Howie Kendrick, Dream Killer
The story starts with promise. Howie Kendrick, a 15-year veteran with a swing-first-and-ask-questions-later game, was doing something weird. Take a look at an extremely specific statistic, current as of August 9 — first-pitch balls in play, by year:
| Year | First Pitch BIP |
|---|---|
| 2008 | 30 |
| 2009 | 33 |
| 2010 | 65 |
| 2011 | 45 |
| 2012 | 70 |
| 2013 | 55 |
| 2014 | 73 |
| 2015 | 65 |
| 2016 | 62 |
| 2017 | 29 |
| 2018 | 19 |
| 2019 | 34 |
| 2020 | 0 |
Of note, I’m only going back to 2008, because that’s the first year of pitch tracking data — Kendrick started in 2006, but those two missing years don’t really change the narrative here. That zero in 2020 doesn’t look all that suspicious — the Nats had only played 10 games — but it looks a little suspicious. It might not be holding a match, but there are burn marks on its fingers. Could Kendrick be changing something on the fly?
Let’s look at a second, more telling statistic. This one is swings at first pitches, again by year:
| Year | First Pitch Swings | Rate |
|---|---|---|
| 2008 | 80 | 22.3% |
| 2009 | 82 | 20.6% |
| 2010 | 127 | 19.4% |
| 2011 | 108 | 18.6% |
| 2012 | 160 | 27.0% |
| 2013 | 138 | 27.2% |
| 2014 | 157 | 23.6% |
| 2015 | 144 | 29.1% |
| 2016 | 161 | 29.8% |
| 2017 | 70 | 21.0% |
| 2018 | 46 | 28.9% |
| 2019 | 80 | 21.7% |
| 2020 | 0 | 0.0% |
Curiouser and curiouser. Kendrick’s game is built around his sweet swing. He puts a ton of balls in play and runs a high BABIP, the foundation of a solid batting line: few strikeouts, few walks, and plenty of line drives sprayed to all fields. It’s hardly a surprise that he swung at first pitches — at a roughly league average rate, if you’re curious, as the average has bounced around between 25% and 30% over the past 10 years.
Why swing at the first pitch? It’s a way out of the strikeout squeeze of recent years. Pitchers are so, so good at striking batters out these days. Every random fifth starter has a wipeout secondary pitch. Good pitchers often have two or three, and they pair them with mid-90s fastballs in unhittable locations. Hitting with two strikes is inhospitable.
First pitches notably aren’t two-strike counts. They’re also fastball counts, and Kendrick is an excellent fastball hitter. Seems pretty straightforward, then: swing early if you get something to hit, particularly if you have some pop on contact, which Kendrick absolutely does.
That’s a great theory. There’s just one problem with it: it hasn’t played out that way in practice. It’s relatively easy to work out how valuable a given batter’s first pitch approach is. A ball moves the count from 0-0 to 1-0; every pitch taken for a ball converts a random plate appearance into a random plate appearance that starts with the batter ahead 1-0. The opposite is true for a taken strike, whiff, or foul ball. Balls in play? You can simply take the difference between the result of that play and the result of an average plate appearance.
It’s all computationally straightforward. In Kendrick’s case, it also wasn’t really working:
| Year | Runs Added |
|---|---|
| 2008 | -2.65 |
| 2009 | 1.69 |
| 2010 | -7.58 |
| 2011 | 1.79 |
| 2012 | -5.46 |
| 2013 | 5.78 |
| 2014 | -1.53 |
| 2015 | -0.19 |
| 2016 | -2.06 |
| 2017 | -1.04 |
| 2018 | -2.77 |
| 2019 | -0.92 |
This way of looking at first-pitch behavior is prone to some wild swings. In 2013, for example, Kendrick had a raft-load of extra-base hits on 0-0. In 2012, those balls got caught. In aggregate, however, the plan didn’t pay off.
One benefit of this type of analysis is that we can easily produce a breakeven production rate, a single number that tells us the wOBA value Kendrick would need to produce when he puts the ball in play on 0-0 to break even given the balls and strikes we already know about.
Here are those values, as well as Kendrick’s wOBACON in each year, the wOBA value for all of his batted balls combined. The 0-0 wOBACON tends to be higher than the overall number, though not hugely so:
| Year | Breakeven | wOBACON |
|---|---|---|
| 2008 | .425 | .375 |
| 2009 | .419 | .391 |
| 2010 | .415 | .349 |
| 2011 | .433 | .415 |
| 2012 | .412 | .371 |
| 2013 | .405 | .390 |
| 2014 | .413 | .367 |
| 2015 | .378 | .368 |
| 2016 | .410 | .327 |
| 2017 | .447 | .426 |
| 2018 | .439 | .405 |
| 2019 | .473 | .438 |
Yeah — it’s tough to get ahead when you need such strong results on contact to break even. As an added check, Kendrick’s career wOBACON on first pitches is a solid but insufficient .387. In other words, despite being an excellent overall hitter, Kendrick has had a consistent first-pitch leak for his entire career. Whether it was swinging too much, swinging at the wrong pitches, or taking the wrong mental approach to the plate, something wasn’t working.
