About that Mariners rotation...

On Thursday, I was a guest on MLB Network's MLB now program, and Drew Smyly was in the news. The southpaw had just been traded to the Mariners in a three-way deal involving both the Rays and the Braves. The deal drew applause from these pages and elsewhere. FanGraphs now projects the Mariners to win 84 games this season and be a wild card favorite. Smyly stabilizes the rotation, relief piece Shae Simmons gives the bullpen length, and everything seems to be in the right place.

I can’t help but feel, however, that perhaps recency bias is giving way to primacy bias in this case. Far be it for me to poo-poo a projection system, but this one just feels wrong to me. In my view, the Mariners rotation is not going to be a strength for them this year, and Drew Smyly isn’t the missing piece to put them over the top.

First, to the projection. FanGraphs uses FIP as the foundation of its fWAR metric for pitchers, and follows similar principles when projecting individual pitcher performance in its combination of the ZiPS and Steamer systems. However, as has been demonstrated repeatedly over the last two years, projection models based around Deserved Run Average (DRA) and Contextualized Fielding Independent Pitching (cFIP) are much more precise, given their ability to isolate the actual performance of a pitcher, irrespective of their parks, defense, catchers, and other critical contextual elements.

Only PECOTA, Baseball Prospectus’s projection system, uses DRA for pitchers. Those numbers won’t be made available to the public for another few weeks. So, we’re left to speculate slightly at what the model will make of the Mariners starters.

Let’s speculate anyway. Both DRA and cFIP are the most predictive publicly available pitching metrics, and might be able to tell us what to expect from the likes of Smyly, Felix Hernandez, Hisashi Iwakuma, James Paxton, and Yovani Gallardo. Here’s a quick-and-dirty table of each of their numbers from last year. (As always, cFIP and DRA- are scaled to 100, and a point above or below is a percentage point above or below league average. Lower is better.)

This is...not promising. Only Paxton appears to be in the very-good-to-great range, and he’s the youngest guy in the rotation. When PECOTA is released, it will consider three years’ worth of data, which will certainly help burnish Hernández and Iwakuma’s numbers. Paxton will also probably get hit, since he only has a year of major league seasoning.

Here is my fear, though. Both DRA and cFIP are extremely reliable year-to-year, and are more predictive than any other measurement we have of what next year might look like for a pitcher. Furthermore, both Hernández and Iwakuma are old and have very recent struggles in their pasts. What if below-average is the new normal for both?

If that is the case, then suddenly you’re not asking Smyly to eat innings in the No. 4 spot. You may be asking him to carry the rotation with Paxton. And let’s not forget that it’s possible Paxton may regress, given his elbow injury last year and relative inexperience.

Smyly will not necessarily be a star in Seattle, either. Yes, Safeco Field is still a pitcher’s park, but Smyly gave up 1.6 home runs per nine innings across 2015 and 2016. He did that in Tropicana Field! The likes of Jarrod Dyson and Leonys Martín will help run down some gappers, but Safeco was a right-hander’s home run haven last year. Against a fly ball-prone southpaw like Smyly, that feels like a witch’s brew.

Suddenly, the 2017 Mariners really are the 2014-15 Royals. They have speed, outfield defense for days, an electric bullpen, and a sketchy rotation. It worked for the Royals, and they blew past their expected win totals, not to mention scooping up a few trophies in the process. The Mariners are indeed in the AL wild card hunt, but they also could crash and burn just as easily. Given the age of the 25-man roster and the barrenness of the farm system, Jerry Dipoto’s necessary all-in gamble for 2017 might come up snake eyes.