To Bunt, or Not to Bunt

[RightFieldCoder]

Since this is such a hot topic in the coach release thread, I decided to start a new one.

 

I agree with the numbers that the chances of scoring with a runner on first with no out is greater than a runner on second with one out, but I believe this is the wrong comparison. You start with the first situation, the second situation is just one possible outcome.

 

Some of the possible outcomes of runner on first, no out (bunt and no bunt) are:

1. Strike/pop/fly out, runner on first one out

2. Throw runner out at second, runner on first one out

3. Double play

4. Runner on second, one out

5. Error or hit, everyone safe.

 

Now, eveyone would rather have #5 happen, but we all know that it's not always going to happen. I believe it is acceptable to bunt if, as a manager, you are willing to give up the chance of having #5 happen to avoid having #3 happen. This can be employed whenever you feel the batter has a good chance of grounding out, including pitchers or slow hitters that tend to hit a lot of ground balls.

 

I know outs are the most precious thing in baseball, and you shouldn't just go throw them away willy nilly. However, if used occasionally, I think the bunt can be a useful offensive weapon. I'm sure if you ran just the pure numbers, you would find the optimal place to put your fielders are the "normal" positions we are all used to, but you could see the Brewers were able to put their fielders in different positions to great extent in certain situations. I think if you study the exact situation and not just general situations throughout history, you can find times to do different things (like play with a shift or bunt a runner over).

[Oldcity]

I just posted this in the fired coach thread, before I saw this one:

 

Quote:

It is also interesting to note that the team with the most sac hits, Washington (91 of 'em) scored the fewest runs in baseball this year. In fact, there is somewhat of a general inverse relationship between runs scored and sac hits this year. The only three teams in the top half in SH and runs scored are Atlanta, St. Louis and Philadelphia.

------------------

 

You just explained it without meaning to. The teams that bunt more are sub-par offensive teams. The Nationals were one of the worst offensive clubs in the league. They hardly had any run producers, which is why they needed to bunt more. Boston, on the other hand, doesn't bunt because... well, would you bunt if you had Ortiz and Ramirez in your lineup? You implied that teams can't score runs because they bunt, but really teams bunt because they can't score runs.

[danzig6767]

For me,it all depends on the situation.What's the score.Who would be the next two hitters.Who is pitching.

 

If it's the bottom of the ninth and i'm down a run with say Billy Wagner or Brad Lidge pitching,i'd bunt the baserunner over as long as it's not my 3-4-5 hitter up.The odds of a home run or stringing a few hits together vs a stud pitcher aren't high.I'd rather move the runner over and hope one hit or sac fly ties the game.

 

If it's the 6th/7th inning vs a middle reliever i'd be much less inclined to bunt unless a crappy hitter is due to bat.

 

It doesn't have to be a same call choice every time,not all situations are the same.

[greeg35]

I agree that it is very situational. If a team is crafted from the ground up in the A's/Red Sox mold, bunting is almost always a bad idea. For many reasons those teams have had success with their philosophies. As Danzig said, for the teams that don't have stacked lineups, "should bunt" situations come up more often due to more poor hitters in the lineup.

 

In an answer to a question in the other thread, I believe bunting is an essential fundamental because

 

a) it's relatively easy to learn and teach, i.e. wouldn't take up too much of your practice time to become proficient at it. (at least I never found it difficult...then again, Randy Johnson wasn't throwing at me http://forum.brewerfan.net/images/smilies/happy.gif )

 

and

 

b) even the best hitting team may be in a situation where they need that one run to win the game, or get an advantage in a pitcher's duel. Sac's and squeezes are a nice option in that spot.

 

For pitchers and bad hitters it is absolutely essential.

[rluzinski]

Quote:

---

This (bunting) can be employed whenever you feel the batter has a good chance of grounding out, including pitchers or slow hitters that tend to hit a lot of ground balls.

---

 

So, pitchers and crappy hitters. Agreed. Whenever a batter has a very poor chance at success (bad hitter or bad matchup for the hitter), bunting is surely defendable. Most other times it's just giving outs away.

