Those lazy, hazy, crazy days of summer seem so far away. It’s been more than two months since the end of the baseball season and the most interesting thing to talk about in January is Hall of Fame voting. Pitchers and catchers don’t report for another month. We have reached the hungry season where even rumors of what some utility infielder might do next year counts as news. Baseball fandom is a cycle of saturation and deprivation.

During the season, it’s easier to remember just how exhausting the game can be, both for the fans and the players. Three daily hours of intense concentration and physical activity and then a plane to catch to across the country for more. But now that the days aren’t filled with worries about whether the right fielder needs a day off to clear his head, it’s a good time to prepare for what can be done about The Grind during the upcoming season.

I've written before about The Grind and how we can see its effect on the players as the season progresses. The effects aren’t always obvious to the naked eye, or even the well-tuned spreadsheet, but they are real. Previously, I’ve found that one way that The Grind creeps into the game is through a slight deterioration in plate discipline as the season progresses. The effect is on the order of perhaps a tenth of a percent of pitches turned into extra strikes, but over the course of a season a team sees a lot of pitches. The team that can hold the line on its players not falling prey to The Grind could save itself a lot of value.

And so to that end, I began looking at whether managers seemed to hold any sway over the silent predator. It turned out that the answer was yes. Some managers seemed to have players who lost their plate discipline more slowly (or even improved a tiny bit) and some saw their players slip away more quickly than we might even expect.

Today, I’d like to do something a little experimental. I want to glance further into the looking glass and see what else we might find about The Grind and its antidote, the manager.

Warning! Gory Mathematical Details Ahead!

I began with a question. If we see the effects of The Grind on plate discipline, where else might we see them? If you’ll forgive the neuropsychological detour for a moment, I reasoned that what is commonly called The Grind is actually a bit of neurological burnout in the brain. Synapses and neurons that fire a lot get tired. They need resting time. This is normally why people sleep. The brain repairs itself while you sleep, but baseball is a game where sleep is at a premium. Sometimes, there isn’t enough time to do the sort of neurological maintenance that needs to be done. But more than that, stress can interfere with that maintenance (or cause more burnout to begin with) as well. While teams can do their best to encourage players to get good sleep, there’s only so much that you can do when the plane lands at 3 am.

But there is something that a team can do about reducing the amount of stress players face. Long pooh poohed by sabermetricians, there now seems to be an understanding that squishy concepts like team chemistry and clubhouse atmosphere can have a real effect, although we are long way away from proving it or understanding how it works. And perhaps more importantly, whether it can actually impact what happens on the field. Using the manager as a proxy for all of those variables isn’t entirely waterproof, but it perhaps rates as “interesting evidence.”

The reason that I initially studied plate discipline as my marker for The Grind was that it represents a higher order skill (neurologically speaking). Recognizing a pitch, figuring out where it’s going to go, and deciding on the correct course of action (in four-tenths of a second!) is something that’s done in an area of the brain known as the pre-frontal cortex. The PFC is the part of your brain right behind your forehead and it controls all of the things that humans are so proud of themselves for being able to do. Complex planning, social cognition, pattern recognition, and the ability to construct a good meme are all housed up there. But the PFC is also an area of the brain that would be very sensitive to The Grind. Ever try to plan something when you were tired?

So, if The Grind affects executive functions like pitch recognition and response, what else might it affect? What about a more basic skill, like reaction time? There isn’t much baseball reaction time data out there publicly. Statcast offers the promise of measures like “first step” for fielders, but thaT data IS not available at this time. Instead, we need a proxy. Thankfully, we have a decent one: Grounders.

Fielding a grounder takes range and footwork, but it’s also a function of simple reaction time. In general, humans react to a stimulus in about 200 milliseconds, but it can vary (especially when someone is tireD) by a few hundred milliseconds. That’s “blink of an eye” territory, but for a ball that will only spend a second or two in its path on the infield, that’s a fairly large chunk of the allotted time for intercept. And we see over the course of a season that infielders get to fewer ground balls.

Here are data from 2012-2016 on ground balls that are fielded by an infielder (whether he turned the ball into an out or not), league-wide, by month.


Percent fielded by an infielder













The pattern above is fairly obvious, but to formalize that I constructed a “number of days since Opening Day” variable for the fielding team as a proxy for The Grind. (I used a similar method in my previous work.) I put that into a logistic regression and it came out as a significant predictor of whether the infielder would get to the ball. Is it possible that over time, players are a little less sharp as The Grind wears on them and that shows up in reaction time?

It would be hard to notice that teams are getting to three-quarters of a percent fewer ground balls than they were three months ago, since that’s one ball in 133, but from April to July we see that big of a shift in the rate at which fielders are getting to worm burners. The average team faces about 1,400 non-bunt grounders over the course of a season. If a team played three-quarters of a percent worse over the course of a season, that’s 10.5 extra grounders that get through that otherwise might have been fielded. Each one of those is worth roughly half a run, so we’re talking about five runs or so. If a manager (or someone) could keep a team from experiencing that sort of summer swoon, he would add half a win to his team’s tally.

(Side bar: I’m not completely convinced that all of that deterioration in results is the result of The Grind. It’s entirely possible that some of that is little injuries popping up that don’t knock an infielder out of the lineup, but do hobble him just the tiniest bit and decrease his range. Those kinds of injuries will happen over the course of a season, but I don’t know that there’s a whole lot the manager can do about it. I think that’s a caveat we need to keep in mind as we interpret these findings.)

