Tuesday, April 15

Boston, Borowski, and bunts

That game SUCKED. No other way to put it. Borowski losing that game in the 9th was as disappointing as it was inevitable. Lobbing straight-arrow slowball at one of baseball's best lineups is no kind of way to win, and I don't think many observers expect Borowski to keep the closer's job much longer with that "stuff." In post-game talks, it's clear Borowski knew he wasn't feeling right - if he really knew that, he should have done the best thing for him and the team by telling Eric Wedge and the medical staff, so that the Indians could bring in Masa Kobayashi or Jensen Lewis. Manny Ramirez not being able to identify for reporters the Borowski offering he hit about 670 feet as a fastball or a changeup pretty much sums it up; comical, but if you're a Tribe fan, black comedy.

PD ace buffoon Bill Livingston said that last night's 9th-inning implosion "continues the Indians' nightmare." I guess that's true if you consider coming within one game of the World Series and then starting a 162-game season 5-8 a nightmare. Try being a Pirates fan sometime.

Anyway, enough about that garbage. I want to talk about bunts.

Let's discuss a scenario from the game last evening. The Indians opened their half of the 5th inning leading 2-1. Casey Blake drew a leadoff walk and Grady Sizemore followed suit, giving the Tribe runners on 1st and 2nd with nobody out. Asdrubal Cabrera then sacrifice-bunted both men over, and both scored on Travis Hafner's 2-run single. Not bad work.

The decision to bunt caused much consternation among the online crowd at Let's Go Tribe, as rabid an anti-bunt crew as you'll find. Those of you who follow baseball know that there's a sharp and growing division between old-fashioned types who favor things like bunts and base-stealing, and new-thinking people who eschew these things on the basis of statistical analysis.

I definitely put myself in with the latter crew - I think sabermetrics and statistical analysis are far better than "traditional" tactics and gut instincts. However, I think it's important to be flexible in one's thinking and not toe a party line either way.

Much of the thinking of the statistical-based people, especially regarding bunts, is based on a very handy device called a runs expectancy matrix. This chart breaks down, on average, how many runs a team should expect to score given the occupation of bases and the remaining outs in a frame. Click that link and have a look to get a feel for it. What it says, reading between the lines, is that outs are much more valuable than stubborn proponents of "smallball" realize.

For instance, most new-thinking people abhor the practice of giving up the first out of an inning to push a runner to second base (let's not get into pitchers bunting for now). The numbers bear this out: expected runs fall from 0.93 to 0.73, a drop of over 20%. The matrix doesn't speak to how the probability of getting just one run changes between these two scenarios, which can be relevant in endgames, but it shows that bunting a leadoff man over is, generally, poor strategy. The same is true for bumping a guy from 2nd to 3rd with your first out, though here I think the odds of getting just one run are almost certainly enhanced.

Getting back to the Indians game, and a situation where the club wanted max runs, not just one, the matrix is a useful tool. The Let's Go Tribe team decried Cabrera's bunt, some of them citing the expectancy matrix. Look at the two scenarios:

1st and 2nd, 0 out: 1.51 runs
2nd and 3rd, 1 out: 1.44 runs

That's 0.07 runs, a difference of less than 5%. About 1 in 15 times this might cost you one run, but generally you end up the same. However, keep in mind that these are average numbers, and are entirely situation-independent.

Consider the Tribe's situation here. Cabrera has looked totally lost at the plate recently. He seems every bit as good a candidate to hit into a DP as get a hit (exaggeration by me). He already has one successful (if ill-advised) sac bunt in this game against this same pitcher. The Tribe's two top hitters are on deck. There is a chance that the sacrifice is unsuccessful, resulting in a 1st and 2nd, one out scenario (0.91 runs expected), which would suck, but Cabrera is an excellent bunter and that chance is mitigated somewhat by the odds of him reaching safely.

Thus, in this case, I think the bunt was a very reasonable strategy. Swinging away would also have been acceptable (or taking a few pitches, considering that Red Sock pitcher Lester had walked the first two batters), but I think Tribe manager Eric Wedge made a good decision here, one that resulted in two runs. Might it have produced more to have Cabrera swing away? We cannot know.

The moral of the story is that it's always good to use statistical analysis to assist your judgments, especially if you don't want Fire Joe Morgan to rip into you, but when you do so, take a couple of precautions. First, make sure the numbers actually agree with your point (which, for the anti-bunters in this particular instance they did not). Second, make sure you are choosing the right numbers, ones that fit your game situation. Then you'll be the smartest baseball fan on the block.


Nick said...

Spot on in your analysis of bunting. I think too many people get caught up in this "all or nothing" idealism; you're either a smallball guy or a stats/gorilla ball guy, when in reality you need to consider the context of the situation. Things like the hitter, the pitcher, the next few hitters, and how all of those players are doing (hot/cold) at that particular time need to be accounted for.

I agreed with AsCab's second bunt, but not his first. It should also be noted that the sacrifice worked out, the Tribe scored a pair of runs.

Andy said...

I did point that out!