Bases Loaded, Nobody Out — What Happens Next?

The Giants, somewhat surprisingly, had the biggest positive difference in bases-loaded, no-out situations in 2019. (via Ian D’Andrea)

It starts with a walk, a single, or sometimes a defensive miscue. Then another, and another. Fans inch a little closer to the edge of their seats. They’re excited — but how excited should they be, really?

Across baseball in 2019, 546 innings featured at least one event (a plate appearance or baserunning play like a wild pitch or runner thrown out advancing) with the bases loaded and no one out. That’s about once every 80 innings of regular season play, or one time every nine regulation games. Of course, it happens more often for some teams than others. Today, let’s investigate who these teams were in 2019 — and whether their facility for driving in runs with the bases loaded and no one out really proved that advantageous in the end.

Finding opportunities

The Angels led all teams with 42 such plate appearances in 2019, spread over 29 innings (several occasions featured more than one such PA in the same inning). They scored 68 runs after loading the bases with none out, which represented about 9% of their season total.

On the other end of the spectrum, the Padres very rarely found themselves in this situation. They sent just eight batters to the plate with three on and none out in 2019, three times each in May and June but just two other times in their 109 contests outside those months. Mariners pitcher Marco Gonzalez, who led all pitchers with 12 batters faced in bases-loaded, no-out situations, saw half again as many batters by himself in those situations as the Padres faced in a full season.

From the batting side, the leaders and laggards in terms of backing pitchers into the ultimate jam are as follows.

On the pitching side, the disparity between the haves and have-nots is not quite so great. Two teams allowed opposing teams to load the bases with no outs in 25 innings, and they’re not exactly unlikely candidates: The Royals and Rangers had the worst (.348) and fourth-worst (.342) on-base percentage against in baseball, and some of those baserunners happened to load the bases. They each faced 35 batters in this situation.

The other end of the spectrum, however, is a bit more surprising. The Giants were only a middle-of-the-pack team in 2019 in terms of run prevention, but they limited opponents to just ten innings with a bases-loaded, no-out situation. Some of their neighbors on this list are even more unlikely: The Tigers and Marlins each limited opponents to 12 such innings. 

Again, the top and bottom five.

Putting the two previous tables together creates an interesting, if somewhat noisy, final product. Here are the top and bottom teams in baseball in maximizing their own bases-loaded, no-out opportunities and minimizing their opponents’.

As shown above, only a few teams fell outside a pretty tight group here: 20 of the 30 teams had a differential between +5 and -5. And while reaching the top of this chart doesn’t seem to guarantee overall team quality, there is a correlation between a low rank and a poor season: Just one of the bottom 10 teams in differential (the Twins, discussed in greater detail below) reached the 2019 postseason.

Bringing Runs Home

Several years ago, Tom Tango of Tangotiger.net shared data on run expectancy by base/out situation across recent major league history and found an average bases-loaded, no-out situation resulted in about 2.4 runs scoring over a sample from 1993-2009 but dropped to just under 2.3 runs from 2010-15. (Run scoring across baseball dipped under 4.3 runs per game in that window after peaking over 5.0 in 1999 and 2000.)

Run scoring is on the rise again in recent years, reaching 4.83 runs per game in 2019 for the first time in over a decade. That didn’t significantly change the run expectation in bases-loaded, no-out situations, however. Major league teams scored 1,258 runs after loading the bases with none out in 546 such innings in 2019, an average of 2.304 per inning that is only slightly more than a hundredth of a run more than the aforementioned 2010-15 calculations.

That number is depressed, albeit only slightly, by walk-offs: 15 innings that included a bases-loaded, no-out situation in 2019 ended with a game-ending run-scoring event instead of a third out. Removing them from the equation only raises the expectation by a few hundredths, to 2.341.

The average also is distorted a bit by a few outlying innings. Twice last season, for example, a team loaded the bases with none out and went on to score ten runs:

• The Phillies did it in the bottom of the first of this game against the Mets on April 16.
• On July 19, the Cardinals entered the top of the sixth inning trailing 7-0 and put up a 10-spot to take a 10-7 lead against the Reds in a game they eventually won 12-11.

All told, a team that loaded the bases with no one out scored at least one run 86% of the time, but exactly one run was the most likely outcome, happening about 28% of the time. The median team scored two runs. Teams scored exactly four runs 83 times, often due to a grand slam.

Half of the league’s teams produced runs in line with those trends in 2019, with 15 of the 30 averaging between 2.0 and 2.5 runs per bases-loaded, no-out opportunity. There were some notable outliers, however, on both ends of the scale.

The Phillies and Cardinals, as might be assumed based on the links above, led the majors with 3.1 and 2.9 runs per inning, respectively, in this situation. Those single innings skewed their samples a bit, but even with them removed the Phillies averaged over 2.6 and the Cardinals 2.5 runs per opportunity. All told, they combined to score about 24 runs more than the average across their 36 innings in this sample.

On the other end of the spectrum, two teams that reached the postseason in 2019 also tied for the basement in terms of converting these specific opportunities into runs. The Twins had 19 innings with the bases loaded and no outs and proceeded to score just 27 runs, just slightly more than 1.4 runs per opportunity. They were held scoreless in these situations five times, tied for the second most in baseball, but their worst situational failure wasn’t one of them.

