A quick review after a couple days since the last edition:

• Part One (ACC)
• Part Two (Big 12)
• Part Three (Big East)
• Part Four (Big Ten)
• Part Five (SEC and Pac Ten)

The working hypothesis, rather than the conventional wisdom of steady improvement, is that most quarterbacks are who they are from the beginning: initially good passers tend to decline or level off after a strong start, mediocre players tend to stay mediocre, and bad debuts are followed with some improvement, but generally level off. The exceptions are the really physically gifted, probably draft-bound players, like Philip Rivers, Vince Young and Chad Henne, who might (or, in Henne's case, might not) look bad or mediocre to begin but evolve quickly into much stronger players.

With this sample size, I don’t think that prediction held up under much scrutiny. It seems pretty clear most teams are starting a freshman quarterback out of necessity, not because he’s destined for stardom, and since even the most hyped freshmen sit a majority of the time, young ‘uns who do play immediately don’t have much greater eventual success than guys who don’t start until later in their careers. But as far as an identifiable trend goes, there’s this among the 41 guys with qualifying numbers as freshmen and sophomores:

The path of both lines is based on freshman passer rating, so follow the invisible vertical lines from year one (red) to the fluctuation in year two (blue). The two lines don’t show much relationship to each other, sequentially –– you could bet that a kid who started out with a rating below 100 would not explode as a sophomore, or that a guy who was over 150 as a freshman would not completely collapse, but for the great majority of quarterbacks in the middle, better numbers as a freshman did not correlate at all to better numbers as a sophomore.

In fact, the most obvious trend in year two is a movement toward the mean, in both directions. Bad and mediocre passers improved, sometimes greatly, but guys who were hot out of the gate almost universally regressed. And it just so happens that the point of departure for diminishing returns is right at last year’s national average for pass efficiency: 136.6. Of the 31 guys performing below that mark as freshmen, 26 improved as sophomores, most of them substantially, although the majority remained below average. Of the ten players who outperformed the average as freshmen, eight regressed in their second year, six of them falling below the average. To paraphrase Big Ten Wonk, the lesson here would be: whatever you think of him, for predictive purposes, regress your view of your young quarterback to the mean.

But as far as who’s progressing and who’s regressing, there are no great insights. Among the guys who improved their efficiency by 15 points or more from year one to year two:

Continue reading this post »

One of the regular assumptions here and elsewhere holds that, whatever a player's performance as a freshman, automatically add major gains to his sophomore projections. He had no idea what he was doing; now he does. Hence: he will be better. He'll probably be awesome, in fact, since who's good enough to start despite having no idea what he's doing? Someone awesome.

Since it's June, and we're just chugging along here, I'm going to spend the next couple weeks testing that assumption – not challenging it, necessarily, but testing it – as it relates to quarterbacks. On what basis, besides their preseason assurances that they're "light-years ahead of this time last year," can we assume young passers are going to improve?

In looking at the numbers in the ACC, it was clear that improving a freshman season was based mainly on being terrible in that season - the only respectable freshman quarterback since 2000 who improved as a sophomore was Philip Rivers, who happened to go on to the best college career in the group by a mile and get snatched up in the first round; none of the other freshman starters were drafted.

The assumption from that is that eventually good quarterbacks may start slow but improve quickly, mediocre quarterbacks may be initially good but then take a step back or level off, and bad quarterbacks start slow, then take a small step forward before leveling off for the rest of their careers, if they can stay on the field. The data from the Big 12 backs up that conclusion to an extent (remember, in the cases of Young and Smith, especially, this doesn't account for extremely valuable running ability, which is more or less there from the outset):

Sometimes you can tell the future, sometimes you can't. Three guys here had very high, NFL-type projections out of high school: Young, Bomar and Freeman. In those cases, you'd probably predict a steady improvement from Year One to Year Two, and in Freeman's case, that's exactly what we see: a terrible freshman quarterback (outside of leading KSU to a high-scoring upset over Texas) who was vastly better as a sophomore – maybe even a little above average, by the standards of a less pass-happy conference. We see a big leap in Young, too – but as a junior, in his world-beating 2005 season, not as a sophomore. In Year Two, Young won the starting job for good but didn't get any better as a passer. Bomar, we don't know; we just have to assume the top-rated quarterback of the incoming class in 2004 would have been markedly improved after taking his lumps as a redshirt freshman.

Damn you, Bomar, will your unrequited talent never sleep?!
- - -

In the cases of Brad Smith and Bret Meyer (who replaced the woefully erratic Austin Flynn), what you saw at the beginning was pretty much what you got the rest of their careers: Smith, always more dangerous as a runner, never had a passer rating in his last three years as high as his mediocre freshman mark, and Meyer, after a WTF, moonshot season as a sophomore, fell back below his freshman numbers his last two years. Smith and Meyer both started as average quarterbacks (from a passing standpoint, anyway) and finished as average quarterbacks. Colt McCoy was destined to fall from his astronomical debut, and did, though he didn't crash: he made a lot of mistakes last year (18 interceptions?), but his completion percentage, yards per attempt and overall efficiency were still above average. Given his very meh status as a recruit and average physical skills, the best guess is that the pretty good Colt McCoy of 2007 – as opposed to very good Colt of his freshman campaign – is closer to the version Texas will get this year, and probably in 2009, too.

Allan Evridge is penciled in to start at Wisconsin this year; I hope they're not expecting too much out of him.

But Donovan Woods? His travails are a mystery, unless you consider the much greater hype that followed him in the form of Bobby Reid, who replaced him as a sophomore and, of course, wound up setting off the most famous YouTube rant of last season. Woods' production was bound to fall in the short-term as surely as Reid's was to rise (and Zac Robinson was no slouch after he replaced Reid last year), but Mike Gundy is probably the only coach in the country with a solid four-year starter at quarterback playing safety and/or outside linebacker.

It still seems the guys who improve from mediocre debuts or hold the line from good starts are the ones with the highest expectations and the best long-term prospects for making the next level; everyone else quickly levels off. We'll see what trends hold going forward.