Home Baseball Run Expectancy and Why Bunting Is Bad

Run Expectancy and Why Bunting Is Bad

is bunting smart

Run expectancy in baseball is simple, and incredibly important, changing how coaches think about strategy, stealing, bunting, and the value of outs and extra bases. Is bunting smart? Want to know why bunting is bad? Let’s start with a definition:

Run expectancy: how many runs we can expect to score, on average, given a specific base/out state.

Base/out state is also simple: it’s a situation, such as runners on first and third with one out. Every hitter comes to the plate in a base/out state; it could be 2 outs and bases empty, or no one out and the bases loaded. There are 24 possible base/out states.

And, for every base/out state, there is a mountain of MLB data that shows the probability of a run scoring in that situation, and how many runs score on average in any given situation (run expectancy). We’ll touch on run expectancy and bunting today, and cover probability another day.

How Has Run Expectancy Data Changed Baseball?


Well, to be precise, it only changes baseball when players and coaches are receptive to it. Understanding run expectancy charts helps a coach and player make good decisions based on the probability that runs will score, and how many we can typically expect.

The biggest takeaway from run expectancy is how incredibly important outs are. Every out causes a drastic reduction in expected runs, so we have to treat them like they’re precious. This is where sacrifice bunting comes in to play.

Why Bunting Is Bad – It Reduces Expected Runs.


In short, because outs are the biggest detractor from a team’s chances of scoring, sacrifice bunting hurts – a LOT. I’m not talking about bunting for hits.

Below are three charts that illustrate how expected runs decrease in the three most common bunting situations:

  • Bunting a single runner from 1st to 2nd (most often used late, when the winning or tying run is aboard)
  • Bunting a single runner from 2nd to third (most often used in the same situation as above)
  • Bunting two runners over from 1st & 2nd, to 2nd & 3rd (this is common when weaker hitters come to bat)

I jumped onto Tom Tango’s website, Tangotiger.net, and used his run expectancy data, which I inserted into the charts below, to make the data easier for you to read. The data is from 5 MLB seasons, 2010-2015.

I recommend Mr. Tango’s book, The Book: Playing the Percentages in Baseball, which can be found Here. I really enjoyed it. The information he posts on his website is free, so please support his research.

Note: Some of the links in my posts earn me an affiliate commission. This doesn’t affect the price you pay, but I thought you should know. I only link to products or books that I’ve used, love and recommend.

Sac Bunt, 1st to 2nd

In the situation below, I’ve highlighted the two base/out states we’re dealing with: runner on 1st with no outs, which then becomes a runner on 2nd with 1 out after the bunt.

Runner on 1st, No Outs: We expect 0.86 runs on average

Runner on 2nd, One Out: We now expect only 0.66 runs on average

RESULT: The bunt reduces the success of the average inning by 0.2 runs, which means that if you bunt 10 times in this situation, your team would score 2 fewer runs than if you didn’t. This might mean two fewer games tied up or won…

Sac Bunt, 2nd to 3rd

This is a common situation, trying to move the potential tying or go-ahead run to third with one out, so that a team can “manufacture” the run. A runner on third with one out will score with a deep fly ball or ground ball to the middle infield if the infield isn’t playing in.

Runner on 2nd, No Outs: We expect 1.10 runs on average

Runner on 3rd, One Out: We now expect only 0.95 runs on average

RESULT: The bunt reduces the success of the average inning by 0.15 runs, which means that if you bunt 10 times in this situation, your team would score 1.5 fewer runs than if you didn’t. This seems counterintuitive, since a sac fly or middle-infield grounder can score a player from third with one out. But, nonetheless, this sac bunt hurts the inning. This reduction is smaller than in the previous, but still relevant.

Sac Bunt, 1st & 2nd to 2nd & 3rd

This one is probably the biggest mistake of the three in youth baseball, because it reduces the team’s chances of having a BIG inning, and big innings often single-handedly win games. The reduction in run expectancy (RE) is only 0.06 runs, and so it basically wastes an out without improving the odds of scoring. 

BUT – remember, that this is MLB data below, and the effect of the double play, which will wipe out a inning faster than anything, is very real. 1st and 2nd has the double play in order, whereas 2nd and 3rd does not.

In youth baseball, especially at 16U and below, this difference is especially important, as double plays are vastly less frequent. In the Major Leagues, when the double play is in order, a double play occurs about 15% of the time.

At youth levels, this might be 5% or lower, which means that 1st and 2nd will produce even more runs than we expect in the chart to our left, since a grounder to shortstop will most likely move the situation to 1st and 3rd with one out (RE = 1.13), rather than 3rd only with 2 outs (RE = 0.35 runs).

