Project for Smitty...

[messageboardsuperhero]

I looked at the innings which our leadoff man got on first base in conference play for analysis. I looked at what the outcomes where when we immediately followed a leadoff one-bagger (Hit, HBP, BB, IBB, etc) man on 1st 0 out. I knew it was bad, but nothing like this.

With man on 1st, 0 out. WE BUNT
Occurrences: 26
Score a run: 6
# of runs : 6

We bunted in that position 26 times, only scoring 6 times, one run each inning. That is 23% which is off the charts bad yielding .2 runs/inning. Thats after having a man on 1st and no outs

With a man on 1st, 0 out. WE DONT BUNT
Occurrences: 66
Score a run: 36
# of runs: 77

We didnt bunt in that situation 66 times, scoring 36 times, with many occurrences of multiple runs. That is 55% of innings we score yielding 1.12 runs/inning, A FULL RUN higher than bunting in the same situation.

55% > 23% of innings we score
1.12 > 0.2 runs per inning

Say what you want about me and this issue. These statistics don't lie.


My post from Sixpack on May 11th, 2012.

My God, I had no idea it was THAT bad... Of course that's from 2012, but I bet the numbers would be pretty consistent this year if we had a larger sample size. Cohen needs to check his ego and let the players hit- it is undeniable at this point.

ETA: And again, that doesn't even count the times we fail to get the bunt down and end up having to swing away in pitcher's counts... Which undoubtably make the "bunt" numbers look better than they actually are and the "no bunt" numbers look worse than they actually are.

[Smitty]

This is where I said bunting HURTS bad hitting teams worse, not the other way around. The meme is bad hitting teams have to manufacture runs. That's true in a sense but sac bunting is NOT part of that. Bad hitting teams being forced to give away a free out just makes them so much worse.

[Bo Darville]

Smitty said he is from Tupelo. I assume he goes to elementary school, so maybe he could work on this at recess.

[Smitty]

Smitty said he is from Tupelo. I assume he goes to elementary school, so maybe he could work on this at recess.

Rent free in your head

[Bo Darville]

Rent free in your head

It's no secret that I don't like you.

[blacklistedbully]

Smitty said he is from Tupelo. I assume he goes to elementary school, so maybe he could work on this at recess.

The dude brings up good points. My only problem with WJ/Smitty is his unrelenting need to ruin game threads with constant bitching about the same thing over & over, as if we don't all already know his POV.

But this is a thread specifically about the question of bunting as a strategy, and more specifically asking for his input. I'm reading it because I'm interested in hearing his details and the discussion, because it's in the proper forum, IMO.

[blacklistedbully]

Quote Originally Posted by Old Will James 6-pack profile View Post
I looked at the innings which our leadoff man got on first base in conference play for analysis. I looked at what the outcomes where when we immediately followed a leadoff one-bagger (Hit, HBP, BB, IBB, etc) man on 1st 0 out. I knew it was bad, but nothing like this.

With man on 1st, 0 out. WE BUNT
Occurrences: 26
Score a run: 6
# of runs : 6

We bunted in that position 26 times, only scoring 6 times, one run each inning. That is 23% which is off the charts bad yielding .2 runs/inning. Thats after having a man on 1st and no outs

With a man on 1st, 0 out. WE DONT BUNT
Occurrences: 66
Score a run: 36
# of runs: 77

We didnt bunt in that situation 66 times, scoring 36 times, with many occurrences of multiple runs. That is 55% of innings we score yielding 1.12 runs/inning, A FULL RUN higher than bunting in the same situation.

55% > 23% of innings we score
1.12 > 0.2 runs per inning

Say what you want about me and this issue. These statistics don't lie.

What stands out to me is, how far above the average we score a run when swinging away. Assuming your stats are correct, we are scoring a run in those situations at a rate that is 3 times higher than the average in MLB over a 10-year period, according to the study we discussed last week. In that study, a run scored less than 19% of the time when swinging-away. We're greater then 55%?

[Smitty]

Remember those were the 2012 SEC numbers... Boyd says man on 1st no outs scored 49% of the time. Swinging away scored 55% for us. Bunting scored 23% for us.

