Evaluating Wakamatsu’s use of the sac bunt
Wakamatsu’s use of the sac bunt before is worth a much more in-depth discussion than we’ve given it so far. The sac bunt is wildly overused, certainly, but it’s also an effective tool at times. What are the M’s doing with it?
There are eleven plays scored as sacrifice hits:
Gutierrez has four
Endy Chavez has three
Kenj, Yuni, Jose, and Rob Johnson have one
Now remember, that’s sacrifice bunts. If you bunt and get on, unless you reach by error it’s not counted as a sac. If the bunt is on for a pitch and then taken off, it’s not in the books so I can’t go back and look to see what the situation was. And (and this’ll be important later) if you bunt for a hit and it fails but advances the runner, you get the sac hit. So if Chavez has a green light to bunt for hits, the manager will take the credit and blame for sac bunts he didn’t order. I’m focusing here on the 11 SH on the book. Oh, and I’m going to use WPA a lot, so if you’re unfamiliar with that, run over here first.
And the last thing before we start, here’s my super-high-level summary of where I am on the bunt: no one but a pitcher should bunt before it’s late in the game, when it can be a good tool in close games.
So, by game:
4/7 vs. Twins
M’s visiting, lead 4-3, 9th inning, 1 out, runner on first. Chavez bunts to advance Balentien.
This is almost a wash. This was counter-intuitive to me, but bunting while ahead’s a much better move than you’d think. Late inning bunts in close games are better than bunting in the second. Not a great move, but Chavez isn’t a great hitter and Gutierrez was up next. WPA-wise, it’s -.011, as close to a wash as you get.
Okay bunt.
4/9 vs Twins
M’s visiting, lead 1-0, 3rd inning, no outs, runner on first. Gutierrez sacrifices Chavez to second.
Generally speaking, you should never play for one run at the expense of the chance at many runs until the sixth inning at the very earliest. The WPA here comes out to -.016, and that’s still not so bad. It’s “flied out with no one on†important. And if you think Gutierrez isn’t an average hitter, it starts to look good.
Okay bunt.
4/10 vs A’s
M’s visiting, lead 5-4, 9th inning, no outs, runner on first. Kenji bunts Branyan to second.
Again, bunting while ahead, surprisingly productive historically. Here we have to consider what we think of Kenji’s hitting. If you see Kenji as hopeless at the plate, this is a good move. If he’s an average hitter, it’s essentially a wash. I’m torn here: I really want to say that until we see Kenji hitting well again, you have to assume the bunt’s the right move. I’m going to say that, actually. The straight WPA is -.007
Good bunt.
4/11 vs A’s
M’s visiting, up 7-5, 9th inning, no outs, runner on first. Gutierrez bunts to advance Chavez.
Straight WPA here is -.003. It’s a wash if Gutierrez is average. Another batter dependent move, and certainly not a bad one.
Okay bunt.
4/14 vs Angels, bunt one
M’s at home, tied 2-2, 7th inning, no outs, runner on first. Gutierrez bunts to advance Johjima.
I don’t like this one. WPA’s -.020, beyond which you’re advancing Kenji in the hope that what, he can score from second on a single? And behind him is Yuni the ground ball machine, which makes advancing on a sac less likely. And while the score’s tied, you’ve got two innings left for the A’s to score – I don’t get it.
Bad bunt.
Bunt two
M’s at home, 2-2, 10th inning, no outs, runner on second. Betancourt bunts to advance Gutierrez to third.
As it turns out, this wins the game. But it’s a good bunt even if it gives up the out to move the runner along. One run wins the game, and moving Gutierrez over makes that one run more likely. Now… the WPA on 10th-inning moves starts to get weird, because the sample sizes drop off so hugely, but this is probably a +.030 bunt. Really – it’s as good a move as two or three of those bad ones.
Moreover, after Betancourt you’re looking at Endy and Cedeno.
Good bunt. Reaaaally good bunt, and it turned out wildly better than that.
4/15 vs Angels, bunt one
M’s at home, 0-0, 4th inning, no outs, runner on second. Lopez bunts to advance Branyan to third.