Getting the first two runners on is the exact thing we want in producing a big inning, and giving the opposition an out hurts that goal, even more so in youth baseball than MLB baseball.

Run Expectancy Is Important!

It tells us why bunting may not be the right choice in most situations. There may be situations where a sacrifice bunt still makes sense, and bunting for hits is NOT included in this analysis of why bunting is bad. Bunting for a hit is completely different, as the goal is not to give the opposing team an out in exchange for a base.

Outs are the currency in baseball – making an out is the worst thing a hitter can do, and NOT making an out, of any kind, provides tremendous run-scoring value for a team, even just a walk, hit by pitch, or single.


  1. I would be interested to see the SAC bunting stats in a walk off situation. That is expectancy to score 1 run vs run expectancy.

    Do you know, or does the book cover those numbers?

    • Well, you wouldn’t really want to use expectancy for that; run probability would make more sense. A single runner on 3rd with 0 outs scores 84% of the time, with 1 out 66% of the time. So the question is, does a suicide or safety squeeze increase or decrease the likelihood of that runner scoring, compared to the 84 or 66%? When the hitters swing, there’s very little risk of the runner on 3rd being erased, whereas with suicide squeezes, and still with safety squeezes, a missed bunt often results in the runner at 3rd being out. So I’d guess the probability he scores on a bunt is lower than normal, but I dont know.

  2. I like the chart and I understand the data behind it. It does all into question the decision to move runners into scoring position. I watch a lot of major league baseball and I understand this data is for all teams combined. Where I get frustrated are teams that seem to not score any runs with bases loaded and no outs a high percentage of the time, not to mention the 2.29 average in the chart. As a matter of fact, just the simple fact that these are averages means that half the league performs below this average. I’m not a big stats geek, but the eye test for such teams says they come up empty the majority of the time with bases loaded and 0 or 1 out. If I only have to push 5 runs across the plate to win more than half of my games, it seems to me that getting a run or two every other inning through manufacturing accomplishes that. I know the home run and the extra base hit are where it’s at, but I’d rather see a box of 100200010, vs 000000300. Wouldn’t a team that is in the lower half of the league, as far as this chart is concerned, might benefit from moving more runners than a team that is built to hit 3 run bombs? Don’t you have to play as you are constituted and not according to charts like these? If you aren’t built to get a lot of dingers, doesn’t it seem counterproductive to wish for the homers that are not coming?

    • Hey Paul, thanks for the comment, and I get where you’re coming from.

      The stats, and I, both disagree with your eye test, which is really not a good way to judge these situations. Bases and loaded w 0 or 1 out…those are the best situations to be in as a hitting team, and the data reflects that. Probability of at least 1 run scoring with a runner on 3rd and 0 out is 85%…which debunks the eye test. And, from my 22 years playing the game, I in no way share your sentiment that teams come up empty most of the time…quite the opposite. It’s hard to prevent fewer than 2 runs from scoring with 0 out, and it’s still very tough to put up a 0 with bases loaded, 2nd and third with 0 or 1 outs.

      Plus, your very reasoning for manufacturing runs is contradictory to your thoughts on the bases loaded…manufacturing a run means trading outs to move runners to 2nd or third base. If you already have 2nd & 3rd, or the bases loaded, that situation is already accomplished for you.

      And, with averages, 2.29 runs is not a average built on 0 and 5 run outcomes – those outcomes are outliers (85% probability means 6/7 times at least one run scores). Rather, the 2.29 figure is built on 1-3 run outcomes in the bases loaded situation, which is consistent with my experience playing the game all these years.

      Not bunting does not mean a team is relying on homers. Rather, it’s just taking into account the BABIP – 3 out of 10 batted balls become hits, so with 1st and 2nd, for example, you get an over 50% chance of scoring those runners just by swinging away for three outs trying for one single or better. If you move them over with a bunt, a strikeout or pop up, grounder to P, 1 or 3 all ruin the manufactured run and force you to need a 2 out hit. In the Majors, only 15% of the time the double play is in order is a double play converted. So the odds are in a team’s favor to swing away – not for homers, but just for singles and doubles with the occasional homer as a bonus.

      • Thanks for the detailed response. I appreciate you taking the time to state your case and I enjoy these articles because it forces analytical thought. We all need to challenge our belief systems or we will fail to improve.

        I certainly agree that baseball has evolved with the focus on sabermetrics or advanced statistics. I definitely believe that the bunt was severely over utilized in the past. I simply believe that today’s game has perhaps tipped too far in the other direction. Often, when a bunt is needed, players are embarrassingly bad at execution. Even David Ortiz was very good at bunting–when he had to. Players need to work on all aspects of their craft.