Swinging away was supposed to yield 1.07 runs in the inning. We scored 1.12 swinging away. 0.2 bunting.



http://www.boydsworld.com/data/ert.html

[messageboardsuperhero]

So you are 3% more likely to score with a man on 3rd and one out than you are to score with a man on 2nd and no outs?

Outs 0 1 2
Empty 32% 18% 07%
1st 49% 32% 15%
2nd 70% 47% 25%
1st and 2nd 71% 49% 27%
3rd 89% 73% 32%
1st and 3rd 89% 70% 34%
2nd and 3rd 89% 73% 32%
Loaded 90% 72% 39%

Boy, giving away that out after a leadoff double on the sac bunt (virtually kill any chance of scoring multiple runs) really looks like a smart move now.***

[Smitty]

So you are 3% more likely to score with a man on 3rd and one out than you are to score with a man on 2nd and no outs?

That assunes around a 95% bunt success rate getting him there as well. Something that obviously cannot be assumed.

[blacklistedbully]

OK, so I looked at your linked tables, and it appears Boyd does not differentiate between, "swinging away" and "bunting" in those ERT tables. So, when he shows, "at least 1 run scored" 52%/49% of the time with runner on 1st and no outs, it's also including runs scored when bunting the runner to 2b. Where is the link showing the breakdown you mention?

[engie]

OK, so I looked at your linked tables, and it appears Boyd does not differentiate between, "swinging away" and "bunting" in those ERT tables. So, when he shows, "at least 1 run scored" 52%/49% of the time with runner on 1st and no outs, it's also including runs scored when bunting to the runner to 2b.

In that situation, the comparison would be "runner at second one out" for the bunt after the leadoff base runner vs "runner at first no outs" for the leadoff single not followed by a bunt. So on, so forth. That's how you make the comparisons. These tables aren't perfectly accurate -- in that they don't account for failed sac bunts. But they also don't account for sac bunt errors/base hits either. For simplicity sake, it's easier to just consider all of that a wash, and probably pretty close to accurate...

Regardless -- there's a wide crater of a difference for us. I'd assume that it's MUCH wider than Boyd's numbers currently, given our almost uncharted failure at situational hitting with RISP and 1 out. It was certainly ALOT wider in 2012...

[engie]

I really need the numbers from this year and last year. I'll do them myself, but I'm not going to have time this week. We're quickly approaching critical mass with this, at least in my head. I wasn't nearly as bothered by it when I felt we had subpar hitters. That's no longer the case. We should have a top 25ish lineup right now talent-wise. And we need to eliminate the obvious/easy stuff to allow that to happen...

Rather than just sitting here and incessantly bitching about this, which obviously makes no sense to the majority of us(both in posting it repeatedly and what Cohen is actually doing), I'm going to do everything in my power to make sure he has all of this possible information at his disposal. Alot of times, I've found it alot easier to present people with the same statistics in a little different manner, and let them do with those statistics as they may. If someone went after Cohen with "you need to change all this stuff", it's obviously going to piss him off and probably just make him dig his heels in even more and wonder "who the F are these people to question me". I've got a few people in his circle that can hand him these stats in a "you need to check this out and see what you think" manner -- without comment or condemnation -- allow him to marinade over them -- and just see what happens. That's what I can actually do to try to help make us better, instead of just sitting here taking the fun out of baseball discussion complaining about it...

[War Machine Dawg]

I really need the numbers from this year and last year. I'll do them myself, but I'm not going to have time this week. We're quickly approaching critical mass with this, at least in my head. I wasn't nearly as bothered by it when I felt we had subpar hitters. That's no longer the case. We should have a top 25ish lineup right now talent-wise. And we need to eliminate the obvious/easy stuff to allow that to happen...

