The State of Sabermetrics in the College Game
Wednesday was Media Day at Stanford University. Some of the most successful coaches in the college game were gathered in Palo Alto to introduce their teams and take questions about the upcoming season. While there were little snippets of saber-awareness throughout, the overall feeling was perhaps more old-school than the professional game.
Up first was Mark Marquess, Stanford coach since 1977. Proud of his team, recently named the pre-season number two in the nation, he probably the most sabrermetrically-friendly of the group. First, he reacted to the new ball. After giving the caveat that the new bat was “here to stay,” he pointed out that decreasing offense was risky in terms of attendance and popularity of the game. He then added something that FanGraphs readers might applaud.
Third and fourth hitters in the pros are not bunting. They are in college. Maybe another year of adjustment will change things, but that’s how people reacted to the bat last year and it wasn’t necessarily a good thing. — Mark Marquess
Marquess’ team averaged a bunt every two games last year. Compared to the MLB average of .34 bunts per team per game, it does put him behind the times a tad, but we have to remember that the run environment in college was lower. The average Major League team scored 4.3 runs a game last year, and his Stanford squad played in a Division I environment that saw only 5.58 runs per game. For him to average about as many bunts per game as the Washington Nationals while playing in a league that scores almost half as many runs means he’s ahead of the average college coach in this regard.
Or he just has better position players than the average college team. One scout said that this year’s Stanford squad could boast top-five-round talent at every position. At least he’s not having his studs lay em down.
Marquess also repeated the adage that ‘pitching wins championships’ with respect to how the Stanford squad would do this year. While that may not be completely true — professional teams to win championships with hitting that is better than their pitching — it does seem that it’s virtually impossible to win championships with a below-average pitching staff.
Either way, it’s a good thing they have top prospect Mark Appel and his 96 MPH fastball atop the rotation, and that big lefty Brett Mooneyham‘s in-jeopardy career has been resuscitated. Marquess seemed confident that the team would find a closer among their good secondary pitchers, which also seems saber-aware. Afterwards the coach agreed to a sit down some time to explore this concept further, so we’ll see. Early returns are favorable.
California coach David Esquer piqued saber cats’ ears when he affirmed that his star player, Tony Renda, would bat at the top of the lineup — first or second. The reasoning? “If anyone’s going to get five at-bats in a game, we think it should be him.” Bravo, coach.
Afterwards, though, the coach muddied the waters somewhat. The thinking was the team’s second-best hitter, Chad Krist, would probably hit fourth. But if Renda wasn’t “getting enough pitches to hit” because he wasn’t getting “coverage behind him,” he’d move Krist up to bat second, and move his traditional #1/#2 hitters to the #8/#9 slot. This is not by the book or by The Book, considering his best on-base guys are now at the bottom of the lineup, and also the fact that he’s referenced lineup protection, which has been historically hard to prove.
And bunting? Cal will do it some, he says, as part of a “mixed” approach — he just doesn’t have the boppers of a Stanford lineup, and offense is “tough to come by” in the early-season, misty college games. Esquer put a saber wrinkle into the conversation, though. Responding to my assertion that research has shown that bunts — at least on the Major League level — are usually bad ideas, Esquer pointed out that it’s “about how you use the bunt.” His team will work on hybrid drag-sacrifice and push-sacrifice in order to increasing the chances of a hit.
“The win for us is if the third baseman plays the ball on the run. The chances of him throwing that ball away or wide is a big deal.” — David Esquer
That does change things. In a tight run environment, with worse defenders and an increased likelihood of achieving a hit, the needle moves a little more in the bunt’s favor.
None of these statements, however, were as saber-unfriendly as the main announcement made by Coach Sam Piraro of San Jose State University. The coach, in his 25th season at San Jose State, affirmed that his star infielder/pitcher Zack Jones would close and probably pinch-hit or play some at designated hitter.
“If we have a chance to win three games in a week, I’ll take that,” — Sam Piraro
He admitted that the scouts wanted him to let Jones start, and play in the field, so they could see what he could do. The player probably would agree, too. But Coach Piraro feels that what the team is asking him to do is already “a bear,” so they are going to stick him in the bullpen first and figure the rest out later.
This is a pre-season Western Athletic Conference pitcher of the year, and a preseason all-WAC pitcher/hitter, and his coach is going to make him a closer and part-time DH. That doesn’t seem like the best use of his resources. Fresno State has won five of the last six WAC titles, and this decision probably won’t make it any harder for them to repeat this year.
It was a mixed Media Day for the sabermetrically aware college baseball fan. Perhaps it’s not surprising — gathering statistics for over 280 Division I college baseball teams is not easy, even if the NCAA does attempt to help.
On the other hand, the top five college baseball programs averaged over 250,000 total attendees in 2010, and even Stanford University (which averaged about two thousand per game in 2010) gets a large (monetary, if not direct) benefit from a winning baseball program. The game is basically the same at any level, and there’s money to be made by figuring out the best practices and discovering college baseball’s “extra 2%.”
An important note: Carolna (that is, UNC) uses a really nice pitch/fx camera and computer setup at their games (that they man with players who aren’t playing- usually yesterday’s starter). I wonder what other teams are doing.
No one cares. What does this have to do with sabermetrics?
I care, but mostly because i want to learn more about anyone named mooneybarger and wikipedia isn’t helping. Also, the increased usage of sabermetrics at the college level seems likely to produce more young players who accept sabermetric ideas about bullpen usage, batting order, etc. which in turn might allow traditional ideas about these things at the major league level to be reconsidered.
This was hilarious. Thanks dude. Fixed it… MooneyHAM.
As a college coach, I care what my peers are doing and how they view the game. This is important stuff.
I thought I enjoyed the article, but then Al alerted me to the face that no one cares. Sorry Eno.
