You might not have realized it, but last night’s slate of games had seven teams reach double-digit run totals. Throughout the night I began to notice there were quite a few games with really high run totals. Then last night while watching the Astros pile runs on the Padres into the 9th inning, armed with a laptop and some SQL code, I set out to research how often this many teams score 10 or more runs on a given day.

MLB Game Days with Teams Scoring 10+ Runs

It turns out this hasn’t happened since 2012, and it has only occurred on nine days over the last decade of baseball. One of those days in 2007 had eight teams break double digits, and all of these days occurred on days with 15 games being played. Yesterday, there were only 14 games played due to the postponement of the White Sox – Orioles game. Most of the occurrences happened earlier in the 2000s when the run environment in Major League Baseball was higher. Today’s low run environment makes days like yesterday less and less likely.

This also gives us a chance to look at how a team’s runs per games are distributed to find out how likely it is that one team scores 10 or more runs in a game. They don’t follow a normal, bell-curve distribution, but instead it’s more skewed (and roughly follows a negative binomial distribution). I’ve illustrated the distribution below using actual results. This represents the percentage of games that a single team scores a certain number of runs.

The run scoring is skewed heavily to the right. I’ve called out the percentage (6.9%) of team-games that have had 10 or more runs scored, so they don’t happen very often. I have the box going all the way out to 30 runs, because that’s the record the Rangers set in Baltimore in 2007. Seven teams having 10 or more runs on one day is obviously less frequent than having just one team do it especially when there are only 28 teams playing instead of 30.

For those of you interested in reading more about run distributions, I have already laid out the math behind modeling run and hit distributions with the negative binomial distribution, and Tom Tango has produced a more accurate but slightly more complicated distribution as well.