Shooting Stars and Falling Stars, Part 3

Today we wind up our look at success rates for college QBs compared to their recruiting rankings. Part 1 described the method of analysis and took the first step of looking at 5-star QB recruits from the 2017-2021 classes, from Trevor Lawrence to Sam Huard. Sadly, with much love in our hearts for him and his family as UW legends, the comparison did not go well for Sam.

Part 2 took it a step further by looking at a much bigger group of over 60 4-star QBs from the same classes, from Sam Ehlinger to Matthew Baldwin (who?), and right in the exact middle of that group was our own Dylan Morris. Literally in the middle, 31 4-star QBs above him and 31 below. They don’t get much average-er than that! (And it was certainly better than any of Garbers, Sirmon, or Yankoff could manage in those classes.)

This final chapter focuses in part on the third recent UW QB of note, our fearless leader and Heisman contender in Michael Penix, Jr. He was the middlest of mid-3* players with a 247sports.com rating of 85, the 66th-rated QB in his class. That’s lower than Jake Haener had in 2017 (86, 40th), and for that matter it’s lower than Nick Tronti (86, 44th), who was Indiana’s QB recruit in 2017. There was nothing in his recruiting profile that necessarily pegged him for greatness, and yet here we are.

For fun, out of the 129 QBs in this analysis, here’s how UW’s recruits ranked overall: Penix #16, Morris #53, Garbers #86, Sirmon #94, Yankoff #103, Huard #104. Hmm, maybe that’s not fun. Anyway...

There were 13 5-star QB recruits from 2017-2021 and 63 4-stars. Simply put, there are WAYYY too many 3-star QBs for me to have the time to deal with them all, so to keep things lined up we are going to focus on the nearest comps to Penix in terms of recruiting profile. To wit, I looked at all of the 3-star, 85-rated QBs from that period, and that group alone is nearly as many as the entire cohort of 4-stars, with N = 53. That feels like plenty to give us a representative sample in terms of sheer numbers, and also taking a slice dead center from the overall 3-star band feels like a reasonable proxy for “all 3-stars.” So, how does Penix compare to other QB recruits with the same profile as he has, and how do all of them stack up against the 5-stars and 4-stars of the QB world? Let’s take a look at Team 85!

How Do the 3-Star 85s Compare to Higher Star Ratings

It’s not good, Bob.

It should surprise no one that the typical 5-star outperforms the typical 3-star, but the degree of difference might be a little startling, especially if you step past the mean value that is artificially inflated by a very small cluster of huge statistical outliers at the top and a loooonnnnng tail of nothing at the bottom.

In comparing 4-stars and 5-stars, we found that exactly 30 of them (10 5-stars and 20 4-stars) had achieved a total BIG+ score (see Part 1 for details) of over 100. That represented 76.9% of 5-stars and 31.7% of 4-stars.

How many 3-stars reached that mark? Three. Out of FIFTY-THREE. That’s 5.7%. You already know one of them, but Penix at 170.45 is actually only second-best in this mark behind Clayton Tune of Houston at 186.49. Penix will almost certainly pass him by season’s end, but those two are way out in front. Memphis’ Seth Henigan just barely crosses the 100 mark at 104.02. Two other QBs had a raw score over 100 but their BIG+ was adjusted downwards for doing it against MAC competition at Western and Eastern Michigan.

A total of 9 3*-85 QBs were able to top the Morris Line - D-Mo’s 50.18 BIG+ score that is the median value for 4* QBs. Of course, that means 44 of them (83%) fell below it. That’s a little deceiving though, because only 4 3*-85 QBs were even more than halfway to the Morris Line, with 40 (75.5%) having a BIG+ score below 25.

The median value for 3*-85s is a lot more indicative of the typical performance for recruits at that level, and it’s not pretty stacked up against 5-stars (producing over 12 times the BIG+ score and, hilariously, 85 times the total offense yards) and 4-stars (over 4 times the BIG+ score and 60 times the total offense). That’s the case because the median value delivered by a 3*-85 QB recruit in this 5-year span was statistically very close to zero. In fact, let’s look at that number zero.

