Kizer or Hogan: Who Should the Browns Start Next Sunday?

We have our first serious quarterback controversy in Cleveland after head coach Hue Jackson benched rookie starter DeShone Kizer and turned the reins over to second-year backup Kevin Hogan.

A day later, Jackson passed when asked whom he would choose to start at quarterback this coming Sunday, saying that he needed to go watch the tape before making that decision.

Clearly Kizer hasn’t looked great, but it’s only been five weeks. There are lots of questions that need to be asked and answered before Jackson makes his decision: Is it too early to move on? How bad has Kizer’s start been in comparison to other similar prospects? And does Hogan have a realistic shot being a long-term answer or is turning to him a waste of time?

The 2017 numbers

With a re-vamped offense line that was ranked No. 2 by Pro Football Focus in the offseason, there was a lot of hope that the Browns would have a functional offense this year, despite the relatively low draft position and inexperience of their 21-year-old rookie starter. The traditional stats show you that the passing offense has been functional, but only with Hogan under center.

PlayerComp%TD%Int%Sack%Yds/AttANY/A
K.Hogan68.47.95.35.09.98.5
D.Kizer50.91.95.77.05.42.6

The numbers are night and day. The most comprehensive passing efficiency stat – adjusted net yards per attempt (ANY/A) – places Hogan slightly below league-leader Alex Smith. While Kizer’s 2.6 ANY/A is more than a yard less than the second worst mark in the league.1

Kizer has been effective rushing with 126 yards and two touchdowns on the ground. But even if we look at a stat like points added that incorporates rushing value added, Kizer still pales in comparison to the value Hogan has generated.

PlayerTeamPts AddedPlaysPA/PPass PAInt PAPI PASack PARun PA
K.HoganCLE5.9470.137.4-3.41.4-0.61.1
D.KizerCLE-15.7203-0.081.8-14.92.9-7.31.8

It a small, 47 play2 sample, Hogan has averaged slightly more points added per play than the points-added leader through five weeks, Deshaun Watson (0.13 to 0.12). Kizer looks relatively better by points added than ANY/A, with less negative per-play numbers than Mike GlennonJay Culter and Joe Flacco.

We shouldn’t look at these numbers are declare Hogan the best quarterback in the league and Kizer the worst, but when we’re talking about a decision linked to who will give your team the best chance to win, Hogan has the advantage.

Benching Kizer at this point could threaten his development. But you also have to consider the development of your other young offensive players. In particular, first round pick – a higher selection than Kizer – David Njoku has thrived with Hogan under center, in contrast to being a relatively inefficient receiver with Kizer.

PasserYds/TargetTarget ShareRec Yds Share
K.Hogan14.20.120.17
D.Kizer3.90.070.06

Using history to estimate likelihood of success

While Hogan has been the superior player so far in 2017, we shouldn’t ignore the fact that Hogan was drafted much later than Kizer. In Bayesian statistics you can make effective use of small samples by combining limited evidence with a strong prior. In other words, we would be less confident that poor early results will continue for top picks, who we expect will eventually be good, than for late-round selections who haven’t had much historical success.

The first step to figuring out the likelihood that Kizer will continue to struggle or Hogan continue to thrive is to look at at historical hit rates for players drafted at their respective positions.

Football outsiders studied quarterbacks drafted from 1994-2016 and classified as “good” those who most would consider franchise quarterbacks. For quarterbacks who were selected in the second round like Kizer, the success rate was roughly 20%.3 This gives Kizer an decent evidence-free chance of success.