On July 23, the Twins entered the bottom of the ninth at Target Field trailing the Yankees 12-11, having given up a two-run homer to Aaron Hicks in the top of the frame. Aroldis Chapman walked each of the first three batters to load the bases with none out, and FanGraphs estimated the Twins had a 74.6% chance to win the game. They got a single run on a sac fly to tie the game at 12 but nothing more and lost 14-12 in ten innings.

The Twins’ companion at the bottom of this list was one of their oldest rivals, the Brewers. The Brewers also scored just 27 runs across 19 bases-loaded, no-out opportunities, averaging almost a full run less than the league average. Their first batter to come to the plate in this situation went 1-for-17 with a single and a sac fly, batting .059/.056/.059 in this split. Their most compelling failure to score, however, started with a baserunning blunder.

On the last day of the 2019 season, all 30 teams started games at the same time. But the Brewers and Rockies, who had nothing to play for, were the last to finish. The Brewers already had clinched an NL Wild Card and knew they were headed to Washington, D.C. The Rockies, at 70-91 on the season, had been eliminated for weeks and were just playing out the string.

Yet somehow, long after 28 other teams had wrapped up for the day, these two squads found themselves in the top of the 11th inning tied at three. The Brewers loaded the bases on a walk, double and hit by pitch. With Manny Pina at the plate and both teams likely eager for a run to score and mercifully end their day, Ben Gamel tried to score from third on a pitch in the dirt and was thrown out. The Brewers did not score in the inning, and the game plodded on, with Colorado eventually winning 4-3 in the 13th.

All told, the Brewers and Twins each left about 16.8 more runs on the table in bases-loaded, no-out situations compared to the league average for their respective volume of opportunities. They also combined to win 190 games, so the difference didn’t ruin their seasons.

Once all 546 innings were divided up by team and broken down by runs scored, there were a few teams that stood out on the positive end and five whose performance was nine or more runs below average:

In all cases above and in the next section, runs are only counted if they score after the bases are loaded with none out. The runs that score before reaching that state still count in the score, of course, but they’re not relevant to the question of how teams perform after reaching this situation.

Preventing Runs

When it comes to escaping the ultimate jam, it again quickly becomes apparent most teams are about the same. As was the case above, exactly 15 teams averaged between 2.0 and 2.5 runs allowed per bases-loaded, no-out situation in 2019, leaving them within a few tenths of a run of the league average.

On the positive side, the Phillies easily were the best team in baseball at escaping this jam. Their pitchers loaded the bases with none out in 19 separate innings in 2019 but escaped unscathed in four of them (tied for the most in baseball) and allowed a total of just 31 runs, slightly more than 1.6 per jam. Their average was more than two tenths of a run lower than any other team. 

In those situations, the Phillies excelled at getting the first batter out. Across 21 batters faced with the bases loaded and none out, Phillies pitchers allowed opposing teams to bat just .125/.190/.188. One of the outs in that sample was by Mets catcher Wilson Ramos. On September 8, he came to the plate in that situation trailing 7-4 in the fifth inning and grounded out. A run scored, but the Phillies had successfully bent but not broken. They allowed another run in the inning on a sac fly but escaped still holding the lead and went on to win.

Despite the 10-run disaster mentioned above, the Mets still did well here also. They allowed 40 runs after 20 bases-loaded, no-out situations, but 10 of them scored in that single outing against the Phillies. They allowed just 30 runs in the 19 other innings, fewer than 1.6 per opportunity.

The Brewers, meanwhile, were the standout on the poor side once again. They allowed opponents to load the bases with none out 17 times and gave up 54 runs following those situations, an average of almost 3.2 per nine innings and a full three tenths of a run worse than any other team. They held an opponent scoreless just once after finding themselves in such a jam and gave up at least six runs on three different occasions. In all three of those games, they were leading the contest entering the inning before giving up a big frame.

Who benefited?

Because the samples are small, the number of runs gained or lost over the course of a season didn’t add up to a lot for most teams. Combining the last two tables to find overall winners and losers shows all but a handful of teams were within eight runs of what would have been expected, given their offensive and defensive opportunities. The difference between the top and the bottom of the list, however, is over 50 runs.

All told, the results here clearly do not correlate well with overall team success. The top and bottom five above each feature at least one postseason team and multiple 100-loss teams. In at least once case, however, these outcomes may have shifted a division race. The Cardinals were over 43 runs better than the Brewers in this situation and finished just two games ahead in the NL Central.

A larger sample likely is needed to determine if these outcomes reflect one or more repeatable skills and what those skills might be. In terms of runs scored and prevented, however, it is clear some teams perform better than others after creating or while attempting to escape the ultimate jam.

Acknowledgements

Readers curious about a specific team or event, or who simply like playing with data, can see the data used for this article here.

This research was made possible by the Event Finder in the Baseball Reference Play Index and by BR support staff, who compiled a list of dozens of events left out of the original search. 

Additional thanks to Tangotiger.net for its work compiling Run Expectancy models for past seasons to provide context to these data.