Skip to main content
ABC News
Home Runs Are Taking Over The Playoffs

This October, MLB has been dominated by home runs. About 43 percent of the runs scored so far this postseason have come via homers, a greater fraction than any year in the wild-card era. That reliance on the home run could make the postseason more random than ever.

The Guillen Number, first introduced by Joe Sheehan at Baseball Prospectus, attempts to measure how dependent an offense is on the home run. It’s simply the number of runs scored by a team through homers divided by the number of runs. As the Guillen Number increases, a team becomes more dependent on the rare occasions in which a ball clears the fences.

This year, the league’s total Guillen Number was 40 percent, the highest it had been since at least 1995, and we saw three of the top five Guillen Numbers ever.1 This is a consequence of the league’s gradual evolution toward more strikeouts but also more power: With fewer runs generated via moving runners on the bases — and more homers than at any time since the peak of the steroid era — a greater fraction of MLB’s offense than ever is coming via the long ball.

A higher Guillen Number can produce majestic homers, but it has other consequences as well. More runs scored via home run makes the outcome of games more random. That’s because homers are both inherently rarer and more variable than other plays: While they are guaranteed to score at least one run, they can bring in as many as four if the bases are loaded. That means the final outcome of a homer strongly depends on whether the prior hitters made it to base, which is itself increasingly rare.

As a more significant fraction of the league’s scoring relies on chaining together these rare events, the outcome of any individual game becomes less predictable. Although it’s not a hard and fast rule, looking across baseball history2, the seasons in which Guillen Number have been the highest have tended to feature the lowest correlations and largest deviations between run differential and winning percentage.3

Whenever the Guillen Number is large, the better team is less favored to win. As an underdog or a weaker team, that works to your advantage. On the other hand, the superior team should crave a lower Guillen Number, one that makes every game more predictable.

Of the teams remaining in the playoffs, the Cleveland Indians and Chicago Cubs had below average Guillen Numbers in the regular season (at 35.26 percent and 39.6 percent, respectively), while the Los Angeles Dodgers were slightly above average, with 40.83 percent. In contrast, the Toronto Blue Jays had the fifth-highest Guillen Number in the game this season, at 46.11 percent.

Considering that they are favored against the Indians, that may work to their disadvantage in the American League Championship Series. However, that elevated Guillen Number would also make the Jays a dangerous team against the Cubs, should they both advance as our projections forecast. In a season marked by more predictability than baseball usually provides, the league’s reliance on the homer this October is providing one last burst of randomness.

FiveThirtyEight: Can Cleveland surprise us again?


  1. None of those three teams — the Seattle Mariners, New York Mets and Baltimore Orioles — are still in the postseason.

  2. Going back to 1950, since that’s as far as Baseball Prospectus keeps data.

  3. I compared the league average Guillen Number to the correlation between run differential and winning percentage each year (r=-.15). I also compared Guillen Number to the sum of squared errors from regressing run differential against winning percentage (r=.27).

Rob Arthur is a former baseball columnist for FiveThirtyEight. He also wrote about crime.