Where and why bunting sometimes is a decent idea, mathematically

[dawgs]

everyone knows bunting can be a good move when playing for only 1 run - almost exclusively in the 9th inning or extra innings of a tied game or down 1 run with a man on base and 0 outs. i wouldn't bunt in the 9th and needing 1 run if there was 1 out though. our problem is we bunt a lot of times when we we shouldn't just be going for 1 run.

[Irondawg]

As always there is a time and place for everything and you have to sometime go with your gut. The problem is that Cohen is going against statistically odds at an alarming rate - much more than other teams i've watched and while it does work sometimes, i don't think we are beating the predicted odds overall.

And keep in mind that bunting in an effort to get a hit is different than sac bunting.

[PMDawg]

I dont have a problem with bunting, per se, but the fact is we suck at it and we over use it. The first inning bunt was extremely idiotic. Doing it when playing with a 3+ run lead (unless very late in the game) is idiotic. Doing it with hot hitters and middle of the order guys is stupid. We also record outs without moving runners on failed bunt attempts at an alarming rate. Our failure rate is much too high to justify doing it more than once or twice a game.

[PMDawg]

Discount double post.

Deleted

[Smitty]

We currently lead the conference in sac bunts with 20.

Texas A&M has 5
Florida has 5
South Carolina has 6
LSU has 9

We are being left behind strategically. The worst teams are bunting, the best teams aren't.

Sac bunts
---------
1. Mississippi State... 20
2. Tennessee........... 19
3. Auburn.............. 14
4. Georgia............. 12
5. Ole Miss............ 11
6. Vanderbilt.......... 10
7. Kentucky............ 9
LSU......……......... 9
9. Alabama............. 8
10.South Carolina...... 6
11.Missouri............ 5
Florida......………... 5
Texas A&M........... 5
14.Arkansas............ 3

[PMDawg]

The saddest part is we probably lead in failed attempts by double digits.

[Smitty]

The saddest part is we probably lead in failed attempts by double digits.

Likely.

It's unbelieveable we have over twice as many sac bunts as all but 4 league teams... And those 4 suck.

[blacklistedbully]

You just ruined your whole point. If you're down one late, or even when tied, you're not trying to score exactly one run, you're trying to score at least one run. Put another way, you're just trying to not score 0 runs.

You definitely want to score 1 run, but obviously scoring 2 runs is also great in this scenario (especially if you're down 1), so scoring exactly 1 run isn't the goal; it's making sure you score at least one run.

The reason your odds of scoring both 0 and 1 runs both go up is because your odds of scoring more than 1 come down by so much. So some of the scenarios by which you would have scores more than one run by not bunting are now reduced to only 1 run, thus increasing the odds of scoring just the one run while not actually increasing the odds of scoring any runs.

How are the odds of scoring at least one run measured? This is the same as not scoring zero runs. So if your odds of scoring 0 runs increase, that means your odds of scoring at least 1 run decrease. So in your own post, your stats show it is still a bad idea.



If I didn't explain myself well enough:
If the percentages are as follows for not bunting (I know these aren't the actual %s but it will explain my point):
0 runs - 30%
1 run - 30%
2 runs - 30%
3+ runs - 10%
then you have a 70% chance of scoring at least one run.

If the percentages are as follows for bunting:
0 runs - 40%
1 run - 40%
2 runs - 15%
3+ runs - 5%
then you have increased the odds of scoring exactly one run, but you'd also increased the odds of not scoring any. Since not scoring any is far worse than scoring more than 1 obviously, who cares what the percentage is of scoring exactly one run? You just decreased your odds of tying or winning the game from 70% to 60%.

You are wrong. If one looks at the stats from a mathematical equation, the odds of scoring one run in a situation where you have a man on first with no outs are at their highest when you bunt the man to second. The price you pay for that is a reduction in the chance to score more than one run. This is the problem inherent with those who use a "total runs scored" approach when analyzing the situation. The numbers get skewed because they are impacted by the innings where multiple runs are scored.