 

The "book" tells you that it's easier to score 1 run with a runner at 2nd and 1 out than a runner at first and no outs. It's a trivial matter to actually LOOK at how often a team scored atleast one run in those situations. Using data from 99-02:

 

Percent of Time Atleast 1 Run Was Scored

 

Runner on 1st, 0 outs: 44%

Runner on 2nd, 1 out: 41%

 

So you have about a 80% chance (approx. success rate of sacrifices) to LOWER your chances of scoring 1 run by 3%. Again, since this is an average it's possible for a sacrifice to be the right call, but those situations are rare. It's just hard to find an advantage in using a tactic that results in giving an out away about 20% of the time.

 

Quote:

---

I'm sure if you ran just the pure numbers, you would find the optimal place to put your fielders are the "normal" positions we are all used to, but you could see the Brewers were able to put their fielders in different positions to great extent in certain situations.

---

 

No, the pure numbers (spray charts) show you should overload one side of the infield for some hitters.

 

References:

 

Run Frequency Chart

 

Chance of a Successful Bunt

 

Article: The Truth About Productive Outs

 

Article: Yet Another Productive Outs Article

[backup catchers]

Russ, do you know what the chances of scoring with a runner at third is with 1 out is? Or could you direct me to a place I could find out? I have to believe it is higher than a man on first with no outs. Just wondering about sac bunts with a man on second with no outs. No one seemed to consider that sac bunt situation. It might be the best arguement for bunting as a fundamental skill.

[et bbal]

Why must every bunt be a straight sacrifice? If you don't want to give up an out why not bunt for a hit? Obviuosly, I did not see every at bat of every game, and I am sure there were others, but I can only remember Weeks bunting for a hit this year. I remember when Steve Trout was playing for the Cubs. He was one of the worst hitting pitchers ever. One spring training they worked with him on drag bunting, and he had a half dozen hits before the league figured it out. It Trout can drag bunt, I have to believe a more skilled batsman would be able to.

[nsimps]

Quote:

---

Runner on 1st, 0 outs: 44%

Runner on 2nd, 1 out: 41%

---

 

Can you really base decisions on mass aggregates? Those numbers don't really tell much.

[endaround]

Huh? They say alot. They say having your .300 hitter bunt runners over to second is incredibly foolish.

[rluzinski]

Quote:

---

Russ, do you know what the chances of scoring with a runner at third is with 1 out is? Or could you direct me to a place I could find out? I have to believe it is higher than a man on first with no outs.

---

 

Are you saying the sacrifice moves the runner over 2 bases, or did you mean runner on second and no outs vs. runner at third and 1 out. Using the "run frequency chart" link provided above:

 

Percent of Time Atleast 1 Run Was Scored

 

Runner on 2nd, 0 outs: 73%

Runner on 3rd, 1 out: 76%

 

Your chances go up 3% but unless you can be almost 94% sure the bunt will be successful it's not usually worth it.

 

Quote:

---

Why must every bunt be a straight sacrifice? If you don't want to give up an out why not bunt for a hit?

---

 

While some players can be very successful bunting for a hits (speedy guys like Pods), it's not usually a good way for the average guy to get on base with. With a runner on, it's a REALLy bad way to get a hit. Here is a breakdown of results of bunting with less than 2 outs and a runner on base for 2004:

 

Double play: 1.2%

Force out: 7.3%

Force out + Error: 0.2%

Fielder's Choice: 0.9%

Fielder's Choice + Error: 0.2%

Error: 0.5%

Single: 8.0%

Single + Error: 1.2%

Strikeout: 6.7%

 

The batter has, on average, almost as good of chance of striking out as he does getting a single. Now obviously many of those strikouts where from pitchers but many of the singles were probably from the Pods and Pierre's of the league. For speedy guys like that, bunting can be an effective tool.

 

Quote:

---

Huh? They say alot. They say having your .300 hitter bunt runners over to second is incredibly foolish.

---

 

Exactly. Because they ARE just averages, I realize very un-average situation exists that can tip the scales in favor of utilizing a sacrifice (Ben Sheets vs. anyone, for instance). The reverse is then also true, however. Even if Brady can successfully bunt 95% of the time (he can't), it's almost never statistically justifiable to ask him to sacrifice.

[nsimps]

Quote:

---

Huh? They say alot. They say having your .300 hitter bunt runners over to second is incredibly foolish

---

 

No they don't. They tell what has happened in the long history of baseball. It doesn't break down anything based on relative strength of the batters.

