Andrew Knapp is Inexplicably Proving Me Wrong

“Finally, one spot would be needed for a backup catcher; this will be given to Andrew Knapp, but the much younger Deivy Grullón has a better bat and should earn that backup job by season’s end.”

Nominally a Phillies fan when the 2020 season, with its new roster and in-game management rules, were confirmed, I took the opportunity to write about how they should construct their team for the year. The quote I used to open today’s article is one directly from my previous breakdown of the Phillies, titled “Weaponizing the Phillies Roster“. Well, now with under a week to go in the 2020 regular season, Grullón is at the Red Sox alternate training site and the Phillies top team catcher leaderboards in fWAR and wRC+ with their combination of Knapp and J. T. Realmuto. I’m left to wonder where I went wrong. Luckily, I’m more than willing to dig into the numbers for an investigation.

Before getting into this year, I would like to defend why prior to the season I made the soon to appear absurd claim that Knapp wasn’t even worthy of a backup role. In parts of three MLB seasons from 2017-19, Knapp had not been a good player. Over 579 PA in that span, he slashed .223/.327/.336 with a 31.4 K% for a 78 wRC+. Added in was the fact that FanGraphs valued his framing at -8.2 runs for his career, and even the catcher positional value couldn’t save him from being replacement level; his career fWAR was 0.1. There were signs of hope, like the fact that in 2019 he had been a positive framer and had underperformed his xwOBA by 22 points, but these were minuscule and, in my mind, more than drowned out by his track record of MLB futility.

Through 74 PAs in 2020, Knapp has made me look very foolish. His slash line is a much improved .310/.446/.466, and his 151 wRC+ ranks fourth among all catchers with at least 50 PAs. Despite his limited playing time as the backup to a superstar (Realmuto) and dropping back to negative as a framer, the 0.7 fWAR Knapp has racked up is both easily a career-high, and good enough to place him 14th among MLB catchers. Clearly something is different about Knapp’s offensive game in 2020. A quick scan of his FanGraphs page confirms this. So far in 2020, he has a .395 BABIP, 65 points above his career mark heading into the season, and clearly an unsustainable pace. This makes sense—74 PA is a small sample, making an outlier BABIP a distinct possibility, and a suddenly inflated BABIP would allow a player to have a slash line well higher than he had ever posted before.

So to review: a small sample size has let Knapp get unsustainably lucky on his batted balls, artificially boosting his profile as a hitter. That’s it, I can close my investigation and go celebrate a job well done by working my way through a crossword puzzle right?

Wrong. BABIP may be fluky during a small sample, but there are some stats that stabilize awfully quickly. In these, it is worth looking at a stretch of just 74 PAs, as they are indicative of a player’s true talent level in that area. Look at Knapp’s 2020 in those fast-stabilizing stats (entering play Tuesday, 22 September) compared to his career rates from the previous three years.

I … huh? Knapp has gotten significantly better in 2020 by taking more pitches and apparently hitting the ball softer. One of these, the former, makes sense, but the latter sure doesn’t.

I’ll address the one that jives with everything I know about baseball first. Before 2020, Knapp struck out way above league average and walked slightly above. As evidenced by the table, he now walks a lot and strikes out just under average. Curiously, his swing profile hasn’t changed all that much, with his overall Swing% staying almost exactly the same. The two things that have changed are large drops in his Meatball Swing% and in his Edge%. Knapp is far better than league average in regards to not chasing pitches. Knowing this, pitchers can’t nibble against him for fear of missing the zone badly, bringing his Edge% way down. Pitches to Knapp are usually distinguishable as either a strike or a ball quite easily. This enables him to take meatballs for strikes and work deeper counts because since he’s unlikely to chase pitches he can afford to fall behind. Knapp’s decline in seeing pitches on the edges is the culprit for his drastic changes in walk and strikeout rates. However, that doesn’t explain all his offensive changes.

In 2020 Knapp’s average EV has fallen drastically, ranking 326th among the 425 hitters with at least 25 batted ball events. He’s not even putting the ball in the air more often, as that career LA going into this year was mainly due to a 4.6° average in 2017; he was at 16.1° in 2019. Yet his xBA, xSLG, and xwOBA have all increased demonstrably. Looking at Knapp’s batted ball data, there actually is a way to make sense of this development. Knapp is pulling the ball significantly more often in 2020, a common adjustment struggling hitters make, but not the key one for Knapp. The percentage of batted balls from Knapp that Statcast classifies as a Flare/Burner is 43.2, way higher than both his career rate and the league average. Balls hit in this category extremely often fall in front of outfielders for bloop singles, or, rarely, down the line for shallowly hit doubles. Pulling balls like this is what is letting Knapp produce so many hits and expected hits, despite his meager exit velocity.

Taking meatballs and hitting the ball softly is quite the unconventional and counter-intuitive way to break out, but Andrew Knapp is doing just that, forcing me to eat my previously written words about him.