How Do the 3-Star 85s Compare to Each Other

Looking at the 76 QBs in our 5* and 4* sample, there was only one QB, Matthew Baldwin at Ohio State, who produced zero total yards. No yards from scrimmage. No passing yards, No nothing. In the 3*-85 group, there were 15 QBs who had zero or fewer yards (Zy McDonald of Louisiana-Lafayette at least got into a game but had the dubious distinction of losing 9 yards rushing). For that matter, there were two guys who as near as I could tell never actually showed up on campus anywhere.

There were another dozen QBs who amassed fewer than 200 total yards. Taken with the 14 zeros and one negative, that’s 27 3*-85 QBs who produced fewer than 200 yards of total offense in their careers. Similarly, there were 26 QBs who never produced a single touchdown. In both cases, that’s more than half of the entire group of QB recruits that produced essentially nothing in their careers. That explains the incredibly low median value for 3*-85s: The bust rate is extremely high.

I suppose you could quibble with calling it a “bust” rate, given that these players were not brought in expected to be stars, but they were brought in expecting them to be able to be developed and challenge for a starting job at some point in their careers. Coaches aren’t going to waste scholarships and roster spots on guys they think just can’t play. On the other hand, a 3*-85 is also not going to get the benefit of the doubt or public pressure to get put into games; they’ll have to work for everything they get. They may even need to transfer for another opportunity, and many did.

Transfers

Of 22 transfers, 4 were moving up to a more prominent position, either hoping for a starting shot like Penix (or even Collin Schlee going from Kent State to UCLA, in case they weren’t ready to start Dante Moore) or as a backup, 4 were lateral moves (in two cases followed by a second transfer down), and 14 were going down a level. Still, a lot of these guys were already starting out at G5 schools or even FCS schools, but fewer than you’d think. Overall, 28 of 53 (52.8%) started at P5 schools, and not just bottom-feeders like Rutgers, Colorado, or Indiana (looking at you, Mike Penix) but schools like Miami, Michigan, Utah, Tennessee, Virginia Tech, and Pitt.

Some guys did well after transferring. Penix, of course, is the poster child for that, but Ben Bryant had a solid run split between Cincinnati and Eastern Michigan, Donovan Smith at Texas Tech and Houston, and Carter Bradley at Toledo and South Alabama; all did better than 5-star Davis Mills at Stanford. Nothing earth-shattering, but solid, productive, decent careers. Most transfers, though, didn’t see much more success at their second (or third) stop than they did at their first, even when moving down a level.

So How Rare is Penix’s Performance?

Very.

Out of 53 QBs in the 3*-85 group:

• How many have more yards than Penix? One: Clayton Tune (at Houston when it was G5).
• How many have even half as many yards as Penix? Four, all at G5 schools.
• How many have even a quarter as many yards as Penix? Four more, 3 at G5, 1 at Baylor.

Tune had a great career at a slightly lower level in the AAC, but Penix will almost certainly pass him by season’s end, making him the outlier of all outliers among players with his recruiting ranking. He and Tune have almost double the score of the next-highest QB (Henigan at 104). Those three are the only 3*-85 QBs to top 100 in BIG+ score, putting them in the top 33 out of 129 QBs in this study. Overall, Tune ranks #11, Penix #16 (right in between Tua at 15 and CJ Stroud at 17 - pretty good company), and Henigan at #32. Penix won’t catch Bo Nix or Caleb Williams, and probably not Jayden Daniels either, but they’re the only active QBs in this group ahead of him. With a monster rest of the season, especially if he gets 2 extra games in the CFP (hey, dream big, right?), he might be able to get past Sam Howell up to #7.

What the numbers show is that the likelihood of getting a Penix-level performance out of a player with his talent rating coming out of high school is incredibly low, less than 5%. Possible? Sure, obviously. It has happened. But it’s like rolling a natural 20 in D&D. It can happen, but you can’t necessarily bank on it.

This analysis looks at QB production on the field; it does not look at specific situations to ask why each player did or didn’t succeed. The presumption is that all college coaches are generally competent at what they do, so the odd cases of actively destructive coaching (Exhibit A: John Donovan) are sufficiently uncommon that, with a large enough N, that kind of statistical noise (along with things like specific instances of poor system fit, injury, getting stuck behind another QB, and so on) gets washed out by the law of averages.