[rluzinski]

No they don't. They tell what has happened in the long history of baseball. It doesn't break down anything based on relative strength of the batters.

 

The numbers am am providing comes from the seasons of 1999 - 2002. If you don't find data from actual game situations relevent in this discussion, I guess there's nothing to discuss at all. I think it's very fair to make generalizations based on aggregates. Steadfast rules would be difficult to formulate from them, but to say they tell you "nothing" isn't fair at all.

[nsimps]

Quote:

---

If you don't find data from actual game situations relevent in this discussion, I guess there's nothing to discuss at all.

---

 

So if the 8th batter gets on base with no outs the pitcher should swing away because there is a better chance of scoring a run?

[rluzinski]

So if the 8th batter gets on base with no outs the pitcher should swing away because there is a better chance of scoring a run?

 

Let's find some of my quotes:

 

So, pitchers and crappy hitters.

 

Whenever a batter has a very poor chance at success (bad hitter or bad matchup for the hitter), bunting is surely defendable.

 

Again, since this is an average it's possible for a sacrifice to be the right call, but those situations are rare.

 

Because they ARE just averages, I realize very un-average situation exists that can tip the scales in favor of utilizing a sacrifice (Ben Sheets vs. anyone, for instance)

 

Perhaps you should reread this thread before diving in to disagree.

[nsimps]

I guess you don't get my point. If those run scoring statistics are so acurate, how can you discount them based on the batter? They also don't take into account any relative strengths of the batters that came up in the situations. Is the aggregate OPS of the batters that came up with a runner on first and no outs the same as the batters that came to bat with a runner on second and one out?

[rluzinski]

I guess you don't get my point.

 

I understand your point perfectly, and it's a valid to question the numbers, but to say they tell you "nothing' is simply false. Statistics use averages to make generalizations all the time, which is all that I'm doing myself. I've been careful not to overstate the significance of the exact probabilities, since the difference between a couple percentage points is insignificant. After factoring in actual sacrifice bunt success rates, it's really not even close for many situations.

 

Is the aggregate OPS of the batters that came up with a runner on first and no outs the same as the batters that came to bat with a runner on second and one out?

 

If we are to assume that weaker hitters are generally called on to sacrifce runners over (creating the runner on second, 1 out scenario), then it could be postulated that the aggregate OPS of batters that came on with a runner on second and 1 out might be higher. How would that help the argument for sacricing though? Are you insinuating that the runner on first scenario usually has better batters feeding it, inflating the probabilities? I can't see how.

 

I suspect that the actual difference in OPS in those situations isn't significantly different anyway.

[homer]

I don't mind bunting, but it's got to be the right batter in the right situation. When the Brewers were playing the Astros a few weeks ago, Brady let off the game with a double. Yost promptly sent Cirillo up to bunt him to third. To me, this is just not a good move. Why go for one run in the first inning? Had they needed one in the 7th or 8th that would have been fine.

[xisxisxis]

I think it's wise to bunt when the likelihood of successfully bunting/moving the runner over is greater than the likelihood of the hitter not making out.

[nsimps]

Quote:

---

but to say they tell you "nothing' is simply false.

---

 

I'm sorry I overstated it, but using means of large numbers of samples is not a very strong statistical approach.

 

Quote:

---

If we are to assume that weaker hitters are generally called on to sacrifce runners over (creating the runner on second, 1 out scenario), then it could be postulated that the aggregate OPS of batters that came on with a runner on second and 1 out might be higher. How would that help the argument for sacricing though? Are you insinuating that the runner on first scenario usually has better batters feeding it, inflating the probabilities? I can't see how.

---

 

There is no way to know from the data given. The run scoring matrix doesn't give any indication of the difference in the situations. They may have simply been scenarios with the 5th hitter getting on to start an inning, etc.

 

As for sacrificing, it depends on the individual batters. I'm not suggesting a .300 hitter bunt a runner over. It depends on the hitters behind a certain player and the particular probabilities of a certain player obtaining a certain result.

[rluzinski]

They may have simply been scenarios with the 5th hitter getting on to start an inning, etc.