But when you dig further into the successes, you see that most of them were taken in the early second round. I think this is a significant difference from Kizer because an early second round selection has only had the worst teams in the NFL pass on them once, whereas Kizer saw multiple teams without a franchise quarterback pass on him twice. Therefore, it’s also relevant to look at third round quarterbacks, whose success rate is around half of second rounders at 10%.4

This means that our expectation for Kizer should be somewhere between 10-20%. But what about the fifth rounder Hogan? All quarterbacks selected in the fifth round and had a success rate of 7.8%.5 That’s not an encouraging rate, but not as dismal in comparison to second and third round quarterbacks are you might expect. Hogan’s prior is worse than Kizer’s based on draft position, but the Browns shouldn’t ignore the on-field evidence and give Kizer the nod no matter how the two are playing.

Visualizing Kizer versus historical successes

I’ve shown before my estimate for true passing efficiency using Bayesian updating. It incorporates our prior of a league-average quarterback and then updates what we believe a quarterback’s true passing efficiency is based on the evidence. The more evidence we have, the more heavily we weight that evidence versus our prior.

I thought it would be interesting to show how Kizer visualizes versus other successful second and third round quarterback. I only have data going back to 2000, but we still get a robust group including Drew BreesRussell WilsonColin Kaepernick and Derek Carr.

 

The colors might be a bit confusing, but Kizer is the line Browns’ orange line heading almost straight down. Past successes like Brees and Carr saw their true passing efficiency estimate fall as they struggled early in their careers, but not nearly as rapidly as Kizer’s. It’s possible that Kizer is able to turn things around after an abysmal , but it wouldn’t follow the pattern of other second and third round successes.

Hogan wasn’t as bad of a prospect as most think

Most don’t take Hogan seriously based on his draft evaluation from last year and the fact that he was a fifth round pick. But he checked a lot of the non-draft-position boxes you like to see in prospects who outperform expectations.

I found fairly simple set of criteria to find those outperformers: at least three years starting, a career adjusted yards per attempt (AY/A) of 8.0 or higher, and more than 20 rushing yards per game. These filters leave you with prospects who won the starting job early, have larger sample sizes, have shown consistent passing efficiency, and have the ability to add value on the ground, which is an undervalued trait.

Hogan joins the group below hitting all those marks: four years as a starter, career 8.7 AY/A  and 24.5 rushing yards per game.

Not all of these guys are been successes, but any screen that identifies Russell WilsonDak Prescott and Colin Kaepernick6 is worth monitoring. Hogan also checked the box of accuracy with a career completion percentage of 66%.

Hogan might not have the traits that scouts love, but he certainly had the numbers to forecast success, nearly matching the career numbers of the Stanford quarterback who preceded him, Andrew Luck.

Time to give Hogan a shot

Beyond the fact that Hogan performed at a different level this season, there are also reasons to think that Hogan’s chance of success moving forward isn’t that different from Kizer’s. It’s the right choice to give Hogan a shot to prove himself. I’m not going to get into the need for Jackson to preserve his job with some wins, which clearly tilts the decision in Hogan’s favor.

If Hogan fails, it’s not as if the Browns have given up on a top selection in Kizer. The Browns were fully prepared to miss all four of the top quarterback selections in last years draft after taking three non-quarterbacks in the first round. When Kizer fell to them in the mid-second round, they jumped to select him, but he clearly wasn’t a priority. The front office had a plan during the draft to let the quarterback position come to them, and that plan included a strong possibility they’d be going into the season with only Hogan, Cody KesslerBrock Osweiler and a mid/late-round pick at quarterback.

For a team that isn’t in a rush to tie themselves and mortgage the future on a signal-caller, the risk of waiting too long to a young starter to turn things around could outweigh missing the opportunity to what else you have on the roster. In the same regard, if Hogan struggles, there’s no reason to hesitate in going back to Kizer and seeing if some time on the bench studying will improve his outlook.

  1. Belonging to none other than Jay Cutler.
  2. Includes all passes, runs, sacks, defensive pass interference calls and runs
  3. Six of 25
  4. Three of 29. Sorry, I’m not counting Nick Foles as a legitimate success
  5. 10 out of 128
  6. Heck, finding Ryan Fitzpartick in at pick 250 ain’t bad.

Leave a Reply