I don’t know about this one. It’s a bad idea to play for one run so early. But the next three batters were Kenji, Betancourt, and Gutierrez. You’re probably not hoping for more than a single from any of those guys, and Branyan’s slow, but he could score from third. But Lopez isn’t a bad hitter, and it’s not a force out situation where you’re worried about seeing Branyan doubled off. On the other hand, it’s Jered Weaver.
The book says it’s a -.014 play.
Okay bunt.
Bunt two
Then behind 3-2, 6th inning, runners on first and third. Gutierrez bunts to move Betancourt to second.
The pitcher makes an error, Johnson scores, and Gutierrez gets to first. I uh… I totally love this play. Seriously. I will trade an out for a run every day of the week. Now I’m not wild about this because Ichiro’s on deck. But Ichiro’s a ground ball machine, and if you pull it off you’ve removed the double play, and if Johnson’s thrown out there’s a good chance you end up with Betancourt at 3rd and Gutierrez at 1st. I dig it.
Good bunt.
4/17 vs the Tigers
M’s at home, trailing 3-1, 5th inning, runners on first and second. Johnson bunts to move Branyan to third, Lopez to second.
Okay, remember when I said the sacrifice was a bad move early in the game except for pitchers? Rob Johnson hits like a pitcher. He’s terrible. Especially in that situation, where the M’s have three straight hits off Verlander and they’re about to turn over the order.
Good bunt.
4/25 vs Angels
M’s visiting, 0-0, 1st inning, no outs, runner on first. Chavez bunts to advance Ichiro! To second.
Bad. Ortega was nothing special, Ichiro can run on his own, and… the first inning on the road against a filler starter just doesn’t make sense. WPA says -.018 but that requires an average pitcher on the other side. As the M’s would prove by rocking Ortega, there was no reason to give him free outs.
Bad bunt.
4/28 vs White Sox
M’s visiting, 0-0, 1st inning, no outs, runner on first. Chavez bunts to advance Ichiro! To second.
No! Stop bunting early! Argh! (-.018)
Bad bunt.
The totals:
4 good bunts
4 okay bunts
3 bad bunts
That’s good. Really. A manager who does this well picking his spots helps his team.The frequency may be cause for concern, but he’s being smart about it. The only legitimate complaint is the early-game bunting, which is inexplicable. It’s not Don Baylor-in-Colorado crazy, but the weight of baseball history shows it doesn’t help teams win games. And if he’d stop that, it’d be great. And if some of those are just Chavez laying down unsuccessful tries for bunt singles, Wakamatsu looks even better.
I did a little Excel work summing this up and then working with the assumptions in each situation. With pretty conservative estimates I come up even, and it’s not hard at all to come up on the plus side. If you’re interested, there’s a ton of really interesting research on this stuff (James Click in particular has some great bunt articles on Baseball Prospectus and in the Baseball Between the Numbers book, and I’d also recommend the chapter in Tango’s The Book)(affiliate links) and I’d welcome additional thoughts and suggestions for follow-up.
I’m not a bunt guy, and I came around: Wakamatsu’s generally been really smart about picking his spots and getting the most for his out. I admit, so far this year I’ve done a lot more tearing my hair about over bunts that weren’t so bad than I have applauding bunts that on closer analysis look pretty good. When I line them all up like this, I find myself reassured.
(this article took ~3h to research and write up)
Comments
46 Responses to “Evaluating Wakamatsu’s use of the sac bunt”
Leave a Reply
You must be logged in to post a comment.
Wow, that was not the conclusion I was expecting you to come to either. Good job doing the research and seeing where it leads you!
I can’t tell — Do the WPAs assume that the bunt works as planned, or do they include other possibilities, such as an error by the opponents, accidentally getting a base hit, or bunting into a double play?
Derek,
Thanks for the research. Nice to see you guys doing analysis on more than Silva’s pitch location and lack of movement.
=oP
It’s cool that you let the research you did lead you to a conclusion as opposed to what most folks do, which is let their conclusion determine what the reasearch means.
