An Extra Inning Runner Study

The 2020 season brought unprecedented rule changes, one of the most puzzling among them being the “extra inning runner.” Ostensibly in an effort to reduce the spread of COVID-19 and speed up play, commissioner Rob Manfred decreed that once a game progresses past the ninth inning, a runner would be placed at second base to begin the frame.

Manfred’s blatantly obvious motives turned baseball fans — a demographic notorious for their acceptance of changes to the national pastime — against it. If there is any defense to be made for the addition of the extraneous runner, it’s that shorter games helped save pitchers’ arms in what’s already been an utterly brutal season for pitcher injuries.

This seismic rule change also created a correspondingly large shift in how teams strategized after a game surpassed nine innings. Teams, even the more sabermetrically inclined among them, began to employ traditional tactics. In order to determine how clubs played with a free runner, I charted every extra inning of the 2020 season.

Laying One Down

The main factor I chose to isolate was the prevalence of sacrifice bunting. Just 15 years ago, when most teams were presented with a runner on second and no outs, they would attempt a bunt without hesitation. But most teams didn’t go that route in 2020.

Front offices were probably all over the run expectancy matrices from last year, which showed a 13% decrease in RE from “runner on second, no out” to “runner on third, one out.” Still, out of 224 extra half-innings, 31 sacrifice bunts were laid down, almost a 14% rate.

So how did the traditionalists do? The Orioles, Marlins, Phillies, and Blue Jays all laid down three or more sacs. In all, 16 teams laid down at least one. And they blew away the competition.

Run expectancy in innings that included a sac bunt was 1.39. Non-sac bunting innings produced 0.84 runs on average. Sac bunting increased RE by 40%! Additionally, sac bunt innings had a 77.4% chance of putting across a run, up from 55.5% in non-sac innings. The skeptics among us, myself included, would say “that’s a small sample,” and “no freaking way.”

A sample of 31 is indeed considered small, especially in baseball. But a 95% confidence interval analysis helps put some doubts to rest. With near certainty, we can predict that the true RE of innings with a sac bunt is between 0.926 and 1.854 runs. That’s quite a large range, but it gives us valuable information about the true, larger sample RE. Extra-inning sac bunting is almost certain to produce higher RE than the alternative, as long as the 2020 rules are in place.

The Quirks

What else happened in extras other than hyper-effective bunting?

For starts, who was the worst pitcher? Since the automatic runner didn’t count as an earned run, I resorted to more crude metrics. Astros hurler Cy Sneed, who sounds like he belongs on the New York Highlanders, allowed runs in four of his five extra-inning appearances. Other strugglers included Kenley Jansen (2/3), Chad Green (2/3), and Scott Alexander (2/3). Going by pure ERA, Shun Yamaguchi’s tough rookie year wasn’t helped by a 36.00 mark in a pair of extra-inning outings.

There were five historic leadoff two-run homers, the first coming from Edwin Ríos (off Sneed). The second was even cooler, an inside-the-park job from Austin Hays.

And with one electrifying swing of the bat on September 3rd, Pete Alonso walked off the Yankees with a leadoff two-run dinger. Freddie Freeman had a similar walk-off jack against the Red Sox on the 25th.

Possibly more impressive than that brand new type of homer were no-run, two-batter innings. There were also five such innings, one of them from the September 3rd contest between the Yankees and Mets. In the top of the 10th, Tyler Wade ran at second and went flying at the crack of a DJ LeMahieu liner to right. But Michael Conforto came jogging in and made a nonchalant chest-high catch, easily doubling off Wade, who was already around third. Then Luke Voit struck out. The end.

Strikeouts of the first three batters only happened four times, and all three were walked zero times. Yuli Gurriel led the league with nine extra-inning ABs, going for 2-for-9. A 1.000 average was compiled by 12 players, led by the Dominican duo of Jean Segura and Willy Adames, who added 4.000 SLGs.

The league as a whole hit .237/.351/.382 in extras. Focusing on that .351 OBP, it was almost 30 points higher than the league average. This corroborates my suspicion that managers used extras as a playground for traditionalist ideals. They issued 38 extra-inning intentional walks in 787 PAs (20.71 PA/IBB), way up from 84 in 3,861 PAs (45.96 PA/IBB) in 2019. The next highest ratio from a 2020 inning was 156.97 PA/IBB in the 9th.

The Manfredifesto

All these kooky changes in playstyle are because of Rob Manfred’s desire to shorten games and increase scoring. Face it: he was in charge during the juicing of the baseball, he has near-unilateral power when it comes to rule tweaks, and the success of the 2020 season was propped up by pure luck. My goodness, when the NFL has significantly better health and safety policies than your league, you’ve gone hopelessly awry. But I digress.

I delight in reporting that the extra-inning runner rule didn’t decrease game time to a statistically significant degree. Manfred’s three batter minimum rule, while a decent strategy twist, only slowed the upward trend of pitching changes per game. Runs per game decreased from 4.83 in 2019 to 4.65, not because of any rule change per se. Not to mention attendance dropped nearly 70,000,000%, but I don’t delight in that.

Assuming that the next major league season will be COVID-19 free, should the extra inning rule remain in place? Judging by its popularity and purpose, the answer is no. MLB keeps trying to become the NBA, in its pace, playoff structure, and even player exploitation (we may have maybe surpassed them there). But baseball fans just won’t go for it.

If you’d like to see my full research, check out the comprehensive extra inning spreadsheet here! Highlighted innings denote a particularly interesting occurrence, although I may have missed a few.