Are sac bunts “productive” outs?

A common conundrum encountered by enthusiasts of NPB is that of the sac bunt. It happens all the time, especially in the Central League where there’s no designated hitter (DH).

And because of the absence of a DH, the pitcher often has to hit. You can occasionally find guys that know how to swing a baseball bat, but in general the pitcher is an easy out. Therefore, most would agree that it’s better to move the runner over than to waste an out (an “unproductive out”, if you will) on a strikeout or a pop-fly.

And then there’s the most-annoying-case scenario: the pitcher hits into a double play, and the team is in worse shape than if he had simply struck out!

But the simple answer to the title of this article is: it depends on who’s doing the bunting.


Like most avid baseball fans, we here at Tsubamegun don’t get too upset about pitchers being told to sacbunt. However, we do get pretty riled up when a guy like Hiroyasu Tanaka is asked to.

Aoki’s on first with no outs? Great, sac bunt! (how many times have we seen that this season?)

The big problem, of course, is that Tanaka is not a pitcher.

Not that Tanaka can’t bunt, because he definitely can, but why would you want a hitter with a .343 on-base percentage (OBP) to play against the odds and ensure that the team is that much closer to not scoring a run?!

And fair enough, there are definitely late inning, tight game situations where the team desperately needs to manufacture a single run, and one method of doing that is to sacrifice bunt.

But it must be kept in mind that bunting in such a situation doesn’t actually increase the team’s chances of scoring a run. Every out brings the end of the inning/game closer, and therefore outs have been shown to be an extremely valuable commodity.

See, there are these things called Expected Runs Matrices that have been around for years, and they consistently demonstrate that having a man on first with no outs is statistically closer to scoring a run than having a man on second with one out. It’s that simple. Not that our manager, Takada, is paying attention, however.

And that is where the sac bunt is used the most often (no outs; man on first)–just like in the Aoki-Tanaka situation mentioned above.

I mentioned yesterday that it might be about time to revisit this topic, and in the unending spirit of producing advice and data that Takada neither cares about nor understands, here are the results of some number-crunching that I did this evening.

Method

Building on the idea that sacrifice bunts with no outs and a man on first rarely lead to a run being scored, and all of the moaning that I did in my sac bunting report from last year, I wanted to find a more simplified way to analyze the idiocy that several NPB managers adhere to. Our own Shigeru Takada ordered no less than eight sac bunts over the last two days, and only one of them was carried out by a pitcher!

A quick look at the circumstances surrounding those eight giveaway outs shows that no runs scored as a direct result of them. They were indeed wasted and ultimately unproductive outs.

However, two measly games does not provide nearly enough data to back up my central point: bunting with no outs and a man on first is nearly always a bad idea. So I decided to do some quick counting on all of the NPB games that have been played since the All-Star break last month. In all, I analyzed 34 baseball games of nine or more innings (which is still not enough to unequivocally support my thesis, but anyway…).

My goal was to see how many runs could be expected to cross home plate as a direct result of a sac bunt. All bunts mentioned in this survey were with no outs and a man on first base. In some situations there were multiple men on base, but this was the exception rather than the rule. The only times that the influence of a sac bunt was negated was when a triple or home run cleared the bases and cancelled the idea that the bunt was a “productive out”.

At the same time, I also gauged how often managers opted to let their hitters swing away with a man on first and no outs. A common reason for instituting the sac bunt is to avoid a double play, so I kept track of double plays as well. Furthermore, I tallied how many runs were scored in innings where sac bunts were not employed.

Because of the DH rule, bunts are a bit fewer and farther between in the Pacific League (23 to the Central League’s 34). However, there were still plenty of them to go around. It should also be noted that I did not attempt to control for fielding errors or other such anomalies that might artificially inflate the number of runs scored. Therefore, it is probably safe to say that this analysis is being quite generous in its consideration of how productive sac bunts have been since the end of July.

Results

There were 57 bunts in the 34 games that I sifted through. Each of them came with no outs and a man on first base, and a total of 14 runs scored following the bunt being laid down. This accounts for roughly 0.23 runs crossing home plate per bunt.

This compares unfavorably with the relative run value of the decision not to bunt in the same situation. There were a total of 156 instances where a man was on first with no outs, and the manager did not call for a sac bunt. I will call these situations, in acknowledgment of the common fear, double play opportunities (DPOs).

