A couple weeks ago, Dellin Betances struck out Tampa Bay’s Kevin Kiermaier to break the Yankees’ single-season strikeout record for a relief pitcher. Mariano Rivera was the previous record holder, punching out 130 batters in 1996 as the bridge to John Wetteland. Betances was one of the few bright spots for this season’s team, and his performance has drawn nostalgia for Rivera’s emergence as a relief ace nearly twenty years ago. Obviously, we can’t compare the two pitchers’ campaigns on results-based statistics like strikeouts for a few reasons, namely because it would ignore the difference in era played. Using rate and environment adjusted stats, we can get an idea of how the two seasons compare.
Dellin has Mariano beat in ERA and FIP, before accounting for league and park adjustments. However, the differences are close enough that I thought ERA- and FIP- (both via Fangraphs) might tilt the scale back into Rivera’s favor. This was only the case for FIP-, albeit by a mere 3%. Incredibly, Rivera allowed just one home run in 107.2 innings in a time when power was up. Home runs allowed are a key component of FIP, and not allowing home runs in a year such as 1996 will go a long way in a league adjusted stat like FIP-. ERA- is another story, in which Betances has a 7% advantage. Not even the hitter-friendy ’96 could make up for the pitcher-friendly ’14 in this case, although Betances’ ERA advantage (0.69) is a bit larger than his FIP advantage (0.24).
In terms of strikeouts, Betances is not only king in total, but also far ahead in punchouts per batter faced. Dellin has fanned 39.6% of opponents, outshining Rivera’s mark of 30.6%. It took Betances 327 batters to reach the franchise record-setting 131 Ks — Rivera faced 425. This isn’t to discredit Rivera’s performance, as 30.6% is an incredible mark as well. It also helps Rivera’s case to point out the fact that league-wide strikeout percentage was 16.5% in ’96, while it now stands at 20.4%.
Baseball Reference also has a nifty season-equivalency tool, in which Betances’ 2014 numbers can be translated to same environment Rivera pitched in 1996:
This portrays Betances’ 2014 as a better performance on a rate basis, interestingly disagreeing with Fangraphs’ notion that Rivera’s FIP performance was better on a level playing field (Mariano’s FIP- is better than Dellin’s).
By fWAR, Rivera’s got Betances beat: 4.3 to 3.2. If you prefer rWAR, same deal: 5.0 to 3.7. In fact, Rivera has Betances beat in a few other seasons by both metrics, but the differences are small enough that the measure’s margin for error might come into play. Nonetheless, the full-win difference between Mo’s ’96 and Dellin’s ’14 is too large to chalk up to error. However, part of the difference is because Rivera pitched more — 17.2 innings and 84 batters to be exact. One could prorate the metrics for the two pitchers to an even amount of batters faced, but this would be unfair, as it would assume Betances would continue pitching at the same level against additional batters.
Grantland’s Ben Lindbergh also tackled the the stature of Betances’ 2014. In particular, I want to focus on his use of RE24 wins (REW) in order to place Dellin’s season in context to the greatest relief seasons in history (post-1988, min. 80 IP – see Lindbergh’s article for explanation of these qualifications). Now, let’s quickly look at how Ben explains RE24 wins:
…REW calculates the difference between the number of runs a team is expected to score in the inning at the start and the end of each play, credits/debits the batter/pitcher accordingly, and then compiles the differences to arrive at a full-season total of wins added or subtracted.
Rivera’s 1996 is tops at at 3.927, while Betances stood at 3.280 at the time of the aforementioned article’s publication (September 19). Dellin finished the season at 3.291, good for fifth-best all-time per Ben’s qualifications. Rivera has a leg up in REW, and I don’t think it would be fair to prorate this stat to an even amount of batters faced, similar to the reason why I didn’t for WAR. We can’t assume Betances would continue to pitch and/or enter the game in similar base-out states in additional innings pitched.
So, who had the better performance? Betances sweeps all of the rate stats (ERA, FIP, K%, BB%), even after adjusting to 1996’s run-scoring environment (with the exception of FIP-). However, Mo’s body of work results in triumphs in rWAR, fWAR, and REW. It’s a dead heat, but I think Rivera gets the nod thanks to his greater usage. Why? Because the more Rivera was used, the more innings that were taken away from lesser relievers. While his rate stats are slightly less impressive than Betances’, he added more wins to the Yankees ledger. But hey, when choosing between a Bentley to a Rolls Royce, one is going to be satisfied either way.