A modest Jacob Turner comparison

Yes, this is another Jacob Turner article.

We all know the Miami Marlins traded the embattled starter last week to the Chicago Cubs for essentially nothing, a pair of non-prospect minor league relievers. The Fish's reason for releasing Turner was ridiculous, and Miami did not have to let go of a 23-year-old pitcher, even if the odds of him becoming a league average pitcher were relatively low. The low-upside move of keeping Turner around was always better than the no-upside move of essentially giving him to the Cubs.

But as we discussed slightly in yesterday's article about the Cubs and Marlins building teams in different fashions, the Cubs went after Turner as his peripherals were at his career peak; Turner struck out more batters, walked fewer batters, and had a better fastball velocity this season than in any other year. So it leads me to wonder how the Marlins are analyzing their players. To emphasize that point, I offer you a modest anonymous player comparison. Names excluded, and actual ERA and FIP numbers taken out. Instead, I give you them in terms of ERA- and FIP-, numbers compared to the expected league average after park adjustment. A number greater than 100 represents a percentage worse than league average and vice versa. These numbers represent data from 2012 to 2014.

In looking at these numbers, what do you see? Player 1 has an atrocious strikeout-to-walk ratio, but his ERA was good in the last three years, and his FIP was better than the other two. However, when you correct for home runs by using fly ball percentage instead of actual homers allowed, his expected ERA by xFIP looks a lot more similar to the other two. Player 2 looks like the most obvious of the three in terms of skills; his ERA, FIP, and xFIP all match up nicely. Player 3 had worse luck with his ERA, but his FIP and xFIP are also comparable to the other two players, and in particular, his performance matches up very similarly to Player 2.

That is an important point in the comparison. Player 2 strikes out more batters, but Player 3 presumably does better on home runs. The home run thing is likely due to Player 3 getting more ground balls than Player 2. Overall, however, the two players have pitched very similarly over the last three years.

The takeaway point here is that, if we look at the most predictive information on this list, the numbers that best predict future run allowance, these pitchers do not look all that different. You may say Player 1 is the best of the three, and Player 3 is the worst of the three, but the difference is likely marginal, perhaps in the range of 0.1 to 0.2 runs per nine innings.

And now the reveal:

The fact that Turner was Player 3 probably did not surprise anyone. The identity of the other two pitchers may have been a tad surprising unless you read through yesterday's article carefully. The Marlins just acquired Cosart from the Houston Astros for two of the team's top hitting prospects, including last year's first-round draft pick. This despite the fact that his walk rate is absurd, he has the lowest strikeout of the three pitchers, and that his predicted ERA based on xFIP was close to 4.30.

Meanwhile, Miami just gave away a pitcher who, despite a 5.94 ERA this season, had a reasonable FIP (the best of his career), was striking out more guys and walking fewer than ever, and had a predicted ERA by xFIP that was almost the same as the team's latest acquisition.

At the same time, Koehler has had the most predictive performance according to his ERA, since his FIP and xFIP are similar to his ERA. He has essentially played to his predicted level, making it easy for the Marlins to properly evaluate him. Thus the Fish have figured him out, correctly placing him as a fifth starter.

The problem is that the Miami front office appears to be evaluating players by: a) their scouting analysis (good) and B) very peripheral, basic statistics (bad). It is entirely possible that the Marlins watched Turner, tried to fix him, and found that they couldn't. But this attempt and mechanical adjustment flies in the face of the fact that his strikeout and walk numbers improved instead of declining, and that the only reason his ERA ballooned in 2014 was a completely anomalous .368 BABIP. That sort of number is just as likely, if not more likely, to be due to random variance and bad luck as it is to be due to poor mechanical work or some other scouting-based analysis of his game.

Turner is not a world beater right now, and his numbers do not project him to be a world beater next year. But he is 23 years old and essentially pitched at the level of Tom Koehler, whom the Marlins would not trade for absolutely nothing. Turner is probably half a rung down from being Jarred Cosart, whom the Marlins just spent a relative prospect fortune on. There is definitely a statistical argument that says that Turner was getting his game on track in the bullpen, and Miami's mistake was to not let him do that when they afforded Cosart and Koehler opportunities. The Fish may still be right in not seeing anything in Turner, but their inability to use statistical analysis of the game points to a greater concern for me than losing Jacob Turner.