It's been a rough season for Masahiro Tanaka. He's pitched to a 4.93 ERA, 4.51 FIP, 3.50 xFIP. The FIP/xFIP gap is massive, and notable. The gap is entirely explained by home runs - Tanaka has gotten killed so far this season, which we all know very well. The question is whether or not Tanaka has been unlucky or just bad.
A few years ago, answering this question would be impossible, but we now have excellent Statcast data. Statcast puts out a statistic called xwOBA, which estimates the expected wOBA of batters against a pitcher based on walk and strikeout rates, exit velocity, and launch angle. The average wOBA for a pitcher is 0.328 working out to about a 4.56 ERA. Tanaka's wOBA so far is 0.335 on the season.
What about his xwOBA? He's just above average at 0.319, suggesting he has been slightly unlucky. What about the trend?
Tanaka wasn't bad in the first few months of the season, averaging a xwOBA of about .300. Everything fell apart in mid-May, and the average batter became Aaron Judge for a month. He remained below-average through mid-June. Since then, he's been good to great.
I think it's safe to say that Masahiro Tanaka is back. He was definitely bad, not just unlucky, earlier this season. Now he's Masahiro Tanaka again: near ace.
In the short term, Tanaka becomes the best #3 starter in the American League. Only 12 starting pitchers hold an ERA under 4.00, and Tanaka is currently performing well below that benchmark. In the long term, opting-out becomes much more attractive for him. Tanaka can claim $67 million over three seasons after the opt out, but an effective Masahiro Tanaka is worth much more than that. Of course, any scenario where Tanaka opts out is one where he is good enough to help push the Yankees into the postseason, so we'll take it.