Friday, October 23, 2009

It's all about the leverage (or why Mariano should have pitched the 7th)

Leverage is a funny word. A loaded word, if you will.

For one thing, it's more or less responsible for the financial collapse of America.

In baseball parlance, leverage refers to an on-field situation--the inning, number of outs, score, pitch count, etc--and how likely the result of that on-field situation is to have an effect on the game.

Example: Bases loaded, no one out, top of the eleventh inning as it was in game 2 of the ALDS is a high leverage situation, because the result of the next pitch (or, if you prefer, at bat) is likely to have a significant impact on the game's final outcome.

Other example: Top of the ninth inning, two outs, runner on second, your team is up 10-1. This is a low leverage situation. At this point in the game, because of the score, whatever happens with the batter at the plate is unlikely to have a significant impact on the game's final outcome.


With this understanding of leverage, one can put forth a very simple concept:

Teams should use their very best relievers in the highest leveraged situations.

Simple, direct, easy to understand, right?

Here's where we have a problem: there is a pervading belief among many, especially those that read only main stream media (MSM), that the highest leverage situations are automatically the ninth innings in close games. In other words, situations that would warrant a "save".

This is not necessarily the case.

The ninth inning can become a high leverage situation, but many, if not most, games are decided before the ninth inning.

In the seventh and eighth innings, for example, you are likely to encounter a situation in which the starter has been removed from the game but where most (okay, right now, all) teams will consider it too early to go to their "closer"--who is often the team's best reliever.

Take tonight's game, for example.

In the top of the seventh inning, John Lackey was (wrongly) removed for Darren Oliver with the bases loaded and two outs.

With the score only a four run differential, this was an incredibly high leverage spot, especially with Mark Teixeira and Alex Rodriguez--two power hitters--due up. Now, one can argue whether or not Brian Fuentes should be a closer, but what is clear is that Oliver is not the Angels' best reliever. Jepsen, Fuentes and Bulger are all arguably or factually better than Oliver.

So the Angels and the Yankees were left with a situation where the Yankees were batting against the "soft underbelly" of the Angels' bullpen in a close game, and a team as good as the Yankees doesn't usually miss.

It's not much of a surprise, then, that at the end of that half inning, the Yankees lead 6-4.

At that point in time, Mark Teixeira's at bat shapes up to be the single most important at bat of the game. Never mind what followed afterwards, but at that very moment, it seemed likely that Teixeira's at bat could very well dictate the final outcome of the game.

Think of it this way: if you're the Angels' skip, are you sure you pitch Oliver there? Or would you be better off, if you're determined to take Lackey out, by bringing in Bulger or Fuentes?


So now we come to the Yankees and the bottom of the seventh inning.

Let's start here with a couple of knowns or supposeds:

A) Mariano Rivera is the best reliever the Yankees have on staff.

A1) Right now, this is likely followed by David Robertson, Phil Hughes, Joba Chamberlain, Phil Coke, Damaso Marte, Alfredo Aceves and Chad Gaudin, with Gaudin not being used except to mop up a blow out or to pitch extended work in extra innings.

B) Pitching the ninth inning of any game is not an easy thing to do, even if it's not necessarily the highest leveraged situation of a game. Just watch this interview after Phil Coke's first Major League save:

Thus, while one tries to address leverage, one cannot ignore a ninth inning where the leading team leads by no more than three or four runs as being a non-factor.

After tonight's loss, there was much conversation about the seventh inning and what should or should not have been done.

There is no arguing that the inning was the most critical part of the game--just look at the WPA graph:

(image credit: Fangraphs)

The bars on the bottom indicate the leverage of a situation--the ninth inning ended up turning into quite the critical situation, but it doesn't matter if not for the seventh.

As it was, a simple refresher:

With a two run lead, AJ Burnett went back to the mound with 80 pitches, and the first two Angel batters quickly reached base. At this point Girardi pulled Burnett for Damaso Marte, who retired two batters while allowing one run. Marte was then pulled for Phil Hughes, and, well, we know what happened from there.

Here's the leverage argument:

Because of the importance of the situation, with the tying runs on base and the Angels' best hitters (Hunter-Guerrerro-Morales) due up, Girardi should have gone to Mariano Rivera.

It's a claim that much of the MSM and their readers/viewers will brush off as being too reactionary, but it's based on the single, simple premise discussed above:

Teams should use their very best relievers in the highest leveraged situations.

At the time, there is utterly no way to predict that the ninth inning will matter or how much it will matter.

What you know, however, is that at the time, the two potential tying runs on base are the two most important runs you want to prevent from scoring if you are the Yankees.

A tie game will automatically favor the home team, since they always have the final at bat and one less half inning to pitch, so it is imperative to keep that from occurring.

Most (okay, all) teams will use these situations for their set-up men, and this is where those like Mike Francesa start oggling over the eighth inning guys. The fallacy here is the the eighth inning guy, almost by definition of the term, is not a team's best reliever, but the team's second best reliever, the one that teams most trust to get to their best reliever.

Has your head exploded yet?

At any rate, after tonight's game, I had a conversation with Ben Kabak of River Ave Blues.

I had previously asked via Twitter how many would have pitched Mariano in the 7th, and a large number of respondents said either that they would or that they would at least think about it.

I then brought up the issue that if you pitch Mariano for the seventh and maybe the eighth, and the Yankees don't score again, you are still left with a high-leverage ninth inning in which you cannot use Mo (the Hammer of G-d is, after all, 39) and have to choose from the remaining pitching staff.

Ben proposed this solution--and I swear your head IS going to explode--

You have Andy Pettitte pitch the ninth.

It's not abnormal for a starter to pitch in relief in the playoffs.

Starters are, by definition, better than relievers, and many teams, even playoff teams, lack a solid relief corps outside their closer and maybe one other guy.

We saw Jered Weaver pitch in relief tonight; in the past we've seen Mike Mussina do it in an ALCS game 7 against the Red Sox. There are other historical examples, as well.

You don't, as Ben suggests, pitch David Robertson in the ninth in a save situation, because there is an intangible aspect to a save that can't be ignored--see the Coke video above.

You go, instead, with the experienced hand.

In this scenario you don't even need to worry about Pettitte's start on Saturday--because you've effectively used Mariano, perhaps piggybacked with Robertson in the eighth, so you still have the lead in the ninth. Pettitte grabs three outs and the ALCS is over.

I am not the first person to suggest that a reliever's most important role is the ninth inning and I doubt I'll be the last.

Some day, a manager will come to the same conclusion and pitch their closer in the seventh to get through another team's 3-4-5. If that manager succeeds, he'll be branded a hero, visionary and genius, and if he fails, he'll be branded whatever his city's equivalent of "Clueless Joe".

The beauty of baseball is that it is a game that evolves, and that someone, somewhere, will try.