So, let’s modify that logistic regression above. To control for the fact that some teams (and managers) are blessed with an infield that vacuums up grounders, I used the seasonal ground ball “get to” rate (just that the fielder got there, not that he completed the out) for the suite of infielders (first to third) that were out there at the time, as well as the batter’s seasonal “avoid the infielders” rate on his ground balls for the season. This creates a control variable for our baseline expectation of how likely it is that a specific ground ball will be fielded.

I asked my computer to draw a line for how each team (and manager) fielded ground balls as the season went on (with days since Opening Day as the unit of measure). I used data from 2012-2016. It produced a line that showed how each manager’s charges progressed as the season went along.

At first, I did a line for each individual season for each manager for which I had enough data. The way that the regressions are constructed, on Opening Day, the regressions would say that everyone is at the same point in being able to field a ground ball, but over the course of the season the identity of the manager starts to make a difference. I looked at where everyone was at 90 days in each regression and predicted the chances that an “average” ground ball would have been gotten to at that point. Depending on the manager, I had five data points for some guys, but one for others. I focused on the managers who had at least four seasons in which they managed.

To see how reliable these estimates are of how “good” a manager is at preventing The Grind from letting extra ground balls into left field, I used a method known as intra-class correlation (for the initiated, I used the AR(1) or auto-regressive first-order type). For those not familiar with the technique, it’s like being able to do a year-to-year correlation, but incorporating more than two years’ worth of data. You can read the result like a more standard-issue correlation. In this case, it came out at .40. This suggests that there’s moderate year-to-year correlation. Normally when we say this sort of a correlation, we say, “Well, one year of data isn’t a great barometer of his true talent.” And that’s true. It’s also not a horrible one, but I’d take one-year reads on this stat with a grain of salt.

But .40 is also the sort of correlation that says “there’s at least some signal in here” and maybe with a two- or three-year window, we can get a better idea of a guy. In player evaluation, that can sometimes be useless. By the time we get a good read on a player, his body has aged enough that we have good reason to believe that he is not that player any more. We have less fear of this with managers.

So instead, I used all of the data from 2012-2016 that was available to look at which managers were the best overall, though with a minimum of three out of those five years in which they were managing. Here’s a top five and a bottom five:

Top 5 Managers

Estimated GB fielding runs saved per year

Bottom 5 Managers

Estimated GB fielding runs “saved” per year

Mike Scioscia


Bud Black


Kirk Gibson


Ron Washington


Ryne Sandberg


John Farrell


Terry Collins


Joe Maddon


John Gibbons


Clint Hurdle



I don’t know that I buy “Bud Black is two wins below average.” In fact, when I ran models looking at plate discipline as an indicator of the effects of The Grind, Black was the best in the business. (I re-ran the original model with 2012-2016 data, and that’s still true.) Maybe the Padres under Black were constantly battling injuries later in the season and that’s what gave him a drooping line downward? Maybe it’s just a marker for managing older players or players who are prone to physically slowing down over the course of a season.

When I ran a correlation between the plate discipline-based metrics and the grounder-based model, the correlations barely hit .10. So that suggests that the effects of ground-ball grind and plate-discipline grind are very different from one another. Maybe this just isn’t as fruitful a line of investigation as I figured it might be.

Do I Have To Go Get That?

This is usually the part of the article where I tie all of the numerical findings together, but I’m left scratching my head on this one. It’s possible that grounder grind is its own thing. Maybe it’s measuring more of the physical impacts of a season that would rob a fielder of his range, while plate-discipline grind is about the mental part. Maybe when more Statcast data become available, we can separate out the effects of the long season on range and reaction time.

I think it’s worth looking at the spread between the top and bottom of that list. That’s a big range, and if a team can find a way to capture even a little bit of that magic it can be a very valuable thing. Maybe this isn’t even The Grind, but some other thing that a team should be paying attention to. I’d love to tie this one up in a bow for everyone, but I’m walking away from this one with more questions than answers.

You need to be logged in to comment. Login or Subscribe
Sciocia's shortstop for much of that period has a combined FRAA of -40. And the third basemen weren't much better. So does this mean Erick Aybar was just really really bad, but he was very resistant to The Grind so he didn't suffer a drop-off from fatigure during the year? Also, you have "John Gibbobs" in there. If you need the typo at least add a "o" so it's childishly funny.
Yeah, this controls for baseline talent. You can be overall amazing, but the slope points downward as the season goes on or you can be terrible, but at least maintain that level of terrible throughout the season. The point is that the power to adjust that slope has value as well as the mean.
And then in 2016, Andrelton Simmons missed a bunch of the middle of the season and then returned for the end, which I would expect would further lift Scioscia's numbers.
Not necessarily, because the control variable/base expectation is fitted off of units. The regression knows that Andrelton Simmons is out there and adjusts accordingly.
Interesting that Hurdle is on the "bad" list, given that the Pirates have been trying to follow the Golden State Warriors model of giving their players more rest.
I wonder how weather might play a factor. Is exit velocity correlated with temperature and how much would the playing surface factor in. I'd assume looking at only games played in a dome would answer the impact of the grind (minus injuries)
It goes downward month-to-month except July into July, and the greatest drop is from May into June. I wouldn't think that indicates weather is impacting it.
You have "July into July" in there. If you need the typo at least add an "r" so it's adultishly funny.
Gawd. It's the worst typo since John Podesta's IT guy said a phishing email is legitimate.
Could a lazy fielder maintain a more consistent level of defense over the course of a season, while a max effort guy (who gets to extra balls when he's physically and mentally 100%) fall off as The Grind takes it's toll?