We can say that, across a 5-year period with 129 QBs and holding other variables to be functionally equal, the chance of getting a Penix-level performance or better (let’s set the bar at BIG+ of 150 or better) is about a 50/50 proposition for a 5-star QB recruit, about 1 in 5 for a 4-star QB, and about 1 in 25 for a 3-star (85) QB.

Is It Penix or the System?

Can we tip the scales in our favor? It’s easy to say 5-star QBs are the best: no duh. But when you look at how much better they are, you have to ask whether you can get similar performance from lower-ranked guys, especially if you aren’t planning to hand out literal millions to recruits to get 5-star QBs on campus.

Penix has done great. He might win the Heisman. He finished 8th last year and is on track for better this time around. Penix has a marvelous arm, but is he a unique magical unicorn as a QB? How did people miss on him so badly as talent evaluators? Or is he just such a terrific student of the game that he took ordinary material and forged an extraordinary career?

There is a glimmer of possibility, though, that maybe KDB/Grubb have a secret sauce that makes any QB a star. They certainly had great success at Sioux Falls. Penix himself was a different player with KDB than without him. His overall numbers at Indiana were solid but not spectacular. He had one terrific half-year in 2019 with KDB, but otherwise he was just okay. His 4-year stint at Indiana, including that dynamic 2019, would have been 68.93, not exactly a big margin compared to Dylan Morris’ 50.18. Right about that Davis Mills level. But at UW he’s added over 100 BIG+ points in 17 games with the rest of the season yet to come. Something is definitely working.

Then, of course, there’s the case of Jake Haener, who we mentioned above as being marginally ahead of Penix in terms of ranking, but only just (rating 86 vs. 85) and who had lots of moxie at UW but little to show for it, who then went to Fresno State and under KDB/Grubb for his first two years there blew up and became a star. His BIG+ score: 127.73, and that’s even counting the penalty for playing in the MWC.

Kalen DeBoer has proven that his system, in at least two cases, has been able to take guys with middling-at-best talent ratings coming out of high school and turn them into all-stars. I don’t know anything about his QBs at Sioux Falls, or at Eastern Michigan for that matter. I know about these two: middling recruits who turned into stars.

That begs the question: Are Haener and Penix unique and special, and those are the only two guys this system could’ve worked with? Or can KDB’s system make any QB look great? That’s not to say that Haener and Penix didn’t have exceptional characteristics that got noticed and refined by KDB’s staff, but on a certain level you can say that about anyone. Everyone is special in some way. Can the KDB system find the specialness in any QB and make them special?

Is the Haener/Penix Performance of 3* Rating/5* Performance Repeatable?

We don’t know, and we won’t find out until the next QB takes the reins. For that matter, we might not need to. Demaricus Davis is a 3-star QB. Will he be the next KDB 3-star success story down the road? Or will Dylan Morris pay off his QB rehab tour with a magical senior season on Montlake? Will Austin Mack flash his mid-4* bona fides under the KDB system? Or will the next Husky QB be another transfer in? Or a parade of 4* and 5* players who see the absolutely sick numbers MPJ is throwing up and decide they want a piece of that?

The system doesn’t work for everyone. It never quite took with Sam Huard. We don’t know why. But so far it seems to be working great with whomever ends up starting.

It stands to reason that better starting talent definitely gives you a better chance to achieve greatness. I’m not going to beat Usain Bolt in a footrace no matter how much effort and how good my coaching and training is. I’m not going to beat Greg Lewis either, and we’re the same age. But, to mix metaphors, I can make a much tastier steak with a nice ribeye than I can with top round.

It’s clear KDB knows what he’s doing with his passing game. It’s worked great to compensate for (supposedly) lower talent, but how much more effective might it be as a force multiplier for (supposedly) higher talent? I hope we get to see it. I definitely trust their evaluation process and scouting and coaching acumen. I’m just also aware that no one person is smarter than everyone else put together, and trying to outsmart the odds is an uphill battle, and the difference in likely outcomes is pretty stark comparing 5* to 4* to 3* QBs. It’s not impossible to beat the odds, but it’s a lot harder and requires precision and luck to try to make sure you don’t miss. For every Michael Penix Jr. there are two dozen Micah Bowens. If KDB has the Midas touch and make turn any QB to gold, awesome, but I hope UW does the best it can to put the odds in their favor at the most important position on the field.

Thanks for reading all of this and GO HUSKIES!