Breaking Bob: The 60-Game Season and the ERA Record

One of the greatest myths of baseball history is the asterisk supposedly added to Roger Maris‘ then-record 61 home runs in the 1961 season. As the story goes, Major League Baseball, aghast that Babe Ruth’s home run record could be broken by Maris in a 162-game season when Ruth’s Yankees only played 154 games, forced Maris’ record to wear scarlet punctuation in order to shame it in the record books. The only problem is that the Maris asterisk never actually happened. Commissioner Ford Frick, who held the job at a time when he was still expected to at least give lip service to the idea of being a steward to the abstract notion of baseball, was simply expressing his opinion; no asterisk ever appeared next to Roger Maris’ name or record.

The truth about baseball’s record book is that its entries have never had much in the way of purity. From changes in the baseball and the mound to whether players could or could not spit on the ball, numbers in one season never really mean exactly what numbers in other seasons do. Not even baseball’s greatest shame — enforcing a grotesque color line that robbed countless star baseball players of their turn in the majors — resulted in a culling of statistics from the game’s first century.

Records tend to be set in an environment conducive to setting them. No dead ball season features a player with a home run total in even the top 1,000 in history; there’s a reason that strikeout records, both positive and negative, are a feature of modern baseball, not antiquity. Bob Gibson’s 1.12 ERA in 1968 is currently recognized as baseball’s number to beat only because the dead ball era was deemed to be too different; Tim Keefe, Dutch Leonard, and Three Finger Brown each boast one better than Gibson’s. And Gibson’s record itself reflected the environment — The Year of the Pitcher resulted in baseball lowering the mound by a third and reducing the size of the strike zone for the 1969 season.

In the “too different” department, 2020 may be the ultimate outlier season, though that will depend on whether COVID-19 lets us have a “normal” 2021. Baseball has played shorter seasons before, but going from 162 games to a mere 60 is a whole new thing. And in 60 games, all those rate-based statistics have the potential to go up in smoke. Among them is Gibson’s aforementioned 1.12 ERA, at risk even in a 2020 season that almost certainly won’t be a pitcher’s paradise. Just last season, Jack Flaherty maintained a 0.91 ERA over the season’s second-half, a number that over a full season bests Gibson and two of those three dead ball era pitchers, falling short only of Keefe. That wasn’t a full season, of course. The Cardinals played 74 games after the All-Star break, a chunk of games nearly 25% greater than the 60 games that will (cross your fingers) be played in 2020. Anything that can be maintained for 74 games can certainly be maintained for 60! You can even draw the line for Flaherty’s ERA a week earlier than that; he had a 0.93 ERA in the team’s final 78 games if you make the mark after the team’s loss against the Mariners on July 2.

And Flaherty wasn’t a fluke. It doesn’t take much additional effort to find two-month “seasons” in which pitchers who threw at least 60 innings also came out ahead of Gibson. Our split data only goes back to 2002, a time definitely not conducive to run prevention, and Flaherty still finds a lot of company, even if we limit the two months to those periods neatly bound by calendar pages:

Naturally, Gibson himself had a two-month chunk of the season much better than his 1.12 ERA: in June and July of 1968, he allowed only six runs in 108 innings, for an ERA of 0.50.

So, as the owner of a projection system, I set ZiPS the task of projecting the probabilities of pitchers in 2020 finishing with a 1.11 ERA or better. The odds are likely shorter than you think. While no pitcher has a “true” talent amounting to a 1.12 ERA and the entire season is a small sample, what helps things out is that ability is not a number written in stone. We have more no-hitters and consecutive hit streaks than we would expect if player abilities were fairly static within even single seasons. The pitcher we’re fairly sure has an ERA ability of somewhere around three likely has a “true” baseline of two at some times and four other times. The more volatile talent is, the greater the chance we’ll find the pitcher who can break the 1.12 ERA record with the benefit of a short season. Buster Douglas didn’t need to be a better boxer than Mike Tyson to beat him, he only needed to be a better boxer than Mike Tyson once.

There are 76 pitchers on our depth charts projected to throw enough innings to qualify for the ERA title. I asked ZiPS to assume that each of their projected inning totals was accurate and gauge the probability that they would have an ERA of 1.11 or better over those innings. Obviously, the top of this chart is full of star pitchers:

In the end, ZiPS expects 0.29 pitchers, on average, to finish the 2020 season with an ERA under 1.12. Overall, there’s a projected 25%, one-in-four, chance that somebody will break the modern ERA record (and in some runs of ZiPS, multiple pitchers). In a normal 2020 with 162 games, ZiPS only projects a one-in-940 shot of someone finishing with that ERA or better.

Baseball has largely avoided asterisks in its history, but 2020, arguably the weirdest year of our lives, may be the season that brings that bit of punctuation out of the mothballs.