Scherzer Reaches New Heights

Steamer, a reputed projection system, offers predictions of the reliability for each player’s projected stat line. (These can’t be found in FanGraphs’ database, but they are available in the raw download from Steamer’s website.) James Shields‘ Streamer projection scored the highest reliability probability for the 2015 season. Shields was basically a 4-WAR starting pitcher for the eight years following his 2006 debut; when the San Diego Padres acquired him, they expected to acquire consistency. Alas, Shields’ MLB-highest 80.8% reliability was, and still is, understandable.

Except Shields cashed in a most unusual age-33 performance, and that’s why more, not fewer, numbers make the sport more special: you can marvel at and appreciate the game through any of an infinite number of lenses, and it never gets old. Neil Weinberg astutely detailed this anomaly, so I will politely not rip open San Diego’s wound as it heals.

I will, instead, turn my focus to Max Scherzer who, similarly to Shields, disrupted an 11th-best 79.3% reliability score on his projection. Unlike Shields, however, Scherzer made even better on his promise.

Scherzer has always been valuable. I think some people forget. Or maybe I’m projecting, because I definitely forgot. Aside from a 2008 partial-season debut, Scherzer has never started fewer than 30 games or generated fewer than 2.2 WAR. Scherzer took his first big step forward in 2012 when he ramped up his strikeout rate from 20.9% to 29.4%. Then he took another step in 2013 when he increased his per-start efficiency from 5.9 to 6.7 innings per game,.

Then last year happened. Scherzer won his fewest games since 2010 as the ace of what was ultimately a profoundly disappointing Washington Nationals squad. Those are the kinds of things that can distract even the most staunchly sabermetric fan from Scherzer’s greatness.

Because he was great, and he was great in an entirely new and impressive way. Sure, it helps that he recorded a career-best strand rate (LOB%) up near 80%. Sure, it helps that he benefited from an unusually low .268 batting average on balls on play (BABIP).

But it doesn’t explain his walk rate (BB%). A walk rate that, after being barely “above average” for four seasons, catapulted into “excellent” territory seemingly overnight. Only Bartolo Colon walked fewer hitters on a rate basis in 2015 than Scherzer, and Colon recorded only half as many strikeouts.

Scherzer’s release points hardly budged. His pitch movements, likewise. It seems like less a mechanical concern than a velocity one: Scherzer gained non-trivial velocity on three important pitches in his arsenal (+1.3 mph on his fastball, +1.3 mph on his slider and +2.7 mph on his curveball).

I won’t say the velocity bumps are causal, but it’s hard to deny the correlation between the increased velocity and more swings induced on pitches outside the zone (O-Swing%) as well as less contact on such pitches (O-Contact%) for all three of the aforementioned pitches. Consequently, each pitch recorded career-best swinging strike rates (SwStr%), en route to career-best strikeout rates on each pitch, en route to a career-best strikeout rate (K%) north of 30%.

To say those gains were only marginal would sell Scherzer short, but he has been, for many years now, an elite strikeout pitcher. A 3-percentage point increase in strikeouts would likely yield much more marginal value for a low-K arm such as Mark Buehrle rather than Scherzer, so the extra few strikeouts are not as big of a deal.

No, Scherzer found more than just extra juice, and not the illegal kind: he found control. He pounded the strike zone with his fastball, and he spotted his slider. Overall, his zone rate increased 4.6 percentage points, and his rate of first-pitch strikes (F-Strike%) also spiked to a career-best — and MLB-best — 71.3%. (I’m hard-pressed to find a different way to say “career-high” or “career-best,” but it’s not my fault Scherzer did so many things in such impressive ways.)

If you throw more strikes, you’ll probably walk fewer hitters. If you induce more swings on the remaining pitches you don’t throw for strikes, you’ll probably walk fewer hitters. And if you induce less contact on those remaining pitches you don’t throw for strikes that induce swings, you’ll probably strike out more hitters. Scherzer checked one of those three boxes in 2012 when his strikeout rate leaped to elite levels. In 2015, he checked all the boxes — and posted his best season ever in terms of FIP, xFIP, WAR, you name it.

On the tip of your tongue: is it sustainable? I don’t know, partly because Scherzer’s combination of velocity and control leads us into somewhat unfamiliar waters. Since 2007, only two pitchers recorded better single-season strikeout-to-walk ratios than Scherzer’s 8.12 K/BB: Phil Hughes in 2014 (11.63) and Cliff Lee in 2010 (10.28). Neither pitcher topped 92 mph with their fastballs. Scherzer throws 94-plus. Harvey’s averaged 95.4 mph on his fastball in 2013, but his 6.16 K/BB is decidedly less impressive.

Thus, finding comps by the ol’ let’s-eyeball-it method doesn’t work so well. I could do (maybe) the next best thing: investigate typical changes in year 2 following big spikes in year 1. I won’t bore you with the minutiae — I hope that you trust me — but this method is not very helpful either in terms of statistical significance. The average qualified starter — in other words, looking at mean outcomes — exhibited no losses in Zone%, O-Swing% or O-Contact% the following year. The distribution, however — and this is what’s important — the distribution of the changes following a big improvement is so wide, so volatile, that making any kind of concrete statement about what Scherzer might do next year — without the aid of more measurements from other variables — would be utterly unwise.

Alas, we are resigned to embrace the unknown. Steamer expects Scherzer’s strikeout rate to remain elite and his walk rate to split the difference between what he did for a long time and what he did in 2015. And that’s pretty reasonable. As a risk-averse owner, I would be reluctant to assume he sustains the same kind of elite control. When Scherzer started fooling more hitters in 2012, he was able to maintain that momentum, so it wouldn’t be unheard of to see him do it again. But this much improvement — well, I guess we’ll see.

Regardless of how you approach things — whether you expect 2014 Scherzer, 2015 Scherzer or someone in between — Scherzer seems like not only an easy lock for the top-10 starting pitchers but also one of the safer 2016 investments in terms of durability and floor production outside of Clayton Kershaw.

Reliable, indeed.