Jeff Curry-US PRESSWIRE - Presswire
bunting discussion with actual facts and less drugs. looking at outstanding one-month performances and their (im)permanence
well, i was going to be nice to mike matheny this week on the bunting stuff.
see, i wrote a slightly unfair piece of satire last week, sending up his proclivity for the bunt. i knew it was unfair, but i figured you should never let the truth get in the way of a good story (or even a slightly enjoyable one).
and i was all set to write up a nice half-apology this week, saying that mike matheny isn't a complete disaster on the bunting front. he's mostly just an ordinary disaster.
then i saw this:
Manager Mike Matheny said he intends to bring a "different look to our bunting approach" next year after having so many unsuccessful bunt attempt cost the Cardinals this season.
Matheny has been aggressive in calling for sacrifices all season and has taken criticism for the timing of some of those calls. But that shouldn't release players from shouldering fault when a lack of execution is to blame. And this has been much more an issue than the actual decision to call for a bunt to be laid down.
"I do believe there's room for improvement all across the board," Matheny said. "I think we could be better. I think they think we could be better."
sorry about all the almost-nice things i almost said about you, mike. you are actually a rolling disaster. you made some terrible, terrible, awful, ugly choices about when to bunt, and then you 1) called for more bunting, not less; and 2) you blamed your players for the lack of success in executing your terrible tactics.
let's look at the facts.
so, to start with a common refrain here - and one which matheny apparently subscribes to:
are we, in fact, terrible at sac bunting?
it's really pretty simple, and the numbers are not that hard to find. we are "successful" 72% of the time.
the league average "success" rate on sac bunts: 71%.
i'm not sure that deliberately diminishing your win expectancy by causing an out but advancing a runner can really be called a success in any sense of the word, but "less appalling that causing an out and NOT advancing a runner" takes too long to say. anyway, we are slightly less terrible than the ordinary team at this terribleness.
i don't know that we shouldn't try to improve our position player bunting for those rare instances where it is appropriate. being better than average at sac bunting seems like a reachable goal. but i do think that matheny is off-base in blaming the players' execution, which seems to be fine.
i suppose that since AL teams don't have batting pitchers, there could be some league differences there. that is, NL teams should have better sac bunt "success" since the pitchers get a lot of chances and do a lot of the bunting. on the other hand, position players might actually have better success because sac bunts may be a surprise with a position player at the plate, while a pitcher is practically certain to bunt with a man on base and less than two outs. i'm sure as shit not going through these numbers by hand to count each pitcher sac bunt and each position player sac bunt to illustrate the point.
the best i can do is feed you some quick NL opponent sac bunt numbers. milwaukee's sac bunt success rate: 65%. atlanta: 71%. san diego: 67%. cubs: 65%. houston: 74%. reds: 66%. mets: 80%. pirates: 73%. dodgers: 75%. giants: 76%. colorado: 73%. miami: 65%. washington: 70%. philly: 78%. arizona: 72%. the NL average is . . . 71%. this makes me question whether the "league average" was MLB or the NL. if it's the latter, i just did a lot of useless math and research.
so, yes, we're basically average on bunting success. you might also note that the sample sizes here are TINY. you're talking about 60-120 sac bunt attempts per team, so the variation of success is really a handful of bunt attempts in a year. having 4-5 more successful sac bunt attempts a year seems like a pretty terrible top priority to have for next year's spring training.
so, having answered question 1 in the negative, on to question #2.
is mike matheny terrible at sac bunting because he orders people to do it too much?
well, that's a terrible question. sac bunting with position players is almost always a bad idea. so if the number of position player bunts is more than 10 a year, the answer is probably yes, in the grand scheme of things. (our number is more than 10.)
is mike matheny terrible at sac bunting because he orders people to do it more than other managers?
that's a better question. we have attempted 96 sac bunts and "succeeded" 69 times. here's other teams' sac bunt attempts across 161 games.
that's an average of 89.2 sac bunts per team across 161 games (i don't think last night's numbers were on-line at the time i did this reckoning).
now, there's two important caveats there. one is, as i said, these numbers include pitcher bunts. a team who has their pitchers bunting 50 times and their position players bunting 40 times is a lot better off than a team with its pitchers bunting 30 times and their position players bunting 60 times, even if they both add up to 90. i don't see an easy way to break that out by team. sorry.
the other thing, which is easier to control for, is that sac bunts only will occur when someone is on-base with less than two outs. so two managers with the exact same bunting philosophy might see very different numbers of bunts, depending on how many times bunting-appropriate situations arise.
while on-base percentage isn't a perfect proxy, since one team could systematically have more or fewer baserunners on with 2 outs (or other circumstances where bunts are particularly appropriate or inappropriate), i think it's a pretty good proxy.
the cardinals had a team OBP of .338, easily the best in the NL. the cubs and the astros had the worst, with a team .301 OBP. NL average OBP was .318.
