June 20, 2013

Skewed Left

Preparing for Baseball's Most Extreme Circumstances

by Zachary Levine

the archives are now free. All Baseball Prospectus Premium and Fantasy articles more than a year old are now free as a thank you to the entire Internet for making our work possible. Not a subscriber? Get exclusive content like this delivered hot to your inbox every weekday. Click here for more information on Baseball Prospectus subscriptions or use the buttons to the right to subscribe and get instant access to the best baseball content on the web.

Subscribe for $4.95 per month Recurring subscription - cancel anytime. Subscribe for $39.95 for one year a 33% savings over the monthly price! Purchase a $39.95 gift subscription a 33% savings over the monthly price! Already a subscriber? Click here and use the blue login bar to log in.

Disaster experts—and really, what better analogy is there for a discussion of major league managers—often use terms like “100-year event” or “1000-year event” to discuss the limits of our preparedness.

There are some situations that are so rare that in a cost-benefit analysis it is deemed not worthwhile to disrupt routines or expend resources in order to prevent them. For example, the United States has not stationed a large military force in International Falls because while there is a non-zero chance, there is a very small chance that Canada will attack us in a ground invasion from Ontario. The parlance is more commonly used for things like floods and earthquakes.

This month in baseball has seen several situations in which managers found themselves somewhat unprepared as games progressed to 16, 18, or even 20 innings. Having talked to a couple of the managers involved, it turns out that this is no surprise.

This month’s Marlins-Mets 20-inning game, Jays-Rangers and A’s-Yankees 18-inning games, and White Sox-Mariners 16-inning game are baseball’s 100-year floods. They are situations that would take a lot of strain to prepare for, in that you’d have to pass up platoon advantages, pinch-hitting spots, and other ways of trying to win games in nine innings in order to improve your chances in the unlikely event of an umpteenth inning.

So does a manager ever go in with a plan even in the back of his mind for what happens if that night’s game goes 18 innings?

“I don’t think you ever think about anything more than 10, 11 innings,” said Cubs manager Dale Sveum, who found himself on the winning end against the Reds, who ran out of pitchers quickly and refused to use Aroldis Chapman in a tie game on the road. (Russell Carleton wrote last week about closer usage in extra innings.)

“You’re trying to win the game, you’re not thinking about that stuff—you deal with it when it happens,” Sveum added. “You’re trying to win the game in nine innings.”

Terry Collins, who wound up having to use Shaun Marcum for eight innings in a 20-inning loss to the Marlins on June 8, said that a manager doesn’t really start thinking about the hyper-long relief plan until the tie game creeps really close to extras.

“It was about the seventh inning when you start to look and see who’s left in the bullpen, and that’s when Dan and I started talking about ‘Hey look, we’re going to need someone to give us three innings here,” Collins said. “That’s how Shaun Marcum’s name got brought up. If we have to use him for x amount of innings, we can always push (his next start) back, and that’s how that happened.

“I don’t think anybody prepares to go 18 or 20 innings, I can tell you that.”

Days, or nights, or days turned into nights like Marcum’s were exceptionally rare until a couple of weeks ago. In fact, of the four longest extra-innings relief outings since 1993, three of them happened on the same day earlier this month. The other belongs, not surprisingly, to one of the prominent knuckleballers of the last two decades.

Pitcher	Game	Innings	Result
Shaun Marcum	6/8/13 NYM-MIA	8.0 (13th-20th)	Loss
Steve Sparks	6/1/03 DET-NYY	7.2 (10th-17th)	Loss
Kevin Slowey	6/8/13 MIA-NYM	7.0 (​13th-19th)	Win, replaced for save
Ross Wolf	6/8/13 TEX-TOR	6.2 (12th-18th)	Loss

Source: Baseball-Reference.com Play Index

When a team gets itself into a situation like that, it’s often because of thinking like Sveum’s or Collins’, which isn’t wrong. It’s hard and probably counterproductive even to predict extra innings when the game is tied. It’s not until a game is tied after the eighth inning that the probability of going to extras crosses 50 percent. According to research from our sabermagician Colin Wyers, who suggested the aforementioned cutoff date of 1993 to give us a representation of something close to the current home run environment, only 22.9 percent of games that are tied after six full innings go extras. That increases to 33.8 percent of the games that are tied after seven and 55.5 percent of games that are tied after eight.

But then a funny thing happens.

I went into this wondering if the percentage of extra-inning games that reached a certain inning to end in that inning would go up or down as the innings got later. I could envision the case for more games ending in a given inning as the inning increased, since either pitchers were tiring or teams were down to their last option with a position player being the extreme. However, I could also envision a lower percentage of games to reach a certain inning actually ending as the innings got later if hitters were just going for home runs.

It turns out it’s almost exactly the same for each of the next four innings in extras.

When a game has gone to the 10th, it has gone to the 11th 53.0 percent of the time.
When a game has gone to the 11th, it has gone to the 12th 53.4 percent of the time.
When a game has gone to the 12th, it has gone to the 13th 52.0 percent of the time.
When a game has gone to the 13th, it has gone to the 14th 52.4 percent of the time.

Over the last 20 years, there have been 316 games that have gone to the 14th inning, and that’s where the percentages to reach the next inning start fluctuating—though not strictly increasing or decreasing—as sample sizes get really small.

From 1993 through Tuesday, there were 47,742 games that weren’t shortened by weather. Slightly over 8.5 percent of them have gone into extra innings, and the numbers go down fairly quickly from there.

Innings	Games exactly this long	Games at least this long	% of all 47,742 games that went at least this long
9	43647	47742	100%
10	1925	4095	8.577%
11	1011	2170	4.545%
12	556	1159	2.428%
13	287	603	1.263%
14	166	316	0.662%
15	74	150	0.314%
16	31	76	0.159%
17	18	45	0.094%
18	15	27	0.057%
19	6	12	0.025%
20	4	6	0.013%
21	0	2	0.004%
22	2	2	0.004%

As the chart shows, the chance of ever getting to a certain point in a baseball game dips below one percent in the 14th inning, and to Sveum’s point, we can see why a manager never plans for 18 innings. One game in 1768 will get there, making it an 11-year event for a given team. A 20-inning game like the Mets and Marlins played on June 8, despite the fact that the Mets just played one three years ago, is a 49-year event for a team in the current run environment.

Those two 20-inning games were the extremes of perceived preparedness, one with reasonably well-rested starters going eight and seven innings in relief and the other with the teams falling all over themselves in the final innings and two position players pitching.

Yet no matter how nice it looks in the box score, you’re never really prepared for that.

Zachary Levine is an author of Baseball Prospectus. Follow @zacharylevine
Click here to see Zachary's other articles. You can contact Zachary by clicking here