As can be expected, a chunk of those DPOs did actually end with the worst-case scenario. Double plays were turned on 20 occasions, accounting for 13% of all DPOs. While double plays are definitely a momentum shifter, it’s difficult to see how a 13% chance of something happening is scary enough to pass up on Tanaka’s 34% OBP. Miyamoto, another Tokyo player frequently called upon to bunt, has an OBP of 33%. In fact, there isn’t a single hitter (with at least 20 at-bats this season) on the the Tokyo Swallows first team that has an OBP lower than 13% (or .130 in baseball lingo).

I’m obviously simplifying things here, but it’s not that hard to see why we might want to limit sac bunts to players who can’t control the strike zone (and draw a walk) or swing a bat effectively.

Aside from the double plays that were actually turned, there were obviously many other outs recorded. However, 101 runs scored from those 156 DPOs. That works out to a simplified run expectancy of about 0.65 runs per decision not to give up an out for free. That is more than 2.5 times the run value attributed to a sacrifice bunt.

One of the limitations of this analysis is that it deals with a very small number of games. A more in-depth coverage of Expected Runs is in order within NPB so that we can see how closely the numbers match the available data from MLB.

However, no matter how small the sample size is, there seems to be little doubt that larger-scale studies will corroborate the available evidence.

Summary

Many a manager’s fallback strategy, the sacrifice bunt, can be attributed to any one of a number of influences. One of them is pure habit. The sac bunt has been an integral part of baseball for as long as these men and their fathers have been alive.

Another reason might be the desire to save face. Getting a guy on second base, no matter what the cost, makes it look like you created a chance to score. And that seems to count for a lot, especially in Japan.

However, in the beautiful words of Little Brother, some of us have already “…reach[ed] the conclusion that every now and then you gotta ask yourself: do you really want to win or just look good losing?”

And some managers have started to listen to the logical arguments (and maybe Little Brother as well) against the kneejerk application of such a wasteful strategy. More importantly, they have shown that it is possible to win while using it sparingly (or hardly at all).

Recently we’ve begun to see an ever-so-slight move away from its use in Japan. Managers like Trey Hillman (Hokkaido Fighters) and Bobby Valentine (Chiba Marines) only occasionally use(d) it. New Saitama Lions manager, Hisanobu Watanabe, last year’s Japan Series Champion manager, rarely bunts at all, and all three of those gentlemen have won a championship within the last four years.

Chunichi’s manager, Hiromitsu Ochiai, has also been known to bunt sparingly, and he won a championship two years ago.

A growing mountain of evidence is showing that their way of playing baseball leads to more runs being scored. More runs often mean more wins, and enough of those could lead to a berth in the playoffs. It’s very hard to deny the intuitive appeal of such a simple equation. Thus, it is shocking that a manager would do anything to decrease his team’s chances of scoring runs and winning baseball games. Sacrifice bunting, except when it’s me or someone else who doesn’t know how to hit at the plate, is simply betting against the odds.

The current survey shows that baseball managers would be wise to be more conservative in their use of the sacrifice bunt. Having competent hitters bunt doesn’t make sense. Betting against the odds will obviously work on occasion, but it doesn’t pan out very well over the length of an entire season.

About Christopher Pellegrini

Christopher is a budding sabermetrician and long-time supporter of Tokyo’s more lovable team, the Swallows. He has publicly volunteered, several times, that he plans to buy the team at some point in the future. When he finally runs the joint, it is likely that he will fine any player who swings at the first pitch or sac bunts (unless it’s a pitcher, of course). Follow him on Twitter: @chrispellegrini

  • Dan

    Ah, expected runs. I compulsively do the quickie ER calculations during games – one of the best sabremetric tools, for my money.

  • Ah the sac bunt.

    Watching NPB all these years has meant that I have learned to accept/tolerate them, but as you say, way more often than not it is an act of complete futility.

    Let’s not forget though that in Japan, it’s not just the manager who is 100% responsible. Often the fans expect that sacbunt to be laid down with that man on first and no outs.

    Indeed fans around us at Jingu often call for the sacbunt (or steal) and get angry with the manager if it is not carried out!

    So the practice seems to be deeply ingrained in all aspects of Japanese baseball culture.

