The Houston Astros’ recent winning streak has livened up what has been a more tumultuous season than many envisioned.
On paper, of course, there is not much for Astros fans to complain about. But in comparison to last season, which felt like something special from the opening pitch, this year has had a somewhat sour note. The bullpen has struggled and been a point of contention, the offense isn’t quiet as deep as before, and the Astros trail the Mariners - the Mariners - for first place in the American League West. What’s happening? A lot of noise.
The Astros are 42-25 while the Mariners are 42-24, but their records could not be more misleading - the Astros have a +127 run differential, best in the league, while the Mariners run differential is +22. The Astros are outscoring their opponents by 1.90 runs per game while the Mariners average lead is 0.33, yet Seattle stills leads the division.
What’s causing such a big discrepancy? The Mariners are 21-9 while the Astros are 6-12 in one-run games.
So am I telling you the Mariners are more clutch than the Astros? Not really. While its true that well managed teams with great bullpens can do slightly better than average in one-run games, high-leverage performance has never been predictive, and extreme cases like the Astros and Mariners are both prone to extreme regression.
We saw it with the Rangers in 2016. It was infuriating to watch, but it was obvious to many that the Ranger’s weren’t nearly as good as advertised. Many vehemently argued against this mind-set - saying it was ridiculous that the Ranger’s should have to “apologize” for winning an abundance of close games - but that was perhaps missing the point.
It’s not personal or an excuse - in baseball, we have many numbers to capture when a player or team is over-performing, and by every measure in the book the Ranger’s 2016 regular season was a fuggazzi. I got into far too many twitter spats with Arlington fans that claimed their team of destiny was too clutch to fail - only to see the team unravel in the post-season and not come close to returning since.
The 2018 Mariners have not been nearly as extreme - their performance in one-run games is the only luck indicator that is off the charts - but the Astros unreasonably awful pitching in late-game situations has only confounded the situation. But I don’t expect either trend to continue.
In order to demonstrate this point, I graphed the “clutch” ratings of Seattle and Houston’s bullpens over the last couple of seasons. As you can see, each bullpen’s performance in high leverage situations has had little to no correlation with the year before, and in many instances we saw a swing to the opposite. Similar clutch “trends”, such as record in one-run games, never hold up outside of a small sample.
If you look at Pythagorean expectation, which predicts a team’s record based on their run differential, the Astros “should” be 48-19 while Seattle “should” be 36-30. Baseball Prospectus’s 3rd-order wins, which go a step further by also predicting how many runs each team “should” have scored and allowed, also predicts the same records for both teams. Both measures actually had the Angels ahead of the Mariners for second in the AL West. Full season projections will paint a similar picture.
The point is not that the American League West race is already out of reach or that the Mariners stand no chance - the point is that you should not think the Mariners stand a chance based on the results of this season. In other words, not much has really changed from last season.
Things could always change - the Mariners could make a trade that puts them in a prime position to steal the AL West - but realistically, as it stands today, the two teams are in different tiers.