Byron Buxton's Delayed Stardom & Small Samples | The Process Report

Byron Buxton’s Delayed Stardom & Small Samples

Most fans are well familiar with the centerfielder for the Minnesota Twins. Byron Buxton has seen his name atop the marquee for prospects going on many years now. He’s the latest in a long line of glove-first true centerfield prospects that the Twins have developed including Carlos Gomez, Aaron Hicks, Denard Span and Ben Revere.

Minnesota Twins centerfield production since 2008

Span was the only one that played as an above average hitter with the team, though all of them went on to greater success after the team gave up trying to develop the bat. Buxton had to feel that he was on a similar path after showing up on plenty of webgems while making outs with aplomb. Then he started to hit a little. Then a lot.

Looking at his rolling average wOBA you can sense a replacement level hitter that enjoyed one good spike to close out 2016, and looks to be doing something pretty similar late in this current year. Thing is, wOBA is a great metric, but it has no idea whether a batter earned that double or that he should have gotten credit for the brought back homer. This is where park adjustments and regression on balls in play can help mightily. Using past results on the various exit velocity and launch angle combinations you can get a good idea of what a batter should have earned on that ball in play. It’s not perfect. Speedy batters and those with an ability to move the ball around laterally are going to be a bit underrated by this. Dead pull hitters that can’t move a bit overrated, but this is as good as it gets, in my opinion.

Plotting his actual park-adjusted wOBA (awOBA*) and true park-adjusted wOBA (twOBA*) you can get an idea for how his actual results compare to what he should have expected. Starting with the expectations you can see a ton of variance early on as his slow start led to a hot streak, but then right back into the basement. The second half of this season has seen him mostly level off as a league average hitter, if not a little better, and his most recent stretch has taken that up another notch. His actual results, however, are all over the place. Especially during the second half of this season where he has been fairly consistent by twOBA*, while his awOBA* shows massive diversions. First being very much unfortunate, but then being much moreso of late. You can say he’s hot now, but you should also mention that it’s making up for his cooler stretch earlier.

Both of these metrics are made of various smaller components that, like a small block engine, can be easily disassembled to check where any flags might be getting tripped. True wOBA* is comprised of both balls in play and those plate appearances that do not end up in a ball in play like when the batter walks, strikes out or is hit by a pitch. The former can be measured using my park-adjusted expected wOBA on balls in play (xwOBA*), which as previously noted is based on the batter’s displayed exit velocity and launch angle. Let’s start there:

Buxton has shown rather pedestrian exit velocity over the course of the season, though it has been ticking up fairly linearly. Until he fell off a shelf recently and then immediately picked himself back up as if he didn’t miss a beat. Conversely, his launch angles have mostly trended down as the season has moved along, until, again, that most recent period where he seems to be elevating the ball more often. Without having watched him more than here or there I would think Buxton would maximize his abilities by keeping his angle a little lower than those looking to jack taters. Seeing him have a little more success in the second half with that lower angle may be confirmation bias or it may be the path that helps him optimize his skillset. As a fast runner that isn’t showing overwhelming exit velocities he would seem to be at his best when he is spraying liners around the yard and stretching them into extra base hits from time to time. Let’s plug these values into the algorithm to see what his xwOBA* looks like based on these inputs:

Over this second half with the progressively higher exit velocities and lower launch angles you can see that his expectations show an average or much better hitter. We see both sides of the luck coin of late first with his underperformance and then his more recent overperformance. I’d expect him to converge with his expected over time, if not continue to overperform slightly as he could be the type of hitter described above that tends to get overrated by this stuff. Other than the start of the season and that ensuing dip a month later he has shown the kind of velocity and angle off the bat that makes him look like a useful hitter with one of the better gloves on the planet. That’s really good! Unfortunately, there is another component to hitting, and that is all of the walks, strikeouts and hit by pitch that doesn’t lead to a ball in play:

This is where things get a little ugly. You may have noticed wide discrepancies between his twOBA* and xwOBA* in the above charts. This is the reason for that. The former factors all events while the latter only considers balls in play. Striking out a third to nearly half of the time is going to suppress the good works he is doing when hitting between the chalk because it replaces balls that sometimes, or even often, go for hits with empty zeroes. He has done well to slowly cut back on the strikeouts, but even of late when he has looked solid you can still see him striking out 20-30% of the time. It’s encouraging to see the peaks getting a little lower and the troughs following suit, but he still has a ways to go if he wants to optimize his skillset. You can strike out this often if you’re putting up the kind of exit velocity that leads to a high frequency of extra base hits, but he isn’t there yet so it would be nice to see this lower a bit.

His walks, on the other hand, have mostly settled in at a close to double-digit rate. They were higher earlier, but with all those strikeouts I would think he was being overly passive, getting into deeper counts and eventually chasing his way back to the bench. There is nothing wrong with being a league average walk guy, which is mostly what he looks like. Seeing him maintain that while continuing to erode the strikeout rate would be very encouraging.

For Buxton to fully evolve he will need to strikeout a little less while maintaining both the walk rate and the demonstrated ball-striking ability. Progress is plainly evident, and it helps to remember that he is still incredibly young for the Show. There is a universe where he continues to increase his exit velocity, and if he can do that it doesn’t really matter what happens to his launch angle as long as he isn’t playing to the extremes. If it increases you’re talking about more extra base hits, and if it continues to lower you’re talking about infield singles and balls beaten by defenders before they can move. This will dictate whether he is thought of as a good average, ok power guy or an ok average, good power guy. I’d prefer to see the former while he’s young, and then transitioning into the old guy skills later in his career much like another great Twins centerfielder, Torii Hunter.

I think the data shows that his most recent scorch stretch is pretty well unsustainable, but there is a lot of good to be seen here, and certainly more than what was showing up at the start of the season.