Nothing in my posts, "ruined my point". You just don't seem to have grasped the point. Broken down, the math suggests that, if you are playing for exactly 1 run in an inning that starts with a man on first and no outs, the very best chance statistically is to bunt that guy to 2nd. Again, even though it reduces your chance of scoring more than 1 run, it actually does increase the chance you will score precisely 1 run. That suggests bunting is a plausible strategy in situations where that 1 run is desperately needed at that point in the game.

Think of it as laying up on the golf course. Sure, you can go for it on your drive, increasing your chance of a lower score. But if you're in a spot where you really just need PAR, it sometimes makes sense to play it safely into the fairway, even if you are conceding a possible birdie. Nobody is suggesting you can win a tournament by playing safe the entire round, but there are times when it just makes sense. If you do it on every hole, you are rarely, if ever going to win, but that doesn't mean you don't do it situationally. Going for it greatly increases your chances of going sub-par on any given hole, but it comes at the cost of reducing the chance of getting PAR. Players who win usually go for it, but they also know when it makes sense to just take the PAR.

It's a situational thing that can't be looked at in the vacuum of overall stats and "run expectancy" over the course of an entire game. It can't really be looked only in the vacuum of, "overall run expectancy in an inning", as that view completely discounts the strategy of, "playing for 1 run". The, "1-run" issue is the big factor here.

Sure, if you want to maximize the runs you produce over the course of a game, automatically bunting every time you have a man on 1B with no outs is not a good way to go statistically, because if you swing away in that situation most of the game, you'll more often score more than 1 run. But if your objective is to do everything you can to try and get 1, then bunting makes sense.

And a big flaw in your statistics is that the percentages of scoring, "zero runs" only goes up about 2%, while the chances of scoring exactly one go up about 6%. Here are the actual percentages for man on 1B, no outs:

Swing away:
0 runs - 57.6%
1 run - 18.94
2 runs - 12.29
3 runs - 6.09
4+ runs - 5.08

Bunting:
0 runs - 59.39%
1 run - 24.63
2 runs - 9.28
3 runs - 4.05
4+ - 2.65

[smootness]

My point is that you're not playing for exactly one run. You're playing for at least one run, so looking only at the chances of scoring exactly 1 run doesn't help you in evaluating it. You have to take your combined chances of scoring 1 run, 2 runs, 3 runs, etc. In other words, your chances of scoring more than 0 runs.

Since you admitted that bunting the guy to 2nd actually increases your chances of scoring 0 runs, you are admitting that it is a dumb move.

Because again, if you decrease your chances of scoring more than one run while also increasing your chances of scoring 0 runs, all you've done is decrease your chances of winning while also increasing your chances of losing. Making it a very bad decision.

The '1 run' isn't the big factor here. It's the '0 runs' since that's what you're trying to avoid.

Your golf analogy would be similar if by laying up, you not only decreased your chances for a birdie but also increased your chances of a bogey. If you simply need a par, then you want whatever play gives you the best chance for doing better than a bogey, not only getting a par.

Like I said, you ruined your own point.

[blacklistedbully]

I've always said that your point on bunting is generally correct, though there are times where I'm ok with it. You're absolutely correct that we've been doing it far too much, though, especially early in games.

The statement, 'While it does decrease your chances of scoring more than 1 run, and also slightly increases your chance of scoring zero runs,' literally means - 'While it decreases your chances of winning, it also slightly increases your chance of losing.' No clue how that's supposed to be an effective argument.

You being, "clueless" on this one has nothing to do with the effectiveness of my argument, or the supporting statistics. You apparently just don't, "get it", and are over-simplifying.

You continue to look at it from too general a perspective. You are not grasping that there are times when a team REALLY REALLY REALLY needs to do their best to get 1 run, not, "at least 1 run", but 1 run to stay in the game. If you know anything at all about risk analysis, you should be able to get this. It is a fact, statistically that the very best odds, in general of scoring 1 run in a man on 1B no outs scenario is to bunt the guy to 2nd. It comes at a cost, that being that you reduce your chance of scoring more than 1.

Of course there are contributing factors that do matter, such as personnel involved, including their relative health, whether they are currently performing at a level that differs from their to-date stats, who is pitching for us, who is pitching for them, etc, etc. Those are thing a coach or manager may know that we don't, and we are relying on our coaches to know those things and base their decision at least in part on those factors.