So in 2020, Kendrick tried something new. In the first game of the season, Gerrit Cole piped in a fastball to start Kendrick’s year, and he was taking all the way:
That pitch might be a strike. It’s certainly not a good pitch to swing at, though, and Kendrick got the result he wanted; 1-0, hitter’s count. Cole came back in the fourth inning with a curveball low, and Kendrick took it as well. Two 1-0 counts against one of the best pitchers in baseball. It’s not much, but it’s honest work, and it helped put him in a position to succeed. He didn’t actually succeed in those at-bats — he struck out once and lined out once — but baseball is a game of small edges, and here was Howie Kendrick, taking his way to a new edge.
What a wondrous tale! A batter renowned for his ability to impact baseballs improves himself simply by not swinging. It’s not quite so simple, of course; if you never swing at any first pitches, you won’t get ahead, because pitchers throw in the zone quite often — 51.8% of the time in 2019 and 2020 combined. You actually need more balls than strikes for not swinging to make sense, because a 0-1 count is worse for the hitter than a 1-0 count is good.
Okay, so maybe the plan wasn’t to simply take every time. Maybe it was an ambush setup; lure the pitchers into a false sense of security by taking a lot, then pounce. Regardless, there had to be something there. Twenty-nine first pitches, 29 takes. That had happened exactly once in his career before this year, over nearly 6,000 plate appearances.
That’s the setup. New behavior from an interesting player, at an interesting time — with the offseason to plan, had Kendrick revamped his first-pitch approach? Would he be hunting meatballs but otherwise staying away from the swing game, a la Mike Trout?
Last night, there was one last game’s worth of data to collect before writing this article — you’ll notice that all the statistics above are through games of August 9. First time up, Steven Matz was clearly a little off:
One down, three or four to go. Matz came back with a better pitch the next time, a changeup on the outside corner that Kendrick was never very likely to swing at:
Alright! 31 straight takes. Hooray for arbitrary endpoints — he had never taken 31 first pitches in a row. Two more, an early night, and it would be time for a juicy article. Howie Kendrick tried this one weird trick — the title writes itself.
Yeah, about that:
Well, there goes the article. That was an excellent pitch: it resembled a fastball to stop any batters from sitting on it, broke away, and still finished in the strike zone. Still, that’s the kind of pitch that you hope taking all the way neutralizes; it’s not a very good idea to throw a breaking ball there if you know the batter won’t swing, because plenty of breaking balls end up out of the strike zone despite pitchers’ best intentions. Stick with the plan, Howie!
Yeah, again, about that:
Another excellent pitch, another swing, and now the narrative is crumbling. Kendrick waited all year to swing at a first pitch, and then he broke his fast on two consecutive sliders that he couldn’t do anything with? Where’s the fun story in that?
Okay, seriously now. Three swings all year, and they’re on three pitches that he probably would have been better off taking. Swinging at three of 34 first pitches isn’t even remarkable, not even the lowest rate in baseball; Mitch Garver and his two swings at 41 first pitches take that honor.
At the end of the day, this is about seeing shadows. I wanted, so badly wanted, the remarkably weird string of taken first pitches to mean something. I dug back and found a hole in Kendrick’s game that a more patient approach might correct, at least theoretically. I found a sweet GIF of him taking a tough pitch from literally Gerrit Cole. Everything was all coming together.
Of course, that’s not how baseball works. Sometimes Howie Kendrick takes 31 straight pitches because he feels like taking, then swings at three in a row because he feels like swinging. Sometimes he takes because he feels like he’s in a funk, then swings because he’s still in a funk and wants to change things up. Sometimes a rare event happens by pure chance, with no underlying meaning at all.
I still want to think there’s something there. Kendrick’s overall swing aggression works for him, but it hasn’t worked on the first pitch; he’s given up 15 runs of value there over the years, a strong outlier for an overall solid hitter. He really should be doing something different to get ahead.
And maybe he really is consciously swinging less often. Maybe he planned on unleashing a first pitch swing as an ambush, only to get punked by Sewald’s excellent sliders. Maybe he’ll continue with this take-heavy approach and refine the times he chooses to swing. Maybe he just had too much coffee before the game yesterday.
That’s all wishful thinking, though. Most likely, I simply saw something that wasn’t there. Sometimes, a cigar is just a cigar. Zero is a fascinating number. One out of 30 is a lot less exciting than zero out of 30. That’s probably the real tale here. Baseball players do weird things all the time, through the magic of huge numbers of trials and random chance. Not all of it has meaning. That doesn’t make it uninteresting, but it does make theories based on 30 pitches a silly way to go about analysis.