Dennisss, the WPA is a measure of the change in state based on the actual events, it doesn’t assume anything related to intent. Otherwise the inability to identify all the uncounted attempted sacrifices would be a bigger issue.
I wanted to also thank you for the time it took.
I admit I have yelled: “why are you bunting now?” at the tv a few times. Thinking back, they were mostly at bunts early in games.
I wonder if Wak and Van Burkleo planned to bunt a lot due to our bad offence, or if it’s more eclectic in nature…
floyd,
I don’t think Wak & VB counted on the offense being quite this anemic. That being said however, in my experience, proponents of “small ball” are rather rabid in their attachment to its ideals.
I think Wak has done a pretty good job thus far, although most folks seem to agree that the team has bunted in some interesting circumstances. He IS a rookie manager, after all.
Well, at least Wak hasn’t pulled a Bob Melvin. Remember when he had Dan Wilson sac bunt a runner over when the M’s were trailing by two runs late in the game. It was awful, but I felt like I was a part of history, watching the absolutely worst bunt that ever was and ever will be.
I dunno, Tek.
Wak using Lopez in the 3-hole when he’s sporting a .300-ish OBA is pretty inexplicable.
Then again, he did go 3-for-8 batting third in the 2 games vs. KC.
I wish some of the Mariners would try to beat out bunts for base hits. In 1972, Rod Carew beat out 27 of 35 bunt attempts. As recently as 1992 Brett Butler had 41 bunt hits. Bunt prevention defense has improved but you only need a 40% success rate for it to be a good play. Ichiro, Chavez, Gutierrez, and Yuni should lay one down every once in a while.
Is there any current and historical data about bunt hits available online?
DISCLAIMER: OK, OK… it WAS the Royals and Sidney Ponson & Brian Bannister.
One thing that strikes me about bunting is that the PPA is likely much lower. Especially early in a game, there is an advantage to getting the starter to throw more pitches. If Yuni is batting this is less of an issue because he isn’t milking the count anyway, but a batter like Gutierrez has some value in terms of looking at a few pitches. I don’t know enough about WPA or the data on PPA for bunting, but it likely serves as an additional justification for less bunting early in games.
Was the opposing team’s infield defense (including pitchers) also taken into account with this analysis?
So at what point does bunting so often, or in certain situations, get into opponent’s heads in a measurable way? I.e. at what point do you see changes in opponent’s pitching, or in their defensive alignment, simply because Wak develops a reputation for bunting at any time?
Or will it always be a surprise, pretty much, until/unless either the batter shows bunt in that at-bat or the situation screams out for one in general anyway?
2 of the bad bunts were Chavez, and I think its reasonable to think he may have been bunting for a hit. If so, its similar to him swinging, but with lower chance of being successfully safe, but much higher chance to get a less-bad out, moving over the runner. If you take those 3 bad bunts plus one that he reaches safely on, it’s probalby a net gain. (I’m not sure what %age he would actually be safe on).
I’ve noticed that for the last 7-10 days, the bunts have disappeared for the most part. Those last few games have also seen the M’s struggle to score runs and win games.
This team has a couple strengths on offense in that they bunt well and make good contact. I’d argue that they need to make MORE use of small ball, be it sac bunting, hit and runs etc.
While I realize the empirical evidence is out there to prove that a lot of these strategies can cost runs, with an offense an anemic as this one, can you wait around for big innings? Why not get your one run where you can, and hope that the team’s defense and pitching can come through and hold the lead. I’d much rather see one run scored than leaving runners stranded and/or hitting into double plays, which the M’s have been doing a lot of lately.
I think early in the season, Wak was TOO enamored with the sac bunt and was using it way too much. I just think he’s done a total reversal now and there needs to be a greater balance as a means of creating some runs.
Very nice work! This is the kind of thing that has me checking this site three times per day.
