- Zach Britton's dominant season in Baltimore has made him a popular AL Cy Young contender, but even in a weak field, he shouldn't be the favorite.
Should Zach Britton be a serious contender for this year’s American League Cy Young Award? That question has been making the rounds this week in response to Britton’s utter dominance as the Orioles’ closer and the lack of a dominant starting pitching performance in the AL. Strong arguments can and have been made both for and against Britton’s candidacy, and both points of view can be supported by sound statistical analysis. It's a question that resonates beyond Britton’s performance and this year’s AL Cy Young race, cutting to the heart of the intention of the annual awards themselves.
For those who haven’t noticed, here’s what Britton has done this season: 50 innings in 52 appearances, with just 25 hits, six runs (three earned), one home run and 16 walks (three intentional) allowed against 59 strikeouts. Britton has converted all 37 of his save opportunities (a league-leading total) and has made a major-league-record 41 straight appearances without allowing an earned run. Though there is still a month and a half left in the season, his current 0.54 ERA and 826 ERA+ stand as all-time–single-season records among pitchers with 50 or more innings.
Despite those gaudy numbers, Britton has yet to get so much as a passing mention in my Awards Watch rankings, and it is not because he has been lurking just outside of the top three; it’s because of how few innings he has pitched. The top man on my imaginary AL Cy Young ballot last Thursday was Cleveland’s Corey Kluber. On an inning-for-inning basis, Kluber hasn’t been nearly as dominant as Britton this season, but in my opinion, Kluber has out-pitched the rest of the AL's starters and has thrown 113 more innings than Britton. By way of comparison, the current AL ERA leader, Tigers rookie Michael Fulmer, has thrown 120 innings all season. Fulmer and Britton combined have thrown just seven more innings than Kluber this season.
My point here isn’t to argue for Kluber as much as it is to argue against Britton. There’s simply no way that a pitcher can be good enough in 50 innings to edge out a pitcher who is having a strong season and has thrown over three times as many innings. Consider this: If we subtract Britton’s 50 innings and three earned runs from Kluber’s season, we’re still left with 113 innings of a pitcher with a 4.30 ERA; the average AL starting pitcher this season has posted a 4.44 ERA. What’s more, Progressive Field has been far more hitter-friendly since its pre-2015 renovations, so it would be fair to say that a 4.30 ERA from a Cleveland starter would be significantly better than average. Kluber’s performance this season has effectively combined Britton’s performance with 113 innings of an above-average starting pitcher. What possible argument could there be for Britton in light of that?
On Tuesday, The Ringer’s Ben Lindbergh did just that, pushing not only Britton’s Cy Young candidacy, but also tabbing Baltimore's closer as the AL's Most Valuable Player. Lindbergh’s argument, built entirely on advanced statistics, is based on the importance of the innings Britton has pitched: According to Win Probability Added (WPA), Britton has done more than any other pitcher this year to push his team toward victory. The outcome of every plate appearance in a game adjusts a team’s probability of winning that game upward or downward; what WPA does is add up all of those changes. That stat credits Britton with having added 4.52 wins to the Orioles' ledger this year. The injured Clayton Kershaw is second among pitchers (still!) with 3.72 WPA. Toronto's Aaron Sanchez leads AL starters with 3.04 but trails his own closer, Roberto Osuna (3.04), as well as Kluber’s new teammate, reliever Andrew Miller (3.08); Kluber ranks 80th at 1.07.
But while Britton leads all pitchers in WPA, Angels centerfielder Mike Trout is No. 1 among all major leaguers at 5.07. To get Britton into the MVP lead, Lindbergh had to go a step further, turning to a stat of his own devising called Championship Probability Added (cWPA), which factors playoff odds into WPA in an attempt to determine what player has most helped his team move toward a championship. Per Lindbergh, Britton had a huge lead in that category earlier this week, leading Nationals second baseman Daniel Murphy, .040 to .027; by Lindbergh's math, Britton has increased the Orioles' chances of winning the World Series by four percentage points. Trout—whose team has a 0.0% chance of winning the World Series or even making the playoffs this season, per Baseball Prospectus’ Playoff Odds Report—doesn’t make the list.
WPA and cWPA are interesting and informative, but to my mind, they have absolutely no place in the discussion about an individual player award. Though far more significant in terms of what they can tell us, win-probability statistics are little more than a sabermetric version of runs batted in, a statistic that advanced analysis long ago devalued as opportunity-based. In order to drive in a lot of runs, a player needs his teammates to get on base. In order for a player to compile a high WPA, he needs the game conditions to allow for a large swing in win probability. And in order to compile a high cWPA, he needs his team not only to play close games, but also to play them in a pennant race. The O's have done all of the above for Britton: The AL East is a tight, three-team race, and Baltimore has been in the thick of it since opening the season 7–0. What’s more, 73 of the Orioles’ 119 games have been decided by three or fewer runs, and Britton has entered just seven of his 52 games with his team ahead or behind by more than three runs. By way of comparison, Cleveland is dead last in the majors with just 61 games decided by three or fewer runs, and just eight of Kluber’s 24 starts have come in games decided by three or fewer runs.
Using context-dependent statistics to decide an individual award artificially limits the field of eligibility for that award. There are some who believe that is the intent of the MVP award. Those individuals use the vagaries of the word “valuable” as an excuse to limit their candidates to contending teams, ignoring the fact that such an interpretation is contradicted both explicitly and implicitly in the official voting rules. But I have never heard anyone argue that the Cy Young award is intended for anyone other than the best pitcher in the league, independent of team performance.
My understanding of these individual awards is that they are intended to recognize the most outstanding player, pitcher and rookie in each league independent of team accomplishments, which are recognized by playoff berths and postseason series wins. The best player is the “most valuable” player, and the best pitcher is the most valuable one, as well—a pitcher who not only dominates the opposition but also consumes innings in a manner that allows his manager to deploy part-timers like Britton more strategically.
Because, ultimately, that’s what Britton is: He’s a bench player, a specialist, and arguments can be made that he’s not even the best in the AL at that. According to Fielding Independent Pitching, Yankees closer Dellin Betances has been better (1.26 FIP in 57 1/3 innings to Britton’s 2.00 in 50 IP). According to Deserved Run Average, the Indians' Andrew Miller, acquired from the Yankees at the trade deadline, has been better (1.44 DRA in 54 IP to Britton’s 2.22). Both also best Britton in BP's Wins Above Replacement Player statistic, which factors in DRA (2.34 for Betances and 2.26 for Miller to 1.63 for Britton).
Britton’s dominance does merit consideration for this year’s AL Cy Young award, and there is an argument to made in his favor. But that argument is based on a method of analysis that tilts the playing field in a way that unfairly favors high-leverage relievers by placing far too much emphasis upon the context of a player’s performance. In most scenarios, context is a good thing, but in this case, it is another word for the performance of a player’s teammates. When it comes to an individual award, that type of context should be irrelevant.