the cubs' lowest number of sac bunt attempts was probably part-dale sweum and part-not having a lot of baserunners. since the cubs only had 95% of the baserunners of the ordinary club, you'd expect them to have 95% of the average sac bunt attempts, or about 84. instead, they had 65.
the cardinals had 96 sac bunt attempts, while one would expect them to have about 6% more bunt attempts based sheerly on the greater number of base runners. that gives me 94 expected sac bunts relative to league average, adjusted for team OBP.
if you want to rake mike matheny over the coals for the two extra sac bunts in a season above what you would expect from a typical NL manager, go ahead. they're bad decisions, no doubt. they take runs away from the team. they're counterproductive. but look at what's happening on other great offensive teams -- cincinnatti, milwaukee, and LA all have significantly more sac bunt attempts than we do. mike matheny's not GOOD on bunting, but there are some far worse offenders out there.
is there anything weird about WHO we're asking to bunt?
yes. looking at our bunt attempts, we see a lot of starting pitchers trying to bunt. that shouldn't surprise anyone; sac bunts from pitchers are usually a good idea. kyle lohse (19), lance lynn (11), jake westbrook (7), adam wainwright (5), jaime garcia (4), and joe kelly (3) comprise more than half of our attempted bunts as a team.
but looking at who's bunting on the position player side makes me shake my head. jon jay has been asked to sac bunt THIRTEEN times. he has a .373 on base average, and he's sac bunting more than anyone on the team except kyle lohse. jay has the NINTH best OBP in the national league, which meanst that he's better than all but 8 players in the league at not making an out. and we're asking him to deliberately make outs to move a runner over. that's completely inexplicable and indefensible.
rafael furcal and daniel descalso have both attempted sac bunts 8 times each. furcal - for as much as the last 2/3 of his season was frustrating to watch - has a .325 OBP. he does not make an out almost a third of the time. it's not as egregious as asking jon jay to sac bunt, but it's still pretty terrible.
the rest of the position player sac bunts are small in number and mostly scattered, with heavy representation of light-hitting utility types: molina (3), chambers (2), greene (4), jackson (1), beltran (1), carpenter (1), kozma (1), and schumaker (3).
even if the numbers are tiny, it's painful to see great hitters like carpenter, molina, and beltran sac bunting. i seem to remember allegations that beltran and at least one of molina's sac bunts were spontaneous. even if carpenter was only asked to sac bunt once, that one call may have been the worst sac bunt call i've ever seen, both in terms of choice of opportunity and outcome.
i have no idea why jon jay is the constant target of sac bunt calls, beyond the simple fact that he often appeared at the plate with rafael furcal on first, due to his spot in the batting order. sac bunting with him is a terrible game plan, which probably cost the cardinals several runs over the course of the season. sure, a successful bunt puts furcal on second, but you're stacking matt holliday, allen craig, and carlos beltran behind those two guys for a reason: so you can drive them in. you're better off, except in a handful of exceptional circumstances, trying to get two guys on to score two runs and keep the rally going. jay is going to deliver in almost 2 of every 5 chances. you have to trust him.
shorter version: stop bunting, especially with jon jay!
here's three player-months from recent cardinals history.
player A - 67 PA, 17H, 5 2B, 3 3B, 2HR, 7BB, .343/.395/.597, .420 BABIP, ISO .254, wOBA .416
player B - 44 PA, 15H, 1 2B, 3 HR, 3BB, .385/.432/.641, .387 BABIP, ISO .256, wOBA
player C - 56 PA, 23H, 3 2B, 2 3B, 1HR, .426/.446/.611, .512 BABIP, ISO .185, wOBA .454
i didn't make any of these terribly hard, and i'm sure some of you can guess both who they are and why they're here.
player A is pete kozma, september/october 2012.
player B is brian barden, april 2009
player C is bo hart, june 2003.
brian barden went on to hit for a .192 wOBA in may 2008, and got 12 PAs for the rest of the season. he has a .252 wOBA as a major leaguer and didn't play pro ball at any level this season.
bo hart hit tepidly in the next three months of 2003 (.257, .211, .282), made a brief terrible appearance in 2004, and struggled in AAA for the next couple seasons before retiring.
so, if you're looking at kozma's numbers across september and assuming that he must have figured something out or learned something, remember these brief fizzles are pretty common, just as the converse is pretty common. allen craig's .304 wOBA across his first 124 PAs in 2010 didn't let on that an outstanding hitter had just come up.
i can understand being tired of hearing about small sample sizes and BABIP, because they are kind of tedious in their own way. but they're tedious because it's so much more fun to give in to a one-month burst of talent (i know, because i remember being very excited about barden). they're tedious because they are so fundamental. people who don't learn from one-month high-BABIP, small sample size crushes are doomed to repeat them.