Rather than just sitting here and incessantly bitching about this, which obviously makes no sense to the majority of us(both in posting it repeatedly and what Cohen is actually doing), I'm going to do everything in my power to make sure he has all of this possible information at his disposal. Alot of times, I've found it alot easier to present people with the same statistics in a little different manner, and let them do with those statistics as they may. If someone went after Cohen with "you need to change all this stuff", it's obviously going to piss him off and probably just make him dig his heels in even more and wonder "who the F are these people to question me". I've got a few people in his circle that can hand him these stats in a "you need to check this out and see what you think" manner -- without comment or condemnation -- allow him to marinade over them -- and just see what happens. That's what I can actually do to try to help make us better, instead of just sitting here taking the fun out of baseball discussion complaining about it...

[Tbonewannabe]

I look forward to your analysis. We finally have highly recruited hitters but we aren't getting the production like I thought.

[blacklistedbully]

nm

[blacklistedbully]

In that situation, the comparison would be "runner at second one out" for the bunt after the leadoff base runner vs "runner at first no outs" for the leadoff single not followed by a bunt. So on, so forth. That's how you make the comparisons. These tables aren't perfectly accurate -- in that they don't account for failed sac bunts. But they also don't account for sac bunt errors/base hits either. For simplicity sake, it's easier to just consider all of that a wash, and probably pretty close to accurate...

Regardless -- there's a wide crater of a difference for us. I'd assume that it's MUCH wider than Boyd's numbers currently, given our almost uncharted failure at situational hitting with RISP and 1 out. It was certainly ALOT wider in 2012...

That is incorrect. You are assuming the % of times a run scores with a man at 1b and no outs is just from, "swinging away". Nothing in that table indicates that. In fact, as it stands, the 52/49 number appears to include all instances of a run scoring from that situation, including those that score as a result of being sac bunted to 2b.

Of course that number is going to be much higher than one that shows only runs scored with a man at 2b with 1 out. Again, it includes all of the times a run scored after being sac bunted over to a, "man on 2b with 1 out" situation. That number is useless unless or until you parse out the number of runs scored that included the aforementioned sac bunt.

It's simple math.

[engie]

That is incorrect. You are assuming the % of times a run scores with a man at 1b and no outs is just from, "swinging away".

No I'm not.

Nothing in that table indicates that.

I didn't say that it did.

In fact, as it stands, the 52/49 number appears to include all instances of a run scoring from that situation, including those that score as a result of being sac bunted to 2b.

You are missing the forest for the trees. As you do constantly on this bunting discussion. How the guy got to second with one out is negligible. If he's got a 49% chance to score from second with one out -- and a 53% chance of scoring from first with no outs. We are statistically less likely to score one run -- and even more statistically less likely to score more than 1 run -- it doesn't make a damn bit of sense to bunt him over. Doesn't matter if "runner on first includes situations where he's bunted to second" -- because the act itself of bunting him to second creates the second circumstance being analyzed and therefore cancels itself in the analysis. Thus, the second situation is now in play -- with a runner on second and one out.

Of course that number is going to be much higher than one that shows only runs scored with a man at 2b with 1 out.

If this is so likely -- then define for me the purpose of giving up the first situation in order to create the second one? You make these statistical arguments -- then you say something that completely destroys your own point. Like you just did here.

Again, it includes all of the times a run scored after being sac bunted over to a, "man on 2b with 1 out" situation.

Which we've already seen, statistically, brings the percentage chance of scoring down. As well as the average number of runs scored declining sharply. So, in reality, you are making a case that the difference is even greater than has been shown statistically by Boyd. The "runner on first no outs" number is, thus, inherently deflated(NOT INFLATED) by including bunting into the second scenario.

That number is useless unless or until you parse out the number of runs scored that included the aforementioned sac bunt.

Boyd's numbers are very telling. But not NEARLY as telling as MSU's same data set analysis where we struggle far beyond the norm with RISP.

It's simple math.

Apparently not.

I gave you A - B = C.
You are giving me A - (B+A) = C - A

[blacklistedbully]

No I'm not.


I didn't say that it did.


