My caveat on Tyler Duffey

Earlier in the week I voiced optimism about Tyler Duffey, with a caveat I promised to explain "tomorrow." Then Monte Irvin died, and I got sidetracked for a couple of days.

Back, now, to Duffey.

He made 10 starts, 58 innings, for the Twins with an ERA of 3.10. Add up his innings in Double A, Triple A and the majors, and he threw 196 innings for the season, a significant increase in his workload from 2014.

What's more he relied heavily on his curve ball, at least in the majors. According to Baseball Info Systems, as recorded in the Bill James Handbook, Duffey threw 58 percent fastballs, 40 percent curves and 2 percent change-ups in the majors.

That's a high percentage of curves. Of the qualified starters (minimum 162 innings) in the majors, the highest percentage of curves was Jose Quintana of the White Sox, and he threw 30.9 percent curves. Only a couple of relievers matched Duffey in percentage of curves thrown.

Anecdotally, heavy use of the curve is linked to arm issues. Steve Stone, longtime broadcaster, has said many times that he knew during his 25-win season in 1980 (for which he won the Cy Young Award) that he was throwing his curve so often that it was going to shorten his career.

Bert Blyleven, whose curve was legendary, might dispute that notion, but Blyleven frequently talks as if pitching injuries are purely random. And it's always important to remember that Blyleven himself was unique in baseball history -- reached the majors as a teen, threw 278 innings at age 20, had no significant arm issues until his 30s and remained a a fastball-curve pitcher into his 40s. There is literally nobody else like that in baseball history.

Duffey's average fastball, according to the Handbook, was a bit above 90 mph, which is hardly prime velocity these days. He obviously doesn't have a lot of faith in his straight change, a trait common to pitchers with standout curves; it's more common to see a pitcher with a high-quality change use a slider for his breaking ball than a curve. Blyleven, for example, never really had a good change; he could change speeds on the bender, and that was enough.

There have been plenty of successful fastball-curve pitchers over the generations, but I don't know that many of them threw 40 percent curves. (I'd love to know what percentage of Blyleven's pitches were curve balls.) And there's my caveat on Duffey: He exhibited in 2015 a mix of pitches unique in today's game, and I'm not sure it's sustainable.