 

As you say, they are simply a mass aggregate of all cases with a particular out and player on base. I understand your concern is that certain situations might tell you something about the players that were involved (if the bases are loaded and no out, it's probably because better hitters where up, which infers worse hitters WILL be up) but I don't think that effect distorts the data to the point where generalizations can't be drawn from it. I don't think that effect would distort the data much at all, really.

 

It depends on the hitters behind a certain player and the particular probabilities of a certain player obtaining a certain result.

 

...and I've mentioned that fact many times in this thread, that's not even up for debate. My contention is simply that it's dificult to statistically justify bunting with a non-pitcher when about 1 out of 5 sacrifice attempts fail anyway.

[GoodGod262]

A quick question about the percentages of a guy on first with no outs scoring 44% of the time. I am just assuming that since you said aggregate percentage that there were numerous outcomes to get the runner on first with no outs home.

 

How many times did that guy on first with no outs turn into a guy on second with 1 out? How many times did that guy end up on third and become a guy on third with no outs? How many times did that guy end up on second with no outs? How many times did the guy on first run into a double play? It just seems to imperfect of a percentage to say one must never try to get the guy on second with one out 99.9% of the time.

 

I do agree that it would be ideal to never give outs up by bunting. But if a situation/managerial decision calls for a bunt I would expect a batter to get it down better than league average of 55%.

[rluzinski]

A quick question about the percentages of a guy on first with no outs scoring 44% of the time.

 

To clarify, the actual runner may not be the one to score (force out at 2B, double would score a run but not the actual runner). We are only concerned with different baserunning/out states and how many runs were scored in each.

 

I am just assuming that since you said aggregate percentage that there were numerous outcomes to get the runner on first with no outs home.

 

Yep.

 

How many times did that guy on first with no outs turn into a guy on second with 1 out? How many times did that guy end up on third and become a guy on third with no outs? How many times did that guy end up on second with no outs? How many times did the guy on first run into a double play?

 

While data of that nature would certainly be useful, it's not really needed for this kind of analysis. Managers use sacrifices to transform one state into another, with the assumption that the later state either generates more runs or generates 1 run more often than the previous state. Managers realize that whatever that benefit is, it must be larger than the negative of a botched sacrifice attempt (force out or double play).

 

Since the data of the states is dependant on an average offensive player against an average pitcher and defense, we simply make generalizations based on those averages. We can conclude that if a batter is of average ability or greater, and the pitcher does not have a substancial advantage over that particular player (lefty vs. lefty for example) the manager is maximizing his chance to score a run by letting him swing. Now, who bats after him is also important, that's why I'm not being overly aggressive in my assessments to begin with. The fact that about 20% of sacrifice bunts fail makes it an easy decision for an average player.

 

Sorry for the long post that basically says the same thing over again.

[backup catchers]

Are you saying the sacrifice moves the runner over 2 bases, or did you mean runner on second and no outs vs. runner at third and 1 out. Using the "run frequency chart" link provided above:

 

I was saying a runner at 2nd and bunted to third. I thought I clicked that chart but must have missed it the first time around thanks.

This is one case where I believe the stats are reliable. Obviously each situation and hitter has to be considered as there is no absolute answer based on stats when it comes to game management. But in a normal situation with a run of the mill average hitter I would tend to go with the statistical averages for the best chance to score. It's hard to argue facts.

[nsimps]

Quote:

---

My contention is simply that it's dificult to statistically justify bunting with a non-pitcher when about 1 out of 5 sacrifice attempts fail anyway.

---

 

But there is no good statistical evidence that says a runner in scoring position with one out is worse than a runner on first with no outs. There is a time and place for bunting, especially when you need to just convert one run.

[RightFieldCoder]

Ok, I'm glad we're basically in agreement RLuz. Average hitter and average pitcher, don't bunt unless it's the end of the game. Situational bunting is ok. Most Brewer pitchers, or someone like Moeller when you know you're going to pitch hitting behind him are ok to call for the bunt. I just thought that you didn't want us to ever bunt. Glad I was reading you wrong.

 

And at no time should we bunt in the first inning. If your pitcher is at the plate in the first inning, the other pitcher is obviously not doing so well. Swing away.

 

The problem with teaching them drag bunting: if they can't get down a basic sac bunt, how are they going to fare trying to drag bunt?