    It does suck that’s for sure, but it may be many, many years before the Japanese managers and fans wake up and smell the ER coffee.

    Until then, I’ll just go back to my trying to ignore/shut out their abuse. And that’s a difficult thing to do with a man like Takada at the helm.

  • Christopher

    Nice article, unfortunately Takada isn’t the only one I’m afraid. I wrote about a similar situation with Tigers last year when either Hirano or Sekimoto were pretty much required to bunt automatically when Akahoshi got on base. This led to Sekimoto laying down 4 bunts in one game and he was a batter who ended up with 53 RBIs in the season! A bunt can be a useful tool to increase the pressure on a pitcher used sparingly but with a good stealer you should never bunt but rely on the steal instead. Purely unscientific of course but I have noticed that double plays tend to happen more with one out which rather defeats the purpose of a sacrifice bunt.

  • Sekimoto is leading the Central League in sacrifice bunts this season with 30 after losing to teammate, Hirano, last year (38 to 47, respectively).

    So far, Sekimoto’s ahead of Tanaka by three.

    Hopefully Tanaka won’t catch up to him.

  • The early 1 run lead to put the opposing team under pressure is valued here. Maybe the ER matrix is different for NPB than MLB? Hillman started laying down more bunts when the Fighters won it all, so who knows? 😛

    • Simon,
      I was hoping that you’d stop by to comment on this! It’s been a while!

      Do you have access to any numbers that would shed light on expected runs in NPB? Does anyone compile that data in Japan?

      CP

  • Christopher

    If it puts the opposition under pressure it’s a good idea but does it? Right at the beginning of the game? The opposition has 9 innings to go and the more savy teams will have figured out that the one run if its scored can be overturned. I would suggest that the one run is more useful to give your own team a boost and settle nerves. I also suspect that the opposition pitcher having given up a sacrifice bunt tries harder to get the next two outs and the bunt ends up being a motivator to the opposition.

    Sekimoto has just added another bunt tonight – Tanaka?

  • The idea of putting pressure on the pitcher is an interesting theory. We’ve all seen plenty of situations where a sacrifice bunt is followed by a walk (or two); one could probably argue that the bunt destabilized the pitcher and led to the free base(s). Whether or not it’s actually true, it’s a common way of thinking, that’s for sure.

    Let’s see what the stats say about this one.

    As Tokyo is currently leading NPB in sacrifice bunts, I decided to look through all of their games so far this season and try to figure out how often an early lead via sac bunt translated to a win.

    On eight occasions (most of which were in the first inning), Tokyo sac bunted with the score tied 0-0, at least one run scored thereafter, and the Swallows went on to win the game. Each time there was a runner on first with no outs.

    Interestingly, the birds have lost only one game this season in which they sac bunted early, got the lead, but ended up losing the game (April 5th, 3rd inning vs Yomiuri).

    At least part of the reason for this is that Tokyo’s relief pitching has been very steady this season, and they have rarely blown leads (knock on wood).

    However, as was demonstrated in the article, sac bunts often don’t lead to a run being scored, and that happened on at least nine other occasions (I am sure I missed several as I was focusing most of my attention on situations where a run scored).

    So as it stands, Tokyo has been quite successful this season, about 50-50, at generating runs if a successful sac bunt is laid in the first third of the game.

    There are, of course, some results that cloud the situation a little. Of those nine failed attempts to grab an early lead, Tokyo went on to win five of those games anyway (two of the bunts that failed to result in a run were in the same game versus the Fighters on May 22nd). We ended up losing the other four games.

    Tokyo’s opponents, on the other hand, were far less successful in terms of early sac bunt-aided runs translating into wins.

    There have been five games this season in which the opposing team got on the board first, but Tokyo came back to win four of those contests. The only team that held the lead was the Saitama Seibu Lions on June 3rd.

    In the end, it looks like the Swallows have been reasonably successful at scoring runs when utilizing the sac bunt early in the game. That observation contradicts my point in the article that bunting early in the game (unless it’s the pitcher) is foolish at least as far as the Tokyo Swallows are concerned.

    Whether or not there is indeed some kind of pressure that is felt by the pitcher is a phenomenon that should be looked into more.

  • Abe Paul

    Should be used if batter is poor, ie NL pitcher, or TWO men on base