All this said, I tend to agree with all those who say we bunt way too much and in situations that don't seem to make sense. The purpose of my posts in this thread was to simply point out that there are instances where bunting is an absolutely viable strategy, and to back it up with statistics.

[smootness]

Scoring 2 runs is not worse than scoring 1 run, though. That's where your issue comes in. It's not always better, but it's never worse.

The bottom line is that if your odds of scoring 0 runs increases, no other statistic matters as that's the only outcome you're trying to avoid.

[Smitty]

It's about the odds of scoring. Smootness is correct. Anything that raises the odds of not scoring at all is bad in this case.

Our bunt numbers compared to the league is shocking. Especially what teams are bunting and not

[blacklistedbully]

Scoring 2 runs is not worse than scoring 1 run, though. That's where your issue comes in. It's not always better, but it's never worse.

The bottom line is that if your odds of scoring 0 runs increases, no other statistic matters as that's the only outcome you're trying to avoid.

You really just don't get it. Your scenario that increases chances of scoring 2 runs also decrease the chances of scoring 1 run. Statistically, the reason the odds of scoring zero go up when bunting is because the odds of scoring 2 or more go way down. Not all of that goes to increasing the, "1-run" probability, a smaller portion of the redistribution goes to, "zero runs". You are getting hopelessly lost in the, "at least 1", or, "more than one" loop.

To grasp this, you only have to consider, "what is the greatest chance of scoring the one run I need this inning?" It's not about maximizing the total potential of, "expected runs". It't about maximizing the chances of getting 1 run. You are making a false argument that, "the better odds of scoring more than one" precludes the possibility that the odds of scoring just one are better in certain circumstances. That is an incorrect assumption, statistically.

[blacklistedbully]

Maximizing the, "potential runs" in a given situation does not, by default also offer the best odds of scoring 1 run. If you can't grasp the logic behind this statement, you will never understand the point.

[LiterallyPolice]

Blacklisted, I hate to break it to you, but smoot has you on this one. You can basically divide the percentages into two groups: Scoring 0 runs (unfavorable) and scoring 1+ runs (favorable).

Also, I saw the exact same percentages that you posted here:
http://www.athleticsnation.com/2013/...f-the-sac-bunt

And the percentages you listed for the "Bunting" scenario are actually the percentages for "Man on 2nd with one out". This assumes the bunt is 100% successful. In other words, you don't take the failed bunt into consideration.

So, if you are looking strictly at the %'s, and you want to score one run, the percentages you list suggest you definitely SHOULD NOT bunt. Moreover, the percentages you list do not consider a failed bunt, swaying the decision even further into the DO NOT BUNT category.

(***DISCLAIMER: This is strictly looking at the %'s. Not specific situations.***)

[blacklistedbully]

Blacklisted, I hate to break it to you, but smoot has you on this one. You can basically divide the percentages into two groups: Scoring 0 runs (unfavorable) and scoring 1+ runs (favorable).

Also, I saw the exact same percentages that you posted here:
http://www.athleticsnation.com/2013/...f-the-sac-bunt

And the percentages you listed for the "Bunting" scenario are actually the percentages for "Man on 2nd with one out". This assumes the bunt is 100% successful. In other words, you don't take the failed bunt into consideration.

So, if you are looking strictly at the %'s, and you want to score one run, the percentages you list suggest you definitely SHOULD NOT bunt. Moreover, the percentages you list do not consider a failed bunt, swaying the decision even further into the DO NOT BUNT category.

(***DISCLAIMER: This is strictly looking at the %'s. Not specific situations.***)

I noticed that too. I am admittedly assuming the author factored that into his calculations. If he did not, the entire article is beyond stupid and fatally flawed. I am giving the author, an admittedly anti-bunt guy, the benefit of the doubt on this, as he is clearly not coming from the perspective of a pro-bunt guy, rather he is admitting that a case can be made in limited circumstances.

[Todd4State]

We currently lead the conference in sac bunts with 20.

Texas A&M has 5
Florida has 5
South Carolina has 6
LSU has 9

We are being left behind strategically. The worst teams are bunting, the best teams aren't.

