Justin Smoak in 2017: What's different, and what's not

If he's not the single most pleasant surprise for the Blue Jays in 2017, Justin Smoak is certainly right near the top of the list. At age 30, he's finally fulfilled the promise that made him the 11th overall pick in the 2008 draft. Even more fortuitously, thanks to the much maligned contract extension he got a year ago, his services are now cheaply controlled through 2019.

How did he get to this level of production? And perhaps more to the point, is it a temporary spike like the first half that Michael Saunders put up a year ago (in which case, as has been increasingly suggested, it might be a good idea to sell high)?

I want to start with what I concluded last August, trying to make sense of the extension:

If Smoak had maintained his career strikeout and walk rates entering 2015, while benefiting from the same rate of production on balls in play [as in 2015-16], he'd be an above average regular [with] something like a (park adjusted) 125 wRC+....

The best reason for extending Smoak is that he's finally living up to his prospect billing as a guy who would do significant damage when he contacts the ball, and it takes time to make the adjustments to fully harness raw power in a good hitting environment. Unfortunately, increased lack of contact looks like it probably will be an unsurmountable obstacle.

I'll come back to that last part, which I intended to further explore last winter and just never got around to, but let's start with the first part. Many viewed Smoak in 2015-16 (99 wRC+) as the same unproductive player he was before (94 wRC+), which simply doesn't cut it for a first baseman. While superficially true, that belied a significant underlying shift in Smoak's ball-in-play outcomes and his plate outcomes:

This is the same chart from last August, extended through the end of 2016. Smoak's ball in play outcomes dramatically improved from below average to very good. These figures are not adjusted, so some of this is generic home park effects. Some of it is division park effects (beyond Seattle/Toronto, the AL East is more favourable than the AL West, and I don't think this is captured in adjustments). Some of it is player specific park effects for a fly ball hitter. Some of it is frankly better luck. But a good portion of it too was Smoak simply making better contact.

Unfortunately, his strikeout rate ballooned driven by a much lower contact rate. We can all remember Smoak flailing at balls out of the zone. That not only tanked his production on plate outcomes, but negatively shifted the mix between plate and BIP outcomes, crowding out the highly productive balls in play (67% to 60% of the total). Those combined effects almost entirely offset the gains on balls in play.

Now let's add 2017 to that chart (to the All-Star break, as 2017 stats throughout are):

On the ball-in-play side, Smoak's production has jumped once again. But really, the 2017 improvement is more incremental than revolutionary: having made the leap from mediocre to very good, he's now in the excellent range.

Drilling down a bit, the difference is entirely about more home runs: Smoak's rates of singles and doubles/triples to balls in balls are exactly the same (22% and 5% respectively). The natural inclination might be to dismiss this as a fluke, but if anything it's the opposite. Smoak is pulling about 25% of his fly balls in 2017, up from about 21% the last couple years (though well below his early career levels). That really matters, as FanGraphs split tool shows for the league in 2017:

• Pulled fly balls: 33.5% HR/FB
• Centre/opposite field fly balls: 6.9% HR/FB

In fact, his rate of home runs on pulled fly balls is actually down in 2017, at 40%, which is a sustainable level for hitters with good power, whereas it was over 50% in 2015-16 (which is not really sustainable).

But even if one expects some regression to his ball-in-play output, the major takeaway here is we're certainly not dealing with a one year fluke here. For the last three years - 1,000 plate appearances and 635 balls in play - he's been highly productive on balls in play. Of the 202 qualified hitters from 2015-17, Smoak ranks 31st on production in balls in play. That he might be further tapping into more power is gravy on the top.

In fact, balls in play are not the main driver of Smoak's success. Below is a "waterfall" chart breaking down the increase from 99 wRC+ from 2015-16 to 145 wRC+ in 2017:

Here the real 2017 difference is clear. BIP outcomes are the single biggest factor, but they represent more of the total outcomes. On a per outcome basis, the improvement at the plate is very similar to the balls in play. The plate outcomes improvement basically amounts to a massive reduction in strikeouts (19% vs. 30%, slightly offset by walks down slightly from 10% to 9%). Fewer strikeouts enable more balls in play, and that change in mix is the third column.

In total then, the massive reduction in Smoak's strikeout rate has directly resulted in just under two-thirds of his overall gain in offensive production (to say nothing of the fact that chasing fewer bad pitches usually means getting better pitches to hit). If we're looking for something to worry about being a fluke, this the area for concern.

What's remarkable about the change in strikeout rate is how singularly it's been driven. Usually, a change of this magnitude is due to wholesale change. But Smoak is not swinging significantly more or less, in or out of the zone. Looking at the Statcast data, his ratio of balls to called strikes is almost unchanged, so it's not that he's more discriminating. Instead, it's all about swinging and missing less, both in and out of the zone. Overall, his contact rate has increased from 72-74% to above 80%.

What's interesting is that's not that far off his contact rate from 2012-13 of 78-79%, so he's not at an unprecedented level by own standards. However, the league contact rate has been falling pretty quickly in recent years, so it's more accurate to look at Smoak's ranking against the league across time:

Contact has never been one of Smoak's strengths. Prior to 2017, his best year was 2012 when he was close to the middle of the pack, before sliding down. Now in 2017, for the first time in his career he's well above the median. That's a pretty stunning gain.

On the other hand, I've always thought of contact as something pretty hard to "fake", which is why strikeout rates tend to "stabilize" quickly and in small samples. It's why I was so down on Smoak in the first place last year. It's pretty rare for players, especially at age 30, to suddenly figure out how to avoid swinging and missing.

To add some historical context: since 2002, there is plate discipline data from Baseball Info Solutions. It tracks the automated data pretty well. Since then, there have been 3,441 instances of a player recording 300 or more plate appearances in consecutive seasons. Smoak's 8.5% increase ranks as the 11th highest.

Can other players who've had large spikes in contact rate inform us about what the future might hold for Smoak? I will return to this in the very near future.

***

To conclude:

• What's not different about Smoak's 2017 is that he's highly productive on balls in play
• What's a little different is him pulling more fly balls and the consequent increase in home runs
• What's very different, even historically so, is the change in his contact rate