The colleges have a lower run environment, so they compensate by bunting. Huh?
yeah, bunting increases the likelihood you can score one run, reduces the likelihood of “big” innings. if it’s true that the run environment is ultra-low, and teams are unlikely to see these big innings, the relative value of a bunt just increased. now factor in that bunt defense is poorer at the college level and can turn some of these “sacrifice” bunts into hits or reaches on error. now there is some real value to a bunt. i don’t see what the problem is here. (though some expected value calculations would be nice.)
EV of a bunt could (*could*) be higher because the fielding’s not nearly as good in college as it is in pro. However, that same fielding discrepancy could mean that you get a lot more people on base just by making contact, wiping out any relative gain bunting would have on swinging away.
And i suppose i was a bit premature to dismiss others’ dismissals of bunting. It may not be the case. I’ll support my points more concisely so they can easily be picked apart:
1) Assuming there is a certain constant with a bunt (a trade of one base for one out), a lower run environment reduces the nominal expected value of swinging away more than it reduces the nominal expected value of a sacrifice bunt. The unknown is the magnitudes of the elasticities- how low of a run environment makes the bunt the sound play?
2) Along the same lines, there may be specific defensive maneuvers associated with the bunt (throwing on the run, knowing the right base to throw to, base coverage) where college kids are especially weak.
These are both hunches. No analysis whatsoever. Not defending them to the death. But as the article states, there hasn’t really been any significant rigorous quantitative analysis done on the college game, so where we’re at is throwing around ideas based on what we do know from MLB sabermetrics, the next step being empirical testing.
USS makes the points that I was assuming, which is great. I probably should have written those up, but there is a difficulty in proving these with the paucity of readily available college statistics in usable form.
No, david, at least in the majors, even assuming that a bunt is successful, it reduces the probability of scoring one run.
However, as Joe points out, that fact may be different in college.
The Stanford coach says that even 3rd and 4th hitters bunt, which I have seen in some of the few college games I watch. That can’t be right at any level.
College to me is interesting becuase [1] the quality of batters at the bottom of the lineup is muc worse than the majors, and [2] the talent level of the picthers is far above the hitters, in part due to #1.
We saw this plainly in the CWS, relievers would come in with guys on base multiple times and the lineup would go down 1-2-3, parituclarly if you played on UNC or against Florida.
Each team has basically 2-3 good hitters and the guys at the bottom of the lineup are essentially automatic outs.
So, a 5 hitter in NCAA might not be as close in talent to the 4 hitter as they would be in MLB.
IMHO, I think applying MLB data and conclusions to the college game is a mistake, especially in regards to IBB, bunting, reliever usage, etc.
It’s true that the run environment and quality of bats makes a big difference. And that the schedule is very different. But I’ll maintain that using your best player as a closer / DH is not the best use of resources, no matter how different the game.
Amen to that, Eno!
Can you imagine a football coach saving his best QB for situations in which his team is leading in close games?
oh god are you talking about shanahan
Rather than make statements, we should be analyzing evidence.
Is it a case of using a star player as a part-timer? Or is it a case of getting one guy to perform 2 roles well?
1. The guy is playing DH to save throws on his arm. I like that. The coach is deliberately not overusing the kid for his own personal gain. He’s keeping his bat in the lineup.
2. He’s using him as a relief ace.
Is basically asking “Is he worth more as relief ace or defensive left fielder?”
Don’t we need to look at: [1] their options in LF, and [2] their options at relief ace?
It’s entirely possible that their option at LF is better than their option at RA, and that by having the star player RA/DH allows the best overall players to play the most.
I suspect that the kid can play LF given that relief aces usually have strong arms. The problem is that relief pitchers in college are used a lot, so having him play LF with the throws between innings, hard throws to bases, etc is actually a detriment to the kid.
I would also posit that NCAA coaches are more likely to use their relief ace in a non-save situation if it a critical moment.
I don’t think the answer is anywhere close to as obvious as you are positing.
Edit your articles, goodness.
dont know if linking to another site is allowed but if anyone is looking for college data collegesplits.com is as the kids say “off the chain”
they don’t say that anymore, I hear “dank” more now. Maybe it’s a midwest thing. Or if you want it to be even more exciting “hella dank” which is some sort of bastardized version of “dank as hell”. I’m not sure how they take a word that basically describes most basements (dank) and turn it into a positive, but they do.
I guess “hell of a lot” is a better way to describe “hella”. I think it’s a dumb word.
I disagree that Stanford is sabermetric friendly. Talk to any person with knowledge about scouting about the “Stanford swing.” It’s designed to eliminate slugging. It is a bunt over the infield.
My feeling is that most coaches want pitchers that strike out a lot of batters and walk very few while not allowing home runs. I’m also going to guess that most coaches wanted these type of pitchers for the last century.
I’d also suggest that unlike their MLB counterparts, college coaches will use their relief ace, wherever, whenever.
I would also go out of a limb and say that coaches strongly prefer and recruit players that hit for extra bases and get on base a lot (I would posit that many coaches respect the “on base any way possible” more than MLB), and most college coaches REALLY value defense if for no other reason it makes coaching a lot easier and gives you a lot of options.
They may not know all of the acronyms and xTERMS, but the sabermetric stats just more accurately measure qualities that have traditionally been associated with success.
We already went through the bunt debate during the CWS and anyone who saw how over-matched the batters were understood that teams were, in fact, playing for 1-run and noted how many RISP the RP left on the bases.
It’s a different environment. Does that mean that in college, walks are bad for batters? No, of course not. But the environment was much more like 1960s MLB than the steroid era (where much of the research on bunting, etc was conducted). I grew up as a huge NCAA baseball fan during the “Gorilla Ball” era. No one bunted during gorilla ball.