clock menu more-arrow no yes mobile

Filed under:

Breaking down Kolten Wong's weak batted ball authority

What Statcast data can tell us about Kolten Wong's contact quality issues

Benny Sieu-USA TODAY Sports

So, this isn't great timing for this article, what with Kolten Wong having a fantastic night at the plate and in the field last night, leading to the Cardinals currently holding a Wild Card spot. This post was planned several days ago, so hopefully this doesn't bum out anyone excited about Wong's great night. The fan in me hopes that Wong's right at the beginning of a long hot streak. The analyst in me, unfortunately, is concerned.

You know the story: he slumped hard the first couple months of the season, with an increase in walks being overshadowed by a weak BABIP and non-existent power. Wong was demoted to Triple-A where he clobbered fools, and now is back and, on the surface at least, is hitting better:

*note: this does not include last night's game*

Wong hasn't been a good hitter, but he's been close enough to average that, combined with above-average base-running and defense at second-base, he's been about as valuable as an average regular. With his production at the plate sputtering in 2016 however, that looks much less true. It's too small of a sample to take Wong's BABIP, ISO, or HR/FB rate as the new normal going forward, but Cardinals fans have become a little uneasy about the Cardinals' $25M investment in the former first-rounder.

Time will tell us the outcome of Kolten Wong's future, but I don't want to wait that long. I've recently started wading into the new Statcast data presented on, with the goal of bettering our understanding of Contact Quality, and finding better ways to describe a hitter's batted ball profile. Last week, I looked at Matt Holliday and Carlos Gonzalez's batted ball profile and came to the conclusion that Holliday's BABIP should be much higher than it currently is. Today, we'll look at home run production, and how Wong's profile helps or hinders that.

Last week I posted a heatmap of the average BABIP of each type of batted ball, grouped by angle and velocity. We'll do the same this week, but for home runs per batted ball instead. Because Home runs are only possible at much fewer angles and velocities than hits in play, this heatmap will focus on a fraction of the area we covered last week. Here goes:

Remember, Statcast classifies batted balls as follows: Batted balls below 10 degrees are ground balls, between 10 to 25 are line drives, 25 to 50 are fly balls, and 50+ are pop-ups. Despite fly balls generally being associated with homers, you can see that quite a few homers occur on line drives, and homers generally become very unlikely before reaching 50 degrees, the separation line between flies and pop-ups.

Last week, we talked about the "Donut Hole", something fly balls (and the higher-angled line drives) display. That's where low velocity batted balls fall in between the infielders and outfielders (bloopers) and high velocity batted balls leave the park or at least land off the wall or otherwise behind the outfielders. Batted balls in between, however, are gobbled up at a very high rate by defenders. We didn't quite see the full affect last week, because we were just looking at balls in play. Here we see that you have to reach about 90 mph before homers start to pop-up more, and even then, it has only a small window in which it is possible.

We generally measure a player's HR production by his HR/FB rate, and that made sense without Statcast. Without specific angle data available, fly balls overlapped with home run possible batted balls enough that it worked. However, with better data now available, we can look specifically at Exit Angle to get an idea of the home run possibility of each angle, and gain a better understanding of what angles are best for homers. Here is average home runs per batted ball at each angle over the Statcast Era:

We saw last week that BABIP by angle has a very obvious peak, and so does Home Run per batted ball by angle, at 27 and 28 degrees. It starts building up rapidly early on in the line drive range, and falls back to earth at a similarly rapid level near the back end of the fly ball range. Let's work with the range from 18 degrees to 42 degrees, which has a combined home run per batted ball of 14.8%, and leave out the remaining angles, which combine for a rate of 0.09% HR per batted ball percentage, and represents just 101 of the 7,418 homers tracked by Statcast in my data sample.

As you can see, batted balls almost have to be at least 90 mph to leave the field in fair territory. Only 13 of 7,317 homers in this group left the bat at less than 90 mph, so they are pretty rare. Once you reach that threshold though, the chances of a homer start to increase much quicker than velocity rises.

