Toronto Blue Jays closer Roberto Osuna was straight-up filthy prior to the All-Star break, with his numbers putting him within an earshot of guys like Kenley Jansen and Craig Kimbrel.
But after dominating to the tune of a 2.06 ERA and .198 wOBA allowed, those numbers regressed to 4.97 and .249 throughout the summer and fall. While he did see some regression in strikeout rate (36.4% to 30.0%) and walk rate (2.3% to 5.0%), it's not as if he was pitching all that badly, evidenced by his 2.56 second-half SIERA (it was 1.98 in the first half).
What did take a precipitous drop were his chase rate (47.3% to 32.9%) and swinging-strike rate (19.1% to 14.4%). These regressions can likely be attributed to what happened with his slider and cutter. Of his three most-used pitches in 2017 (fastball, slider, cutter), his slider and cutter induced the most swings and misses while also forcing opposing hitters to chase out of the zone most often.
When we look at the value of these pitches throughout the course of this season, it's easy to see why his strikeout rate and walk rate both worsened while also allowing a lot more hard contact (20.3% in first half, 36.0% in second half).
| 2017 | SL Usage | SL Value | CT Usage | CT Value |
|---|---|---|---|---|
| First Half | 25.3% | 6.7 | 20.5% | 3.0 |
| Second Half | 19.4% | 2.0 | 30.6% | 0.6 |
It appears that Osuna made a significant leap between 2016 (2.87 SIERA) and 2017 (2.27 SIERA) because the performance of these pitches improved. He depended on them more than ever, too -- his fastball usage went from 69.3% from 2015-16 to just 48.0% this past season. The link between that first-half dominance and second-half regression appears to live in how those secondary pitches perform.