How Do Games End?

When it was announced that the MLB was going to have to seven inning doubleheaders this year, I was curious about seven innings games. While there are numerous resources to research that I yanked the data from Retrosheet and used the play-by-play data they provide. As typical, for me, a dive into play-by-play data, things turn sideways quickly.

Play-by-play data is the historical record of baseball. It offers more insight into the game even if we know that information may be an inaccurate accounting of the events as umpires are not always the best witnesses and judges.

The fine folks at Retrosheet are a dedicated group of baseball historians who have one small goal: to record the play-by-play data for every baseball game then can. As of today, they provide the play-by-play data for nearly every game since 1916, including over 170,000 games and 13,000,00 events. The best part? They make all this available for free. They also provide software tools to investigate the event files and pull a TON of information out of it. The Chadwick Baseball Bureau offers additional tools and information. So, humbly standing on the shoulders of those giants, I moved from investigating seven innings.

Why?

When I asked the play-by-play data to show me all the games that ended in the seven innings and what was the last event play of the game something interesting popped up.

Four seven-inning games ended with an intentional walk. This is equal to the number of games that finished with a triple or home run, combined! Odd way to end the game.

So, I checked all the games to see how often this type of situation happened.

In over a century of games, an intentional walk had happened six times as much as interference.

Looking at all the ending of games I show about 90% of the time, games end with an out or a strikeout. Games end with home runs about 2.4% of the time.

Since play-by-play data is the historical record of baseball. We should expect the data to show a few trends. One trend I will assume that 90% of the time games end without outs. Since pitching has changed, relief pitching increasing, more pitchers throwing harder, etc. one can assume that strikeouts would be more prominent at the end of games. At the same time, we know that home runs increased starting in the late 1990s, and we should expect that more home runs in baseball itself should translate to more games ending with home runs.

Breaking down by decade, outs + strikeouts is around 90%.

The change from the 1910s to the 1920s was the only time that strikeouts go down significantly. Since then, strikeouts have increased. There is a small drop between the 1960s to the 1970s with strikeouts. In the late 1960s, pitching changes to the mound and other various aspects of baseball were made that pushed the game towards a contact game for a bit. Since the 1990s, strikeouts are increasing and teams are using fresher pitchers at the end of the game. The assumption about strikes holds.

What about game-ending home runs?

We see what would be expected. Home runs increased steadily as the game changed from the 1910s to the 1960s. You would expect baseball to show some volatility three decades after the 1960s. Games were played in bigger, multi-purpose stadiums that kept home run numbers down. In the 1990s, teams started building newer, smaller stadiums. At the start of the 2000s, those new smaller stadiums had taken over. Players were getting bigger, and pitchers were throwing hard. While pitchers were getting more strikeouts, they were also giving up more home runs. The assumption about home runs is shown in the data.

We can see this by the play-by-play data and we can see it in history. The ending of games has changed as one would expect. It should be interesting have this year to see if this pattern holds true. Retrosheet also has play-by-play data for postseason games, it will be interesting to see if these assumptions care over into the postseason.

The play-by-play data tells a story. It is a historical record of baseball. Odd tales are available there to be found. Let’s get back to those.

Did you figure out how six games ended in intentional walks yet?

Those games were called because it rained. The last time it happened was July 31, 1992. The Red Sox lost the second game of a double here when the game was called in the top of the 6th inning.

The game-ending interference?

August 1, 1971, Reds playing the Dodgers, bottom of the 11th inning, bases loaded. Yep! A game-ending bases-loaded catcher’s interference by Johnny Bench. Manny Mota was attempting to steal home. Johnny Bench stepped in front of home plate and was called for interference. Manny Mota scored, game over.

And who was the runner at second base when the play started? Bill Buckner.

Photos Adapted by Zach Ennis (@zachennis on Twitter and Instagram)