clock menu more-arrow no yes

Filed under:

The Rockies’ record and run differential tell different stories, both wrong

New, 41 comments

What the Rockies record after 14 games tells us.

The Rockies are 9-5 to start the season. That’s great so far, but digging deeper into record estimators reveals more of a mixed bag for the Rockies. In particular, the Rockies’ record and run differential tell conflicting stories, while the more sophisticated record adjusters put them in between the two.

Let’s start with the raw wins and losses. The Rockies have won nine games, and they have lost five games. The numbers are small, but a 9-5 record is pretty impressive. The winning percentage is .643. Extrapolate that for an entire season, and it’s a 103-59 team. This is an “on pace” conclusion, and it’s best taken with a grain of salt (Greg Holland is also “on pace” to save 81 games, and I’m “on pace” to live forever). The most significant thing we can take away from this start is that those nine wins aren’t going away. The Rockies own ‘em.

A major reason the Rockies have such a good record so far is that they’re 5-0 in one-run games. One-run run games in general are less indicative of a team’s overall quality than they are in seeing on which side of a coin flip they’ve fallen in a given timeframe. The 2016 Cubs, for example, were just 22-23 in one run games. That’s typical. It’s true that the Rockies’ bullpen has been excellent so far this season, and they should be credited for helping the team hold on to those five wins. But sooner or later the Rockies will enter those late innings down instead of up, and the bullpen won’t be able to help. There are always outliers like the 2016 Rangers (36-11 in one-run games), but I wouldn’t count on the Rockies to come close to replicating that.

The raw record doesn’t tell the whole story. A different measure of how well a team is doing, and it relates a bit to record in one-run games, is its Pythagorean record. Pythagorean record is based on run differential. The foundation of Pythagorean record is that it moves beyond pure results. Instead, it bases the record on how many runs a team has scored against how many runs they have given up. It is supposed to tell us something about how well the team has truly performed against competition.

The Rockies aren’t doing so well in this area. In the National League, only the Pirates, Cardinals, and Padres have a worse run differential than the Rockies’ -7. Translate that differential into a record, and the Rockies are playing like a 6-8 team. That .440 win percentage extrapolates to a 71-91 record. The difference between the Rockies’ actual record and this performance based measure is pretty big. The idea behind Pythagorean record is that it’s more reliable about a team’s overall talent and is therefore more trustworthy when guessing how the team will do going forward.

But while run differential provides a reasonable record adjustment, there are more sophisticated approaches. FanGraphs’ BaseRuns and Baseball Prospectus’s 3rd Order Win Percentage both adjust for additional underlying factors beyond run differential, such as how many runs a team “should” have scored based on the situations the team found itself in throughout the season, as well as quality of opponent. Here the Rockies find themselves in the middle. By FanGraphs’ estimation, the Rockies should be 7-7—not bad, but not what Rockies fans are hoping and seeking after this great start to the season. Baseball Prospectus’s measure is more optimistic. It thinks the Rockies should have a .547 win percentage right now. That would translate to about 88 wins. It isn’t 103, but it sure as hell isn’t 71.

We have what the Rockies have done so far, as well as a few estimations of what they should have done. Taking the extremes, we can say with a whole lot of confidence that the Rockies will win somewhere between 71 and 103 games. The truth of the middle is where we find the most nuanced record estimators. What we can say about the Rockies’ record so far is that the raw record paints too rosy of a picture, while the run differential is a grimmer portrait than really exists. There’s a lot of baseball left to reveal more truths. I bet we’ll find it somewhere in the 80s.