he be good for another two innings Friday? Regardless, his effectiveness in those games will be affected by his workload up to this point. Leyland should take a lesson from Ken Macha, who used everyone but his two best relievers (Justin Duchscherer and Huston Street) to get out of Game 1. If you're never going to use your worst relievers, then take them off the roster and use that spot to keep Chris Shelton on the roster. Doing anything else is just a waste of roster spots, which are a valuable resource.

No comments: