If there is one thing that drives Blue Jays fans nuts about Kevin Pillar, it is an alleged lack of discipline at the plate that results in very few walks and limits his on base potential. His walks (or lack thereof) are practically a punchline even. If he would only take a few more pitches, and work the count a little more, he would be much better off—or so goes the conventional wisdom.
And yes, in an ideal world, Pillar would walk more. But we don't live in that world, and the reality is that Pillar has a significant flaw, in that he cannot consistency identify and lay off offspeed pitches. Consequently, he has a nasty tendency to flail at pitches out of the zone and look completely hopeless in the process.
This is usually a recipe for a high strikeout rate—think Juan Francisco—and that can only work if it's paired with huge power, which is definitely not Pillar's game. Yet since becoming a regular at the start of 2015, Pillar has managed reasonable production. The chart below shows Pillar compared to the league average for overall offensive production (wRC+), as well as the four major determinants of that: strikeouts rate, walk rate, getting hits (BABIP), and power (ISO). It also shows his rank among the 245 players with 500+ plate appearances, and the percentile at which he ranks.
Pillar's production is essentially right at the lowest quartile, but this is just looking at regulars which is a higher comparison group (the median wRC+ is 103). Unsurprisingly, Pillar has one of the worst walk rates in MLB, at a tick under 4%. His BABIP is almost exactly average, with below average but not insignificant power. But where he really stands out is his ability to avoid strikeouts, at a rate almost 6% less than average. This is easily his greatest strength as a hitter.
So how does a free swinging player who is very vulnerable to breaking balls avoid striking out so well? To strike out requires at least three strikes, and at least three pitches. Swinging early in the count and looking to put the ball in play is the antidote, especially in the case of Pillar since most pitchers lack the command of their secondary pitches to use them early in the count and rely on fastballs earlier in the count. And Pillar does exactly this, swinging about 50.5% of the time, or about 4% above the league average of 46.5%. Perhaps not coincidentally, he also puts the ball in play about 4% more often, 22.5% to 18.5%.
The flip side of course is that walking requires taking four balls, so an aggressive, contact oriented approach will curtail opportunities to walk. Perhaps this seems excessively obvious, and it is elementary, but the point is that for Pillar, not walking and not striking out are fundamentally linked—and barring a sudden improvement in pitch identification and recognition, will remain so. Therefore, the question is not whether Pillar could or should walk more, but whether the tradeoff is worth it.
Below is a table showing the run expectancy relative to average for three events in 2015–16: walks, strikeouts, and the average Pillar ball in play (based on his frequency of 1B, 2B, 3B and HR). I've also included wRC+ since that may be a more familiar scale.
Again, relative to the average to all events, a walk is worth +0.30 runs, and a strikeout about −0.26 runs. Pillar's are almost exactly in the middle (above the average of all events, but below the average for league batted balls, I should mention, especially in light of the first chart showing him below average).
The key insight here is that for Pillar, trading one fewer walk for one fewer strikeout is exactly a breakeven proposition: 0.30−0.26 is 0.04 runs, which is same output as two of his average balls in play (leaving aside hit by pitches, the only alternative to walking or striking out).
Going back to that first chart, Pillar has walked 4% less than league average, but struck out 5.8% less than league average. Those differences—again, those two are linked—create value. If Pillar was instead league average at both, his wRC+ would fall from 91 to 88 (about 2–3 runs over a full season).
In fact, we can go further: if Pillar could maintain that tradeoff, the ideal would be that he'd never walk. At that rate he's strikeout about 9% of the time and put the ball in play 90% of the time (figure a 1% HBP rate), and have an expected batting line (based on 2015–16 production) of .289/.300/.429 (that's 15 more points of BA, −7 OBP, and +26 SLG).
There is one further question: having established that Pillar's plate outcomes are actually quite decent, are they optimal? If he swung less, could they be even better? This is actually quite tricky and involved, and I don't have a definite answer, but I don't think there's anything major being left on the table.
For starters, it's worth pointing out that Pillar is not actually a completely undisciplined free swinger:
With no strikes, Pillar actually swings less than league average. This is interesting, because "no strikes" is first pitches and counts where the batter is ahead, but this passivity does not extend to other counts like 2-1 and 3-1 where Pillar is very aggressive. If anything, there might be an argument Pillar should swing more, at least on the first pitch, since he may be taking good pitches. His call% (called strikes/pitches taken; higher is better for a pitcher and lower for a batter) is 54%, compared to 46% league average, so it seems like the extra pitches taken are strikes.
With one strike, Pillar really starts swinging, my guess is to avoid falling to two strikes. It's also worth pointing out he's a phenomenal one strike hitter, which a career BA of .367 and SLG of .545 (compared to league average around .320/.490, and remember that overall Pillar's batted balls are below average).
With two strikes, he's forced to defend and that's where he's at his most vulnerable to expanding the zone. What if he consciously approached two strikes by swinging less? I played around with some numbers, and holding all else equal, swinging 5% less (still above league average) would improve his wRC+ by about 1 point, adding about 0.7% to his walk rate and only 0.3% to his strikeout rate. That's a good tradeoff, though relatively modest. The bigger issue though, is that all else would not be equal, and pitchers would in time counteradjust. But it may be something to explore.