/cdn.vox-cdn.com/uploads/chorus_image/image/62872912/827424778.jpg.0.jpg)
As we put another week in the books of a seemingly endless and unmoving offseason, it can be hard to remember that baseball is just around the corner. Pitchers and catchers will be reporting before you know it, and you’ll be able to avert your eyes from the slow-motion trainwreck of free agency. To that end, I spent a good deal of time this week updating some descriptive statistics I looked at throughout the 2018 season to see how various players finished out the year. Partially I just wanted to come up with some new ideas for what to write about in 2019, but I was also genuinely curious to see what became of statistical oddities I picked up on during the season. Just to give one throwaway example, I wrote about Jordan Hicks when he was on an absolutely insane run of striking out a third of the batters he faced and not allowing any walks. He was still good on the season as a whole, obviously, but never quite so much as in the two-week period surrounding when I wrote about him. If you’ve ever heard of the Sports Illustrated cover jinx, you can totally see how this happens. For someone or something to be worth writing about, it has to be notable or weird. Regression to the mean exists, though, and so inevitably we can’t have nice things. No one writes about Matt Carpenter after he has a week where he hits just okay- it’s always after he’s been great or terrible.
I’m happy to report that I found one exception to that, however, and that’s going to form the first of my two vignettes about place discipline today. Back in July, I developed a new statistic for plate discipline and found that Kolten Wong, of all Cardinals, performed best on my metric. It wasn’t just getting intentional walks from batting eighth, either, as the statistic strips out intentional walks. If I’m being honest, I basically dismissed it in advance at the time. No need to wait for someone to regress to the mean if you don’t even believe their performance is real to begin with. Well, the joke’s on me, because Kolten Wong actually got *better* at the plate in the second half of the season. Before I go any further, let’s look at a list of how everyone on the team performed in 2018. As a quick reminder, NOC+ is indexed to 100, so a 120 NOC+ would be someone who could be 20% below average on contact and still be an average hitter:
2018 Cardinals, NOC+
Name | NOC+ | K% | BB% | HBP% |
---|---|---|---|---|
Name | NOC+ | K% | BB% | HBP% |
Kolten Wong | 112.12 | 14.90% | 6.90% | 3.50% |
Yadier Molina | 110.27 | 13.10% | 5.80% | 1.80% |
Greg Garcia | 109.22 | 17.90% | 9.20% | 1.90% |
Jedd Gyorko | 108.22 | 19.20% | 10.90% | 0.70% |
Jose Martinez | 106.08 | 17.60% | 8.30% | 0.30% |
Dexter Fowler | 104.46 | 22.50% | 11.40% | 0.90% |
Matt Carpenter | 104.43 | 23.90% | 12.90% | 0.90% |
Marcell Ozuna | 103.17 | 17.60% | 5.80% | 0.50% |
Yairo Munoz | 99.86 | 22.00% | 7.10% | 1.20% |
Tommy Pham | 99.3 | 24.60% | 10.40% | 0.50% |
Paul DeJong | 96.59 | 25.20% | 7.00% | 2.50% |
Harrison Bader | 89.08 | 29.50% | 6.60% | 2.60% |
Francisco Pena | 79.02 | 30.90% | 2.20% | 0.70% |
Tyler O'Neill | 63.46 | 40.10% | 4.90% | 2.10% |
First things first- good job Kolten Wong. Being a target for baseballs propels him to the top of the list, but even purely his walk and strikeout numbers are quite solid. It’s also worth noting that Tyler O’Neill’s plate discipline numbers should terrify you. There’s absolutely a good player lurking in there, and yesterday’s ZiPS projections have him at a more manageable 32%/8% split, but he was more or less impossibly bad in 2018, and he simply won’t be able to repeat those numbers in 2019 and be a valuable contributor. Back to Wong, though: if you’re looking for a way to create a stable value floor, this is exactly how to do it. Wong ran only a .275 BABIP last year. He hit for no power- a .139 ISO, well below league average. Despite all of that, he was nearly league average from the plate, all because he essentially willed himself to positive outcomes when he didn’t put the ball in play. Combine that with elite defense, and you’re looking at a tremendously valuable player, power or no.
Initially, I was going to try to explain why Wong is so good at this essential but hard-to-evaluate skill. As it turns out, though, it’s incredibly hard to evaluate! Wong’s central skill appears to be doing everything pretty well. He doesn’t swing at balls very often. He’s patient in the zone, swinging less than league average. He makes a lot of contact. He gets hit by a lot of pitches. There’s no one skill, but they all add up together nicely. In the process of trying to figure out what made Wong so good at this skill, however, I looked back to 2017. Guess what? Wong was great at it then too! Again, lest you think it’s all unintentional walks, those are stripped out. Even without them, Wong ran a 7.5% walk rate and got hit by a pitch another 3% of the time. He’s simply always solid. That was an absolutely startling find for me, because I simply had no idea Wong was doing anything special. Lest you think it’s an effect of batting eighth and getting a lot of intentional unintentional walks in front of the pitcher, that effect certainly isn’t the case league-wide. This isn’t a great way of doing the study, and worth coming back to later, but as a first pass I looked at the NOC+ of all 7th batters and all 8th batters in the NL in 2018. Not only did the 8th batters not receive a bump, they had a lower unintentional walk rate than 7th-place hitters. Clearly there’s a bit of bias because I’m not using the same population of hitters, but any kind of eighth-spot bump is hard to find.
This quick nod to Wong’s 2017 stats leads me to my second quick hit on NOC+. I mentioned when I created the stat that I wanted to see how it held up year-over-year. To do this, I generated samples of every batter with at least 300 PA in both 2017 and 2018 and checked for year-over-year correlation. I won’t beat around the bush here- it’s super high. 60.2% of the variance in 2018 stats can be explained by the 2017 stats, which is pretty incredible. As we say in the statistics business, that’s a high R-squared. For comparison, wRC+, everyone’s favorite all-in hitting stat, carries only a 28% R^2 from year one to year two. NOC+ also does a decent job of predicting next year’s OBP- 2017 NOC+, for example, has a 28% R^2 with 2018 OBP. It doesn’t do a very good job of predicting overall batting lines, but that’s pretty reasonable, as it says absolutely nothing about what happens after you get your bat on the ball. Any stat where Andrelton Simmons and Kris Bryant grade out similarly is going to miss on the specifics, but I’m alright with that for now. Building a tool that predicts next-year OBP better than current-year OBP feels like a good start.
What’s the next step for my toy stat that has so far exceeded my early expectations for stability and predictive power? Well, at some point it’s going to be time to add in park adjustments, but that sounds like a pretty daunting project at 9pm on Friday. Instead, let me leave you with this thought on Kolten Wong. Wong’s an imperfect player, and that’s probably not changing any time soon. It doesn’t mean that he’s not incredibly valuable, though. It’s easy to see it in the defense, but Wong’s offense is the reverse of his defense- subtle, not flashy, but consistent. I’m as guilty as the next fan of wishing Wong could just do something, anything, at the plate to complement his defense. It looks like I’ve been missing the forest for the trees, though. I wish I had a better conclusion for you, but that’s basically all I’ve got. Kolten Wong’s offense isn’t great, but it’s SO much more stable than I’ve given him credit for.