You are missing the forest for the trees. As you do constantly on this bunting discussion. How the guy got to second with one out is negligible. If he's got a 49% chance to score from second with one out -- and a 53% chance of scoring from first with no outs. We are statistically less likely to score one run -- and even more statistically less likely to score more than 1 run -- it doesn't make a damn bit of sense to bunt him over. Doesn't matter if "runner on first includes situations where he's bunted to second" -- because the act itself of bunting him to second creates the second circumstance being analyzed and therefore cancels itself in the analysis. Thus, the second situation is now in play -- with a runner on second and one out.


If this is so likely -- then define for me the purpose of giving up the first situation in order to create the second one? You make these statistical arguments -- then you say something that completely destroys your own point. Like you just did here.


Which we've already seen, statistically, brings the percentage chance of scoring down. As well as the average number of runs scored declining sharply. So, in reality, you are making a case that the difference is even greater than has been shown statistically by Boyd. The "runner on first no outs" number is, thus, inherently deflated(NOT INFLATED) by including bunting into the second scenario.


Boyd's numbers are very telling. But not NEARLY as telling as MSU's same data set analysis where we struggle far beyond the norm with RISP.


Apparently not.

I gave you A - B = C.
You are giving me A - (B+A) = C - A

Dude, you are still not getting it! I am not arguing that bunting is THE solution when you have a runner on first with no outs. I have NEVER taken that stance, and have, in fact stated we do it too early and too often. My previous argument on bunting, in another thread, was that there are stats that back up a suggestion that the odds of scoring exactly 1-run can go up when bunting in a man-on-1-b-with-no-outs. There are also real stats based on a 10-year analysis of MLB that show there are times when, if you're really playing for that 1 run, it can make sense to bunt the guy over, depending on who the players are at that given time. But this strategy also has the negative effect of reducing the chances of scoring more than 1 run.

As far as this argument, in this thread, I am responding to your assertion that I somehow got it wrong when I pointed out the linked, "Boyd's ERT tables" arrived at the 52/49 number by including runs scored when the runner was bunted over to 2nd, resulting in a man-on-2b-1-out situation. The fact that the odds of scoring a run go down from there has NOTHING to do with the fact that the 52/49 number INCLUDED all the runs that DID score from the subsequent man-on-2b-1-out situation.

This is not to argue against the numbers being better, in general by swinging away, rather that the 52/49 number DOES include EVERYTHING, including subsequent bunts.

In looking at the Boyd stats again, the overall % difference between scoring at least 1 run from 1st with no outs and scoring from 2nd with 1 out, is 3% (2% for '11-'12). With that narrow a gap, you are clearly in the area where the, "who's at bat, who's OB, who's pitching, etc" are factors that could suggest that bunting is a potentially better option. For instance, if you really, really need to do all you can to get 1 run in, have a guy AB that isn't a good hitter and/or is a good candidate to be doubled up, but IS a good bunter, it might make sense to bunt. But early in-the-game is not a time to be playing for 1-run, as it comes at too high a cost as it relates to potential total runs scored.

Consider this - as much as you want to look at the 2-3% increase of a run scored when comparing 1-on-1b-no-out versus 1-on-2b-1-out, have you looked at the difference between 1-on-2b-1-out versus 1-on-1b-1-out? % drops by 15. So, if you have a damn good bunter at the plate, he should be able to get you to the former. Only the better hitters are going to do better than that swinging away the vast majority of the time. But again, it comes at the cost of lower probability of scoring more than 1 run.

ETA - by, "good bunter" I mean a guy that can be relied upon to drop it down consistently. An especially good bunter who is capable of dropping one to 3b at the right speed has a damn decent chance of getting a hit. Even Bill James research shows a batting average in, "Zone 6", even on sac attempts averages out to a .291 batting average. Of course, a problem is we are bunting guys who aren't that skilled at it.

[engie]

What's the "goal" of this argument you are pulling out of thin air? The sake of arguing?

Dude, you are still not getting it!

I'm not the one failing to get it. With later quotes in this post, I literally don't think you know the difference between sacrifice and drag bunting -- it's going to be impossible for me and you to come to common ground in that case...

My previous argument on bunting, in another thread, was that there are stats that back up a suggestion that the odds of scoring exactly 1-run can go up when bunting in a man-on-1-b-with-no-outs.

