Dissecting Manuel’s Decisions to Bunt

Perhaps it’s the fact that the Phillies dropped two consecutive games to one of the most beatable teams in the MLB. Perhaps I’m just overreacting to the first few games of the season, which always seem ever-important after five months without meaningful baseball. Maybe it’s the fact that hindsight, of course, is 20/20.

But I really, really hate bunting.

There’s something inherently frustrating about voluntarily giving away outs. After all, assuming a 9-inning game, a team only has 27 outs to go around, and willingly discarding them is usually a senseless maneuver. Basic statistical analysis tells us that, on average, bunting will result in a loss of expected runs. That is, if used less than optimally, the bunt will almost always hurt a team in the long run.

In the first three games of this season, the Phils have shown that they’re going to carry out their “small ball” plan of attack that was described by Charlie Manuel in spring training. Pressured by the fact that much of the lineup has been devastated by injuries, Manuel went on record saying that the team would be working on things like stealing bases and bunting—the kind of asinine strategy that you’ll often hear TV announcers like to call “manufacturing runs”. This type of approach is why you saw Juan Pierre, the man who OBP’d .329 last season, lead off for the Phillies today.

What I’d like to do is break down each of Charlie’s decisions to bunt thus far. Is my frustration justified, or have these bunts had no real effect on the outcome of this first series? Let’s take a closer look at each individual situation. Keep in mind that run expectancy is not perfect; it fails to properly take context  into account. Yet we can do our best to subjectively gauge that effect as we go along—bunting Victorino is much worse than, say, sending Brian Schneider out to sacrifice.

Saturday:

1) Top of the 1st, no outs, men on 1st and 2nd. Rollins up to bat, Karstens on the mound.

It’s not everyday in baseball that you find your 3-hole hitter laying down a bunt, yet that’s exactly what occurred. Instead of letting Jimmy hack away, Manuel made the decision to move the baserunners over at the expense of an out. Run expectancy tells us this:

Expected runs to end of inning, prior to bunt: 1.43
Expected runs to end of innings, after bunt: 1.29
Actual runs scored in inning: 1

Whereas there is a clear loss in expected runs, the difference here is admittedly marginal. Pence ended up singling to center, and while we can’t know what Rollins would have done if he hadn’t bunted, I don’t find that bunting Rollins with no outs in the top of the first is a defensible move.

2) Top of the 9th, no outs, man on 1st. Nix up to bat, Hanrahan on the mound.

Prior to this attempt, Laynce Nix had only laid down sac bunts four other times in his 9-year career, the last of which occurred in 2010. Given that information, one would be inclined to think that it wouldn’t go so well, and it didn’t. After fouling off the first attempted bunt, Nix popped up the second behind home plate, briskly making that first out. By the numbers:

Expected runs to end of inning, prior to bunt: 0.85
Expected runs to end of innings, after bunt: 0.65 (given that the bunt would have actually worked)
Actual runs scored in inning: 0

One caveat: expected runs isn’t exactly a perfect metric in this scenario. In a tie ballgame in the later innings, the focus shifts from scoring as many runs as possible to simply scoring a run. With that in mind, this move theoretically becomes less frustrating. Yet the fact of the matter remains that Laynce Nix is probably not very good at laying down bunts, and he managed to mess it up.

Sunday: 

1) Top of the 6th, no outs, man on 2nd. Victorino up to bat, McDonald on the mound.

Over the past few years, we’ve seen that Manuel has a tendency to bunt Victorino for no discernible reason. Doing so makes even less sense these days considering that Victorino has become the team’s most productive hitter, Hunter Pence aside. Regardless, Victorino opted to sacrifice an out in order to give Rollins a shot to knock Pierre in from third base.

Expected runs to end of inning, prior to bunt: 1.06
Expected runs to end of innings, after bunt: 0.9
Actual runs scored in inning: 0

Well, Rollins ended up striking out, and the inning ultimately came to an end with a 0 on the board. Again, we can’t possibly know what would have happened had Victorino decided not to bunt, and it’s entirely possible (and even likely) that he would have made an out anyway. But failing to let a great hitter swing way in a crucial spot is simply an inexcusable mistake.

2) Top of the 7th, no outs, man on first. Galvis up to bat, Hughes on the mound. 

Even though the numbers would tell you differently, I believe this is a largely defensible move. Galvis has still failed to record his first major league hit and he has looked terribly uncomfortable at the plate. This is one of those situation where context truly trumps the overly-theoretical nature of run expectancy, and it’s our responsibility to make a subjective adjustment. Regardless, let’s have a look.

Expected runs to end of inning, prior to bunt: 0.85
Expected runs to end of innings, after bunt: 0.65
Actual runs scored in inning: 2

It’s important to note that the 2 runs were the result of an error—Galvis ended up reaching base due to a bad throw. Indirectly, bunting Galvis turned out to be a pretty lucrative choice. Errors and misplays aside, this is one of those times when a sacrifice bunt isn’t such a bad thing.

Though I find myself in stark disagreement with three of the four decisions to sacrifice bunt thus far, these are admittedly minor moves. Had the Phillies won today, I probably wouldn’t be writing this, and subsequently I wouldn’t be upset over a game or two that won’t make a glaring difference in the long run. However, the long run should absolutely be a concern. These seemingly inconsequential managerial moves can add up over time, and if the Phillies continue to waste outs at this rate, this sort of stuff will actually start to matter. Lost runs here or there pile up into lost wins. Come September, the Phils will probably see themselves in a position where lost wins become the difference between October baseball or going home. Here’s to hoping that Manuel lets the runs manufacture themselves.