Soooo there’s a lot of work on whether you should be playing for one run when your team can’t score anyway (the obvious example — if your team has 9 Rob Johnsons, you’re going to do a ton of bunting), but the threshold for when it’s a good play is shockingly prohibitive. Even the most anemic major league offense is better off letting everyone hit away until close-and-late situations. Check out the Click/Tango references for more.
Great article Derek.
I know it would be much harder to cull out the data… but I bet these numbers would improve if all of the attempted bunts were examined.
I’m not sure about it, but if you include all the bunt hits, the errors, and the pop-ups and other failed bunts, I imagine the net results would be more positive than what you’ve got now. …And that doesn’t even include the at-bats where the hitter shows bunt to bring in the infield, only to punch it through on the next pitch.
It’s not, as it turns out. Bunts don’t just succeed beyond expectations, they fail a lot: bunt doesn’t get down, runner’s left swinging away 0-2, the bunt’s not good and the lead runner gets thrown out. Again, check out the Click/Tango stuff — I can’t quote the fail/success rate off the top of my head, but the positive effects aren’t so huge they outweigh what happens when it goes wrong.
Mind-blowing work, DMZ. You couldn’t do anything more thorough unless you watched all 30 games from start to finish. Bravo. Here’s my two cents: the next task is to tally up the all Mariners’ bunt attempts and the times they showed bunt and then see the situation and WPA and, if possible, what Wak’s intent might be. This would be perfect for outsourcing – you could ask your readers with MLB.tv to each watch one game (I’m sure there’s somewhere close to thirty people willing to do this over the space of a week or two) and report every last hint of a bunt attempt. You could have a sign-up sheet or something. I’d be happy to watch a game or two that I missed originally in the MLB.tv archives and do that. Again, though, great work.
Also, I like how you told us how long this took to research and write – doing this occasionally is very healthy for the reader. It gives someone like me a much better appreciation of the process, effort, and energy that goes into feeding the content-demanding beast.
Why is there an advantage?
The more times a starter goes through the lineup, the worse he is. The Book has shown this.
So realizing that, you’d want a starter to stay in as long as possible. Once you get the starter out, a reliever will generally go through the lineup one time max, and the pitcher will have the advantage.
There’s the issue of “wearing out” a bullpen, I suppose, but I’d wager any sort of impact would be realized in the *next* series, rather than the current. So you’d be helping another team, not necessarily your own.
I guess the most glaring ones for me are Chavez trying to bunt Ichiro over in the 1st.
1) Ichiro attempting to steal 2nd would probably be a better option. You have something like a 80% chance of Ichiro getting 2nd without an out and then you have a 30-35% chance of Chavez still getting on first even if Ichiro is thrown out. I would think this would be a more optimal situation.
2) The 2nd hitter in your lineup should be good enough that his expected value from hitting would be higher than the expected value from a sac bunt attempt. The fact that Chavez is bad enough with the bat to make having him bunt even be an option is an issue that needs to be addressed during the line-up construction.
Is there good treatise on finer points of WPA that would explain pros/cons of treating extra innings as equivalent to the 9th? (I assume it has something to do with progressive decline in offense/defense/pitching due to use of pinch runners, depletion of bench & bullpen, etc.)
Agreed. I am not a bunt guy either, mostly because they are hard to do well and typically come with a huge ‘I AM GOING TO BUNT’ Wile E. Coyote sign. However, if a team with as much speed as we have bunts properly and in the right situations it can be extremely effective–something Wak seems to know.
What I worry about is this next series and whether we will ‘regress to the mean’ of the last few years and begin our slide back into the basement of the AL West. We lose against Minnesota and we may be in for a pretty serious nose dive based on what we have coming up: Texas (can’t seem to beat them), Boston, followed by LA…ugh.
Heck, if we can bunt our way to beating Minnesota, Texas, Boston and LA, through our brief NL vacation and past Oakland… Great!
I’d be curious to know two things:
1. How fast do you have to be to make a bunt (sac or not) an attempt at a legitmate hit? This may change the equation for Chavez.
2. How poor of a hitter should you be to make a Sac Bunt better than a strike out?
Ichiro, Chavez, Gutierrez, and Yuni should lay one down every once in a while.