Which was a bad position and wide-ranging argument you made, based on a single situational snapshot playing for a walk-off win. That's the only time in a baseball game that you are ever playing for exactly one run. Smoot took that argument apart. But you not getting his point then virtually assures that you won't get mine now, so I don't really know why I'm trying...

There are also real stats based on a 10-year analysis of MLB that show there are times when, if you're really playing for that 1 run, it can make sense to bunt the guy over, depending on who the players are at that given time.

Link? No one said to "NEVER" bunt. You've never heard that come out of my or anyone else here's mouth. You are creating phantom opposition. And you are still hung up on your "one run" ideal.. Only applicable as a home team in a tie game in the final/extra innings. Otherwise, in reality, exactly one run is NEVER the actual goal of the inning, thus not even the actual goal.

But this strategy also has the negative effect of reducing the chances of scoring more than 1 run.

Which is why it should be employed rarely...

As far as this argument, in this thread, I am responding to your assertion that I somehow got it wrong when I pointed out the linked, "Boyd's ERT tables" arrived at the 52/49 number by including runs scored when the runner was bunted over to 2nd, resulting in a man-on-2b-1-out situation. The fact that the odds of scoring a run go down from there has NOTHING to do with the fact that the 52/49 number INCLUDED all the runs that DID score from the subsequent man-on-2b-1-out situation.

Still missing the forest for the trees.

52% scoring from 1st with no outs.
49% scoring from 2nd with 1 out.
If the odds went down after it happened, then the second situation is NOT HELPING THE FIRST SITUATION, BUT RATHER HINDERING IT.

How can you possibly argue that "that second data set significantly effects the first". The only way it effects the first is to artifically deflate a number that should be more like 53%. Hence not supporting your previous assertions about bunting...

This is not to argue against the numbers being better, in general by swinging away, rather that the 52/49 number DOES include EVERYTHING, including subsequent bunts.

Trees. Forest. Everyone knows the "numbers include everything". What's your point? What does that change with significance about the numbers? Or was it just to argue for the sake of arguing?

In looking at the Boyd stats again, the overall % difference between scoring at least 1 run from 1st with no outs and scoring from 2nd with 1 out, is 3% (2% for '11-'12). With that narrow a gap, you are clearly in the area where the, "who's at bat, who's OB, who's pitching, etc" are factors that could suggest that bunting is a potentially better option.

Not really. Not to sacrifice.

For instance, if you really, really need to do all you can to get 1 run in, have a guy AB that isn't a good hitter and/or is a good candidate to be doubled up, but IS a good bunter, it might make sense to bunt.

Where did anyone ever argue "all bunting is bad and that you should never sac bunt"? Why are you arguing against a position nobody is taking?

Consider this - as much as you want to look at the 2-3% increase of a run scored when comparing 1-on-1b-no-out versus 1-on-2b-1-out, have you looked at the difference between 1-on-2b-1-out versus 1-on-1b-1-out? % drops by 15. So, if you have a damn good bunter at the plate, he should be able to get you to the former. Only the better hitters are going to do better than that swinging away the vast majority of the time. But again, it comes at the cost of lower probability of scoring more than 1 run.

Link? You are now to make theoreticals say what you want them to say... I'm not interested in your theoreticals in this thread. I didn't start it for that purpose.

ETA - by, "good bunter" I mean a guy that can be relied upon to drop it down consistently. An especially good bunter who is capable of dropping one to 3b at the right speed has a damn decent chance of getting a hit.

Holy crap -- now we're talking drag bunting?
I have never once condemned bunting for a hit in any situation.

Even Bill James research shows a batting average in, "Zone 6", even on sac attempts averages out to a .291 batting average. Of course, a problem is we are bunting guys who aren't that skilled at it.

Drag bunting is different from sac bunting.

You need to take a little break and think through your position a little better rather than using a number you found on "better odds to score exactly one run(as if that's a goal 99% of the time) and throwing together a bunting philosophy off of that number, which is a very minimal participant in overall bunting situations.