Defense and Contact and the American League Cy Young

There are five spots on a Cy Young ballot. This season, in the American League, only the first two are going to matter. Last year, Justin Verlander received the most first-place votes, with 14, but he was ranked second on just two ballots. Rick Porcello, meanwhile, received 18 second-place votes in addition to his eight at the top spot. Porcello ended up the winner.

This year, Chris Sale and Corey Kluber are likely to occupy the top two spots on every ballot, and whoever receives more first-place will probably take the award. The pair have produced very similar statistical records this year, so voters will have to split hairs to decide which of the two is more deserving. Let’s begin splitting some of those hairs right now, examining defense, contact quality, various sorts of WAR, and how they all influence each other when it comes to the two best pitchers in the American League.

To provide an initial tale of the tape of sorts, here are a few numbers with which most readers will be pretty familiar.

The strikeout and walk numbers are pretty equivalent. Sale has the edge in homers and innings, leading to a lower FIP and therefore WAR. Kluber has a much lower BABIP, which helps him lead in ERA. If you want to factor for league and park, we can do that. Sale’s FIP- is 50 to Kluber’s 57, and Sale’s ERA- is 61 to Kluber’s 54. Part of Kluber’s ERA advantage comes from that BABIP. He also has an 81.6% left-on-base percentage compared to Sale’s 76.7%, though, so part of the advantage is simply due to sequencing.

So those are some of the most basic numbers. What about each pitcher’s value in terms of wins? As you know, there are different versions of WAR on different sites, each valuing certain aspects of a player’s game in a slightly different manner. The table below shows a few of the WAR numbers from a few different sites.

The first row features FanGraphs’ main version of pitching WAR. Based on FIP, it adjusts for park and league, as well as infield flies. Due to Sale’s advantage in FIP, as well as his slightly higher innings total, he has the lead here.

RA/9 WAR is also found at FanGraphs. It essentially substitutes runs allowed for FIP. Because Kluber has allowed fewer runs this year, he comes out a bit ahead.

Baseball-Reference starts with RA/9 and then factors in opponents, park, and team defense. Kluber come out well ahead due to the advantage in runs allowed. Baseball-Reference also estimates that the Red Sox have played better defense than Cleveland, while Kluber has faced slightly tougher opposition than Sale in a slightly tougher hitting environment. The defense, park, and opposition account for roughly one win, which helps explain the relatively small difference between the RA/9 WAR and the much larger one at Baseball-Reference.

Finally, Baseball Prospectus uses DRA. As you can see, the numbers here essentially even, with Sale’s greater workload accounting for his advantage.

As far as team defense goes, it does appear as though the Red Sox have been superior to the Indians this season, both by UZR and DRS.

Both methodologies place the Red Sox’ defense ahead of the Indians’ by roughly the same margin; however, it isn’t fair to assume that because an entire team has been better on defense over the course of the season, that the defense was better for two players’ particular starts over the course of the season. At FanGraphs, the question of defense is left out almost entirely for purposes of WAR. Baseball-Reference makes some assumptions, but makes an effort to include defense in pitcher value. There’s a three-win gap between FanGraphs and Baseball-Reference and a large portion of that gap is based on what happens when a ball is hit in play. To further examine that area this season, we can turn to Statcast.

Below, we have xwOBA and wOBA for each pitcher. Strikeouts and walks are included and xwOBA does not factor in the difficulty of the opponent, but it should remove any effect of ballpark and defense. All stats below come from searches performed at Baseball Savant.

Sale has allowed the higher wOBA, but actually a lower xwOBA. What does this tell us? It suggests that, when you include quality of contact along with strikeouts and walks, stripping out ballpark, defense, sequencing, and opponent, and don’t worthy whether a batted ball actually left the yard, Sale has been slightly better than Kluber this year — maybe to the tune of four runs, or roughly 0.2 per nine innings. If you believe Kluber has faced slightly tougher opponents this season, that 0.2 number is probably half that. The difference above shows that Sale has been a bit unlucky while Kluber has been basically neutral to slightly lucky. Kluber’s runs allowed look better not because he has yielded weaker contact, but because his defense and sequencing has been slightly better than Sale’s.

What happens when we want to focus more on that contact? Here are those numbers, taking out strikeouts and walks, leaving only batted balls.

The disparity between the two players for xwOBA on contact again gives a slight edge to Sale, but the gap between the actual results grows larger. Kluber has been a bit luckier on those batted balls. What if we remove homers?

Here we see Kluber actually has a slight edge when it comes to balls that remain in the park. We know Sale has allowed fewer homers (18 to Kluber’s 20) and his xwOBA on homers is about 80 points lower than Kluber, which helps explain how Sale’s overall xwOBA on contact is lower but not lower when homers are removed.

That isn’t the real story of the last table, though. Sale’s wOBA is nine points lower than expected, which shows some solid defense from the Red Sox or perhaps a little luck on the directional angle of batted balls (both pitchers have similar pull/center/oppo profiles fro both grounders and flies). Kluber has a whopping 38-point difference between his actual wOBA and his xwOBA. Perhaps Chris Sale has benefited some from defense, but it would appear that Kluber has been a much bigger beneficiary this season.

We can repeat the above exercise and look at only batting average and we get the same results:

This post isn’t a battle of WARs, or at least it isn’t intended to be. It’s intent has been to help find a little bit of context for the cases that Chris Sale and Corey Kluber have for the Cy Young. Kluber has allowed fewer runs than Chris Sale. If that is what’s most important to you, then that makes a pretty definitive case. He hasn’t actually pitched any better than Chris Sale, however. The Statcast data suggests Kluber has benefited more from his defense than Sale has and not the other way around, despite what our team defense tells us. Depending on the weight you give to a pitcher’s control over what happens on a ball in play, the gap between Sale and Kluber in WAR maybe shouldn’t be as high as it is for you, but even accounting for a lessening of that gap, Chris Sale has probably produced more value than Corey Kluber based on their performances this season.