When the Washington Nationals played the Milwaukee Brewers last week and Edwin Jackson needed one more out to escape a jam, it looked like he might have found it: Ryan Braun grounded up the middle, and Daniel Murphy slid, tumbled, turned and flipped the ball to Stephen Drew for the attempted force out at second base.
But while the call on the field was that the runner was safe, the replays on the broadcast were oh, so close.
Looked to me like the throw beat the runner, but I'm just one guy. When the Nationals sent the call to New York for a review, the -- wait, the Nationals didn't send the call to New York for a review. A run scored on the play. Travis Shaw followed it with a home run for Milwaukee. And instead of being down 1-0 going into the bottom of the fourth, Washington trailed 5-0.
We probably shouldn't get upset at managers very often. Managing is hard. It happens fast. The manager has far more information about the situation than we do, and more incentive to get it right than we do, so a healthy amount of deference is due.
But it kills me to see a manager waste a replay opportunity. Replay is a miracle of living in the future. It's a hard-won political victory. It's the cure for the cruelest baseball affliction: a botched call.
And, for a manager, using it is practically free. Most managerial decisions come with a cost: A sacrifice costs an out, a pitchout costs a ball, a pitching change costs a pitcher -- but challenging a call is almost free. Watching Dusty Baker walk right past it without bothering to pick it up made me wonder: Are managers making the most of the replay?
Baseball's replay-review era began in 2014. A few months before the first overturned call -- on, coincidentally, a Ryan Braun infield single -- Dan Brooks and Russell Carleton had warned at Baseball Prospectus that smart managing might lead to an almost absurd number of game-delaying challenges because of, paradoxically, how few overturned calls there would be.
The logic was that the cost for an unsuccessful challenge was tiny. All a manager lost was his right to make another challenge, an inconvenience that itself would last only through the sixth inning (since changed to the seventh), after which time umpires could instigate unlimited reviews. The evidence suggested that there are actually not that many incorrect calls, so a manager who lost his challenge was unlikely to miss it before the seventh-inning rules kicked in. Hence: Managers should spend their reviews liberally:
This may seem counterintuitive, but managers should be losing challenges. A lot of them. So many of them, in fact, that the best managers in terms of maximizing "run production" gained from challenges will almost certainly be the worst managers in terms of challenges won percentage. It's kind of like that old adage, "You miss 100 percent of the shots you never take." Well, since there's essentially no cost for missing, any time managers see a challenge opportunity, they should take it.
-- Dan Brooks and Russell Carleton, Baseball Prospectus
We are now four years into expanded replay reviews. We've seen baseball's rules adapt to the technology; on transfer plays, and someday perhaps on what qualifies as possession of a base. We've seen baseball players adapt to it, with harder slides and more physical tags. But have we seen managers adapt?
It's almost undeniable -- though not quite, as I'll get to -- that managers have been too cautious with their reviews. We can see this in how willing they are to be wrong.
In the first inning of regular-season games, 70 percent of manager-requested challenges successfully overturn a call, according to replay records at Baseball Savant. Managers, in other words, collectively refuse to risk their challenge unless they're 70 percent sure they'll win the review. Unsurprisingly, they get riskier the closer they get to the seventh inning:
1st: 71 percent
2nd: 65 percent
3rd: 67 percent
4th: 58 percent
5th: 55 percent
6th: 50 percent
7th: 47 percent
8th: 40 percent
This is clear evidence that managers are doing cost-benefit analyses here. The lower the cost, the more likely they are to instigate a doomed review, supporting the fundamental logic of Brooks and Carleton's smoke-'em-if-you-got-'em argument. But it's also strong evidence that they're exaggerating the costs, especially early in the game. A manager gets a call overturned in the first seven innings in only about 9 percent of games. The other 91 percent of the time, there is no practical cost to losing a review even on the first batter of the game. As Brooks and Carleton hypothesized, there aren't enough overturnable calls to make the loss of a challenge very punitive. A manager might justify 10 extra early reviews if it got him just one extra overturned call.
Maybe the Nationals wouldn't have gotten that call overturned against the Brewers; who knows what angle they saw in their video room. But getting it overturned would have been worth, on average, 1.4 runs, according to traditional run-expectancy tables. (As it turned out, it ended up costing them four runs.) On the other hand, wasting a challenge in the fourth inning costs, on average, about 1/30th of an expected run. (A 5 percent chance of overturning a call worth about 0.7 runs.) They only have to be right every 30 times to justify it in that specific situation.
Managers, or the staffers they have in video rooms, are increasingly accepting this. Before Major League Baseball made it harder to ask for replays by instituting stricter time limits this year, the frequency of challenges by managers (with the exception of umpire-instigated challenges) was steadily going up, from one every 2.3 games in 2014 to one every 1.8 games last year. This might come at the cost of some lost challenges -- this year's success rate on manager-requested challenges will be the lowest ever, at 51 percent -- but the cost of a bad challenge is pennies compared to the gain of a good one. The raw count of calls overturned goes ever up.
But there's something else odd in the numbers. As established earlier, managers appear to be risk-averse in early innings, while casting a much wider net in middle and later innings. Yet the raw number of calls overturned is actually higher in early innings. There have been 305 overturned calls instigated by managers in the first inning, the most of any inning, but only 267 in the fifth, 241 in the sixth and 266 is the seventh.
Part of that is there's more offense in the first inning, when the top of the order bats, and part of it might be managers passing up later-game chances to challenge when they're down 7-1. But if there were a lot of overturnable calls out there, waiting to be caught by a wider net, we would expect to see more calls overturned in the fifth, sixth and seventh. If a wider net doesn't catch them ... well, you can't catch something that doesn't exist.
If you watch baseball a lot, you'll see a handful of calls this year that a manager could have challenged, that he probably should have challenged and that you might be screaming at the TV for him to challenge. Those plays exist, they can change a game (as Edwin Jackson knows) and a manager might be mathematically justified asking for 25 or 50 more umpire reviews just to pick up one or two of those calls a year. But the evidence suggests that managers are actually doing a pretty good job maximizing their challenges already, without turning things into slowed-down farce.