They do — at least the first three, maybe not Yuni. What I don’t understand is why guys that get the shift treatment (Jr. for example) don’t practice bunting and do it pretty much every time the shift is on. If you can bunt your way on even half the time the defense will be forced to stop doing it.
Speaking of Junior, in light of the Manny Ramirez situation, in a weird way don’t Ms fans sort of feel good that Junior has been on a steady decline and isn’t very good anymore? That’s what happens to older players who don’t juice.
Junior use to try to bunt against the shift back in the day. He always fouled the pitch off and got himself into an 0-1 count. He stopped doing it. I suppose that why the Juniors and Giambis don’t bunt. They are not that good at it.
I’m not sure on the accuracy of the fangraphs bunt numbers, but they have the Mariners leading MLB in bunt hits (11) with a success rate of 44% (so they have some information that says there have been 25 attempted bunts.) Given your 11 identified sac bunts, that would be:
11 bunt hits
11 sac bunts
3 bunts for outs, no runners advanced
That seems like a pretty good record if it’s accurate. The league averages so far this year are 29% successful in the AL and 22% successful in the NL.
Of course, that doesn’t count the other stuff Derek mentions as bad side-effects of bunt attempts (I hate seeing guys get cornered into an 0-2 count, since hitters are .179/.210/.269 after 0-2 counts), so if the team’s been doing that a lot, that’s something to count against both the bunters for failing to get the bunt down and Wak for giving poor bunters the go-ahead.
I do think that there’s a real advantage for contact hitters drawing in corner infielders early in the count, but I’m not sure just how big it is.
Junior use to try to bunt against the shift back in the day. He always fouled the pitch off and got himself into an 0-1 count. He stopped doing it. I suppose that why the Juniors and Giambis don’t bunt. They are not that good at it.
That’s part of it, certainly. For Giambi, being slow matters, too.
Another thing is that if you have legitimate power, it’s a big trade-off to limit yourself to a single (or an error-assisted “double” at the very best.)
However, it makes more sense for someone like Joe Mauer, who is a pretty good bunter, sees an infield shift occasionally, doesn’t have much power, and also has a pretty significant platoon split. Often after a game where he’s bunted, he’ll mention that he didn’t feel comfortable swinging away against the pitcher (often a left-handed pitcher) and he’s 18-for-28 by fangraphs’ count, which is high enough that you could almost argue that he might not be attempting it often enough.
That seems like a fairly unique case, though, in that most of the guys that get the infield shift have pretty good power and don’t want to give that up. Even without a shift, my intuition is that good bunters without much power should try to take advantage of poor corner infielders more often than they do.
4/14 is the only one I’d really debate on, but that’s just because it’s bad for my blood pressure to see Yuni at the plate with a runner on first and one out.
The chapter on bunts in The Book is excellent and everyone should buy that book and read it.
Previously I had believed (based on reading other sabermetric articles) that bunts were pretty much exclusively bad, except for at the end of tied games.
But Tom Tango describes how while the RESULT of a sacrifice (get an out, move runners over) is nearly always bad, the act of bunting is justified much more often than that, based on looking at all the results of bunts, including errors by the defense and successful bunt hits.
The result is that many times late in games, it is correct to bunt, especially with a lead. (When behind its still generally bad).
Other factors are important beyond the base/out state and the score, such as the speed and bunting skill of the bunter (chance of successful hit, etc), the wOBA of the batter (bunting is better the weaker the hitter is, obviosuly), and also the way the defense is playing.
If you never bunt and the opponents know this, they will play back and turn some hits into outs. You need to bunt enough that you sometimes fool the other team, and dont allow them to play a strategy of always playing deep.
This is like in poker, where you need to bluff enough that your opponents will sometimes call your good hands, thinking they could be a bluff.
Regarding the lack of Mariners bunting recently, I believe ths is due to a combination of not having baserunners, and not having situations where bunting is good (late in close games, especially games where you have a small lead).
The fact that they are not bunting in the situaitons they have had where bunting is a bad idea, is a good thing.
