How Not to Scout a Quarterback, Part 2

Last week, I tried to explain why college stat lines are a really bad way to try to evaluate a college prospect when it comes to the quarterback position. However, if you don’t know what you’re looking for or you don’t have access to EJ Snyder’s wonderful breakdowns (here or here), you might want to lean a bit more on stats. For that matter, you might have done your film watching and now you’re curious about whether or not what you see shows up in a measurable way. What can we say?

Let’s look at the college stat-lines from nine quarterbacks receiving a lot of attention.

This is the pool that I will be comparing to the 18 “good” quarterbacks and 17 “bad” quarterbacks that I identified last week, as defined by their performance once they reached the NFL.

First, there is a lot of worry regarding Mitch Trubisky’s limited playing time. Some worry that it speaks to his ability that he sat behind Marquise Wilson. [edit: Williams]. For those of you who feel this way, I remind you that Jordan Howard sat behind Jeremy Langford. Stuff happens and sometimes better players sit behind less capable counterparts. Coaches make questionable decisions at all levels of the game. By itself, Trubisky’s delayed ascension to a starting role doesn’t mean a lot.

A more valid concern may be that Trubisky simply hasn’t played enough to be evaluated effectively. For this, the total number of games doesn’t really matter. After all, Trubisky technically played in 31 games, per Sports Reference-CFB. What really matters is how many times he’s had to throw the ball in a game—the total number of passes attempted. This is a significant limitation, as he has only attempted 572 passes, and 447 of those came in the most recent year. How low of a number is that? It’s not a lot, but it is roughly twice as many passes as Cam Newton had attempted in college when he was drafted #1 overall, while it’s 123 fewer total passes than DeShone Kizer. Put another way, Trubisky is closer to DeShone Kizer’s total (695) than Deshaun Watson (1207) is to Patrick Mahomes (1349).

Of the “bad” quarterbacks listed in my last article, only one of them had fewer college passes than Trubisky (Mark Sanchez), whereas two “good” quarterbacks were in the same boat (both Newton and Matt Flynn). In other words, there is no statistical warning bell here, so long as what’s showing up on tape looks ready for the pros.

Second, there are accuracy concerns with some of the top candidates. As noted elsewhere, watching the games themselves will explain a lot about the reality behind the numbers, but the worst completion percentages in this group belong to Nathan Peterman (60.1%) and Brad Kaaya (60.6%). Bad, but those totals are both better than the college completion percentages of 5 of the 18 “good” quarterbacks drafted in the last ten years. Those totals are better than the completion percentages of Matt Ryan (59.9%) and Matthew Stafford (57.1%), but worse than Jimmy Clausen’s 62.6%. What’s interesting about this is that both Ryan and Stafford went on to be leaders in completion percentage in the NFL, as did fellow sub-60%er Tyrod Taylor.

Next, there’s adjusted yards per attempt. This figure takes into consideration not just whether or not the player can complete a pass for some real depth, but also the ability to earn touchdowns while limiting interceptions. Football Outsiders uses it as a major component in their own QBASE, their formula for projecting college quarterbacks into the NFL. Trubisky leads the way (with 9 AY/A), while Dobbs (6.9) and Webb trail (7.3 AY/A). If I had to build a case against one of these quarterbacks purely on the basis of stats, it would be here. Nine of the worst quarterbacks in my pool of 60 recent draftees had sub-8.0 AY/A, whereas only five of the “best” group did. And two of those five have some legitimate question marks attached (Matt Flynn and Nick Foles).

Add into these concerns Dobbs’ inability to double up on the touchdowns to interceptions playing against college defenders and he’s relatively safe to eliminate from the pool, at least on a statistic level. Likewise, when looking at the low passer ratings turned in by Webb and Dobbs, and I feel safe trying to recycle another team’s discard instead of picking up one of these two.

Okay, what about the other way? Does anyone jump off the stat sheet in a positive light? Yes. Without question.

Trubisky has the highest completion percentage, the highest adjusted yards per attempt, the highest passer rating, and the highest TD:Int ratio. Watson, with the second-best completion percentage (only 0.1% behind Trubisky), third-best AY/A, third-best TD:Int ratio, and second-best passer rating seems like a fine choice. Mahomes is basically the reverse of Watson (4^th-best completion percentage, second-best AY/A, third-best passer rating, and second-best TD:Int balance). Any of these three show stat lines that would be hard to separate from the “good” group of recent draftees.

The next tier would be formed (statistically) from Kelly, Kaaya, and Kizer. They have promising AY/A, passer ratings at least a little above the 145 averaged by the “bad” group of recent draftees, and decent completion percentages. Of these, Kaaya would be the front-runner because he has more total pass attempts and the best TD:Int ratio.

However, here’s an interesting thought. Cody Kessler had a 67.5% completion percentage over nearly 1300 attempts. He managed 8.9 AY/A and a 156.4 college passer rating. He managed 4.63 touchdowns for every interception. On paper, he was basically as good as a Trubisky or Watson. He was the 93^rd pick in the draft.

If Pace were to go after one of the top three, I would have trouble arguing with the choice using numbers alone. After Myles Garrett, it’s hard to say that there is a transformational player in this draft.

However, if you made me pick one of these quarterbacks without anything to go on but the stat sheet, I would wait until the second round and pick the best remaining from Trubisky, Watson, Mahomes, Kaaya, or Kizer. The stats and the historical comparisons make them the five best bets (I would skip over Kelly for the same reason I would have skipped over Manziel). This means that I would basically be giving up on getting the best of this crew by missing out on Trubisky and Watson for sure, and maybe even missing out on two of the other three. I would instead be hoping that a better team would give any one of these quarterbacks a better chance to develop.

Let’s hope Pace knows more than I do.

All stats come from either Pro Football Reference of Sports Reference-CFB.