        I still believe there is a place in the game for bunting other than against infield defensive shifts. Besides, if we always went with the numbers, no one would play the lottery 😉

        Thanks again.

  3. Dan, I think this info is extremely interesting. What would be really great is if there was an analysis of college level baseball, with the assumption that the talent level is lower and how that might effect outcomes (hitters not as good at situational hitting , infield defenses, etc.

    • Hey Desi, I completely agree – I read this stuff and constantly wonder how it applies, and how the numbers change in lower levels of baseball. In the Majors, outcomes are much more certain, as you know. But a sac bunt in 13U baseball probably becomes a hit 15% of the time (just guessing), so it’s very different. And, the disparate levels of play makes run expectancy tables much less relevant. Gamechanger just added BABIP to the stats they track, maybe they’ll start keeping base/out state data in the future.

  4. The MLB data likely does not accurately represent RE in youth baseball games. In fact, given that youth baseball scoring environment is generally higher than in MLB, I would expect that the true impact of sacrifice bunting is probably even more detrimental in youth baseball. More passed balls and wild pitches, as well as increased SB rates, for example, would tend to further temper the impact of sac bunts. In other words, a lot of those runners are going to advance, anyway, so why give up outs with sac bunts?

    • Yep, I agree – that’s what we’re faced with now, figuring out how these new stats apply to amateur and youth baseball, and how the implications change. And, you’re right that sac bunts are likely even more costly because of how often players steal, and how often a bunt can go for a hit, even when intended to be a sacrifice. Doesn’t make sense to just give up an out when they’ll give you a base anyway.

  5. Very interesting- every situation is different but knowledge is power. And an informed decision is a better decision. Statistics don’t lie- Thanks for the info/ learning.

    • Agreed – we can’t necessarily rule anything out, but statistics are doing a good job of explaining what maybe we couldn’t see before. Thanks for reading! Dan

  6. Love this stat and I’m not a fan of the bunt for advancing runners. How do we factor in the running making it aboard, the runner swiping and extra base? Do we assume this is countered by the likely hood of the batter not executing the bunt?

    I’ve seen similar results in the sport of Softball, but it’s more likely a runner makes it aboard with well executed bunt.

    My current view point is, let your hitters hit and fast runners steal.

  7. Also a follow up question. Whilst talks about run expectancy, this chart factors the likely hood of multiple runners scoring. I’d really like to see the expectancy of the lead runner scoring in these situations? If this does increase the likely hood of the lead runner scoring, which I expect it would, comparing the influence this additional run has on the result of the game over rolling the dice for the additional runs?

    • Run probability charts would be the stat to look to in that case, not expectancy. Probability just tracks the likelihood of at least one run scoring before the inning ends. I dont have a way of pasting one into this comment, but tangotiger.net has them (if I dont have one on this post).

  8. This is all wrong. Look at the chart. You forget one very important thing. The result of the at bat. Runner on first with no outs the run expectancy is.86. Like the charts says if the sac bunt is sucessful then runner on 2b with 1 out is a.66. Now if the manager doesn’t bunt and the batter gets out (so many varibles in regards to the htter at the plate but remember even the best hitters get out 70% of the time plus the risk of grounding into a double play) then you have runner at first with one out for a .51 run expectancy which is worse than runner on 2b with out. Of course if the batter hits into a DP (again varibles with the type of pitcher on the mound). Played and managed this games for 40 plus years. Run expectancy means nothing to me. If I have a player that can’t hit but can lay down a bunt and it is a close game with runners on base, I am bunting everytime.

    • Hi Bud,

      In the run expectancy charts, the numbers reflect all outcomes until the end of that inning in which that base/out state occurred. So, the 0.86 runs we expect with a runner on 1st with 0 outs includes all the times a runner will ground out, ground into a DP, etc. They took all instances in MLB where there was a runner on 1st with 0 outs, and with every outcome that occurred after, the number was still 0.86 runs.

      So, the same is true for a runner on 2nd with 1 out: 0.66 runs. This is the average of everything that comes after. Choosing 2nd with 1 out is choosing a situation in which less runs score over time. If you choose that situation 100 times – by bunting to it, for example – it will cost your team 20 runs on average.

      This is not to say bunting is never a good idea – I didn’t say that in my article. Context still matters, and bunting in the final inning to put the winning run on 2nd increases win expectancy, even if run expectancy is diminished. Context still matters.

      Sabermetrics and advanced statistics are challenging many of our long-held beliefs. It’s up to all of us to have an open mind to see how we can use the data to simply make better decisions. Thanks for the comment. Dan


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.