I wanted to measure how much both angle and velocity affect a player's HR/FB rate, but as you can see, using fly balls as the denominator isn't optimal and isn't necessary in the Statcast Era. So I used home runs per batted ball in between 18 and 42 degrees inclusive, which you could also call home runs per home run possible batted ball. Those are both very clunky names though, so I am definitely taking suggestions for snappier ones. For short, we will simply refer to them as HRPBB for now.

Like last week with BABIP, I devised three metrics: angle-based xHR/HRPBB rate, which takes each batted ball between 18 and 42 degrees and assigns it a HR% based on the average home run per batted ball rate at that angle. I also did the same for a velocity-based xHR/HRPBB rate, which took the average home run per batted ball rate for each specific velocity. I then created a total xHR/HRPBB rate, which took both the angle and velocity of each batted ball into account, and assigned it a HR% based on the heatmap shown at the top of this post. I also included a home run possible batted ball percentage (HRPBB%) which states the percentage of recorded batted balls a player has in between 18 and 42 degrees. Finally, I also tracked the percentage of Home Run possible batted balls that had a velocity under 90 mph (less than 90 mph HRPBB%), and thus, were very unlikely to become homers.

Let's look at how Kolten Wong has fared at these metrics from 2015 to now, compared to the league average scores over the Statcast Era:

I think these stats sum things up pretty well. Wong is about average at getting the ball into home run possible angles, and has a roughly average distribution among those home run possible angles, but he simply lacks the Exit Velocity to do damage through the air. Wong was only slightly worse than average at getting above the "Donut Hole" in 2015, but is doing much worse in 2016.

So what we have in Kolten Wong is a hitter who is average at generating the angles required for home run production, but lacks the ability to get above the velocity threshold required to hit for real power. This affect was small in 2015, but it has ballooned this year. It's also discouraging to see that things haven't really improved since returning to the big league club,

There's better news on the BABIP side. Using my xBABIP calculator from last week, Wong's 2015 xBABIP came out at .313. For 2016, it was .285 before his Minor League assignment and .302 after. He's shifted against 19% of the time. That's not all that large, but it should take a bit of a bite out of his BABIP ability. However, his speed should help when he's not shifted against. In light of the poor power numbers, it's a relief to see reason for the BABIP to bounce back.

Between last week and today, we see the 90 mph threshold is much more important for flies than for line drives and grounders, where Exit Velocity and value have a more linear relationship. To further illustrate this, here's three buckets of batted balls, as well as their wOBA by exit angle: one is all batted balls recorded by Statcast, one is all batted balls over 90 mph, and the other is all batted balls under 90 mph.

As the graph indicates, home run possible batted balls have the largest difference in value between above 90 mph batted balls and those below. What that means is that hitting at a high velocity at those angles is more important than any other range of angles you could construct. If Kolten hit mostly grounders and line drives, the affect of his reduced velocity wouldn't be nearly as large.

Wong's a guy that coming up was described by scouts as having "some pop", though it usually came with the qualifier of "for a smaller guy/second baseman". What's more interesting is this analysis might not be an indictment of Wong's raw power as much as his inability to square the ball up.  To see what I mean, we can look at Wong's 2016 angle profile, provided by

In the first graphic, we see that Wong has a spike in batted balls right at the 30 degree mark. That angle is passed the peak HR production angles of 27 and 28, but they're still up there. In the second, we see that he is only hitting those balls on average a little over 80 MPH, which means a lot of those batted balls are ending up in the Donut Hole, where batted balls go to die. I think what's also important is how spread out the distribution of batted ball angles are. It might not look like it to you if this is the first time your looking one of these, and I'm certainly no expert as I've only looked at about a dozen different players' angle distribution in this way. But this is one of the worst ones I've seen in that short time of looking at them. Stephen Piscotty has the best that I've seen, so let's look at that for reference:

Compared to Wong, Piscotty has much fewer grounders under -5 degrees, which grade out weakly unless hit very hard. Piscotty also hits much fewer balls over 30 degrees, when home run production has already peaked and BABIP is non-existent. Piscotty hits the ball harder across the board. But I wonder, is that related? Perhaps Wong is losing Exit Velocity because he's not squaring up the ball very well? OK, last angle profile, I promise: here's Wong in 2015:

It's still not as good as Piscotty has been this year, but in 2015, Wong's batted balls were more tightly grouped together, with less pop-ups and less low-angled grounders. Hypothetically, this could be evidence that Wong hit the ball squarely more often in 2015 than 2016. On the velocity side, I don't know what's going on around zero degrees, I assume that is some error with the data, and Kolten Wong wasn't hitting those balls at an average of about 5 MPH. It does reminds us that this Statcast is still new, and it's based on new technology, so it's not going to be perfect right off the bat. Regardless of some kinks, I still think Statcast gives us much more than we knew before about batted ball quality.

So, with the caveat that we still don't know how predictable this new Statcast data is, what conclusions can we draw from this? At the very least, we can say that Wong isn't getting unlucky in terms of home run production. He's simply not going to hit for much power if he can't get his batted balls over 90 MPH more often. And even when he was getting them over 90 MPH more often (in 2015), it still wasn't enough to make him even close to average in expected home run production.

Some might say this backs up something fans have been saying for years about Wong: he should change his swing and try to hit more grounders and line drives, taking advantage of his speed to hit for a higher average and more doubles. Maybe get rid of that aggressive leg kick. I don't agree with that. Kolten Wong's swing is Kolten Wong's swing, and it's something that has evolved with him for over a decade. Wong may not be the type of player to completely change his swing. Regularly people depend on muscle memory, baseball players are slaves to it. Even if Wong were to attempt to do so, it should definitely take place in the off-season, not during a Play-off race.

What I can say, is that it makes me a bit more pessimistic about Wong's future outlook than I was beforehand. His extension for $25M over five years (more like $23M over 4 after this year, plus an option year) certainly isn't something that's going to restrict the team's ability to make moves, but it also looks a bit worse with Wong's weak performance thus far. And while it's still only been 200 PA, and Wong may very well just be in an extended slump, the expected power production still didn't look like much in 2015 either.

Some may object to trying trade Wong at the deadline, and call it selling low. But perhaps it would be selling while a team can still believe in his future. Wong does have first-round pedigree and plays an above-average position on the defensive spectrum (pretty well by my eye). He's also locked into a contract that is still very cheap by baseball standards, and already has two roughly average, mostly full-time campaigns under his belt.

It's hard to judge how teams would value him due to the range in his performance, but a team looking to compete over the next four to five years could view Wong as someone they could buy-low on now and get average production for much lower than the cost of average players in free agency. The cash-conscious Rays and their revived star Evan Longoria come to mind, though of course the Cardinals would have to add on much more talent to make it happen.

Just 200 PA of bad production shouldn't cause the Cardinals to try to just dump Wong's contract, especially without knowing how predictable these numbers are. That would be a rash decision, particularly because this analysis hinged on home run production, and Kolten Wong could still end up as a perfectly average player even with lackluster power. The contract is still a nice surplus even if you think he only has a 50% chance of being average going forward, and the worst case scenario simply isn't all that bad at that level of investment. However, he's probably not going to gain Exit Velocity or better his ability to square the ball up as he ages, or completely retool his swing. Add on that on average, players decline after they enter the league, and 2015 may be the best we see from Wong.

If the Cardinals can find a team that believes in Kolten, and is willing to part with some real win-now value for him, I'd love to see them pull the trigger. Holy cow, we're over 2500 words at this point, which even for me is long. Thanks to everyone who saw this through to the end, and I hope you enjoyed it.