Earlier, Wakamatsu established the mariners as a team that bunts while also being successful, and thats a good thing, because now their opponents have to play in a way that they expect the bunts, which will help the Mariners chances of getting hits.
I suppose that why the Juniors and Giambis don’t bunt. They are not that good at it.
I suspect that’s why as well, but I also suspect bunting the ball hard towards third is a skill that a professional hitter can develop with practice.
Another thing is that if you have legitimate power, it’s a big trade-off to limit yourself to a single
Is it? I’m not so sure if this. It requires further research. Suppose you could bunt your way on 50% of the time, what type of a hitter would you have to be to make it better to swing away? Taking into account that a shift would be in effect. I suspect that you would have to be Bonds-in-his-prime-esque, but I could be wrong. I’ve never seen any articles addressing this.
I’m not a baseball expert like most of the guys on this site, but I have a theory about the amount bunting in the first few weeks.
I think Wak may be bunting more than he normally would in order to test his guys in certain situations and get them used to bunting in real games, kind of an extended “Spring Training” for bunting.
Once he figures out what he has, he will bunt less often and pick his spots more carefully.
Do you think that may be what’s going on?
Any player who could bunt their way on 50% of the time would be better than anyone except probably Bonds at his peak. .500 OBP is amazing even if its all walks.
Of course, the other team would just play way in to prevent it.
I like how many of the bunts turn out to be acceptable because of how bad some of the starting batters are.
Good stuff, Derek.
EXCELLENT article. I learned lots.
I’m curious. I [you’re off topic]
Bloomquist analysis:
He sucks. It’s 50 at-bats.
Happy?
Be careful to read too much into small sample sizes.
EDIT: Dave beat me to it.
Any player who could bunt their way on 50% of the time would be better than anyone except probably Bonds at his peak. .500 OBP is amazing even if its all walks.
Indeed, considering that even bunting all the time, you’d still draw some number of walks (Betancourt-level maybe, but not nothing), and probably get hit by pitches every now and then as well.
Also, if you’re going to bring up Bloomquist out of the blue, at least put in a token effort to make it relate to the topic at hand, it’s not that hard. That is all.
Also, I want to commend DMZ on this post and to join in his recommendation of the relevant chapter in Tango’s “The Book.” Just finished reading that book myself and agree that it had a very interesting and thorough evaluation of bunting strategy and effectiveness.
You guys are just lucky you don’t love NPB! The SB is the real religion of Yakyu-sekai (there is a rival sect, which demands that non-satellite televised games end at 9.28 no matter what the situation). I’d rather they just cut out the still beating heart of a randomly-selected intelligent fan over home plate than watch my team SB in the first three innings.
And even with years of experience and acclimation to the pseudo-philosophy of Yakyu can’t enjoy the spring and summer HS tournaments (the biggest annual sports events in japan, by far), because EVERY team almost literally bunts EVERY time in EVERY goddam first inning, second inning, third inning…
Derek, you’re awesome! This is excellent work – Thanks!!!
I’m also a bit surprised at the findings… maybe I’ll yell at the computer screen less when the guys are bunting tonight.
Then again, maybe not. I’ll probably be yelling at them either way, so whatever.
I don’t know what to do with myself without a game thread. My brain itches.
I think there’s something you can take for that.
I know what you mean, it’s weird. I imagine Derek is in a meeting, or out on his bike, or has forgotten about the whole time zone thing again…
Of course, we could probably get much the same effect by just going back to either of the past two game threads and slowly scrolling through them, substituting “Jakubauskas” for “Washburn” and the names of various Twins players for the Royals. Otherwise, the details of the M’s offensive ineffectiveness hardly matter.
One of the more enjoyable bits of microanalysis I’ve seen. And I do think it’s true, as SeasonTix suggested, that Wak was doing some extra bunting early in the season as a test. A lot of the guys he’s bunted with most were guys who were in the WBC (Kenji, Endy, Lopez), so he wouldn’t have had enough of the spring to see what they could do.