Sac bunts
---------
1. Mississippi State... 20
2. Tennessee........... 19
3. Auburn.............. 14
4. Georgia............. 12
5. Ole Miss............ 11
6. Vanderbilt.......... 10
7. Kentucky............ 9
LSU......……......... 9
9. Alabama............. 8
10.South Carolina...... 6
11.Missouri............ 5
Florida......………... 5
Texas A&M........... 5
14.Arkansas............ 3

We lead the SEC in OBP and RUNS SCORED. And yet, we are "the worst team".

[smootness]

You really just don't get it. Your scenario that increases chances of scoring 2 runs also decrease the chances of scoring 1 run. Statistically, the reason the odds of scoring zero go up when bunting is because the odds of scoring 2 or more go way down. Not all of that goes to increasing the, "1-run" probability, a smaller portion of the redistribution goes to, "zero runs". You are getting hopelessly lost in the, "at least 1", or, "more than one" loop.

To grasp this, you only have to consider, "what is the greatest chance of scoring the one run I need this inning?" It's not about maximizing the total potential of, "expected runs". It't about maximizing the chances of getting 1 run. You are making a false argument that, "the better odds of scoring more than one" precludes the possibility that the odds of scoring just one are better in certain circumstances. That is an incorrect assumption, statistically.

:facepalm:

You think I'm talking about run expectancy, which is a flawed statistic, and I'm not. The reason run expectancy is flawed is because it may give you the average number of runs you're expected to score, but it doesn't tell you the odds of scoring vs. not scoring.

But you brought up the fact that your odds of scoring 0 runs at all increase by bunting. That's bad. It's actually the only bad outcome. Scoring 1 run in a game you're losing by 1 is good, but scoring 2 is even better. Scoring 1 run in a tie game is good, and scoring 2 is just as good. But scoring 0 in either case is the only bad outcome. So you yourself said that bunting in that scenario is bad, you just didn't realize it.

I really don't know how else to explain it. I'm not trying to increase the total of expected runs. I'm just trying to make sure I don't score 0. Like I've said multiple times, you said that bunting both decreases the chances of winning while also increasing the chances of losing. That's the worst of both worlds. If the odds of scoring 2+ runs decrease while the odds of scoring 0 also decrease, you would have a point. But that's not the case.

[blacklistedbully]

"Over the past couple of decades baseball analysts have seemingly discredited the bunt in all but the most obvious situations. Much of their evidence is based on the use of an overall run expectation table that reveals a loss of run potential even with a successful sacrifice. These overall expected runs tables, however, fail to differentiate between the innumerable possible scenarios of the ability of the hitter at bat and those following in sequence. Subdividing the data by batting order position allows a look at more finely dissected sequences of player ability. Although most of this analysis still indicates a successful bunt does not increase the run potential, it certainly shows that it in certain base/out situations it is not as detrimental as commonly believed. In fact, disaggregating by batting order still averages over many different player ability sequences, suggesting that in a number of instances a bunt may actually increase the run potential.

The Retrosheet play-by-play data allows us to partially test this hypothesis that managers can outperform the run expectation tables by a selective employment of the bunt. While not conclusive, the data here is clearly suggestive: in some base/out situations teams do increase the run expectation with a sacrifice bunt beyond the overall run potential implied by the ERT. And furthermore, even in those cases in which the runs expected over the remainder of an inning after a sacrifice bunt are less than that derived from the ERT, the decrease is typically less than the derived value. In addition we can assume that a manager typically bunts in those situations in which the specific sequence of batters is inferior to the average reflected by the ERT.

It is in the case of playing for one run, however, that the overall aptitude of managerial decisions shows up most clearly. As table 6 reveals, when managers bunt they usually increase the likelihood of scoring at least one run in the inning. And this increase is materially greater than that suggested by simply looking at run probability tables. While the bunt should and will remain a controversial managerial decision, it is clear that managers use it more judiciously than a cursory analysis based on the run expectation and probability tables would suggest."

Dan Levitt is the co-author of Paths to Glory, winner of the 2004 Sporting News-SABR Baseball Research Award. He manages the capital markets for a national commercial real estate firm.

[smootness]

I don't know what all the statistics show. I was going based on the numbers you gave, which are the exact opposite of what you just quoted.

So if you are trying to now disprove your own numbers, ok, but I'm not sure what to do now...