Jeremy Johnson has one career start, that being last year's opener when Nick Marshall was suspended for the first half. Beyond that, his only appearances with more than five pass attempts happened in 2013 against Western Carolina and Florida Atlantic. Even so, he's generating quite a bit of attention.
Some of that is simply that he's now spent the most time in Gus Malzahn's system of any highly touted starting quarterback. Malzahn's current stint at Auburn ties his previous one for the longest time he's spent in one place at three seasons, and while the third season guys back then of Barrett Trotter and Clint Moseley match Johnson's three years in the system, they weren't as talented as Johnson is. After what we saw Malzahn do in short time with guys like Paul Smith, Cam Newton, and Marshall, Johnson stands to do quite well for himself.
There has to be more to it than that, though. I don't think merely spending three years under a well regarded coach is enough to put someone with so little playing time at 20/1 odds for the Heisman Trophy. Let's dig into this question some.
The closest analogue to Johnson's situation I can think of is Tyler Wilson at Arkansas. He barely played as Ryan Mallett's backup except for the 2010 Auburn game when Mallett went out with a concussion. Wilson came in and performed admirably, going 25/34 (73.5%) for 332 yards (9.8 YPA), 4 TDs, and 2 INTs in a little over a half. The Razorbacks couldn't keep up with the Tigers' high powered offense, but for a guy who had never played in such a pressure-packed environment, that was a great job.
In his first half against Arkansas last year, Johnson was even better than that. His line was 12/16 (75%) for 243 yards (15.2 YPA), 2 TDs, and no INTs. I will throw out there that Arkansas's passing S&P+ defense was 120.5 last year, which is better than 2010 Auburn's 110.8 in the same category. The '14 Hogs probably grew into that number as the year went along, so it's probably a bit high for where they were in the first game. Let's just say that the two defenses in question were in the same neighborhood.
So Johnson might project better than Wilson based on that one measure, although Johnson had the advantage of knowing he was going to play and getting reps in practice befitting that knowledge. Even then, being better than Wilson does not necessarily a Heisman candidate make, as Wilson was never a finalist for the award.
I will take a different approach now, looking at adjusted passing yards per attempt (AY/A). It's a stat that comes from the same kind of place that passing efficiency does, as it adjusts for the positive value of touchdowns and the harms of interceptions. It is yards plus 20 times TDs minus 45 times interceptions all divided by attempts. Johnson is not immobile, but he figures to do much more passing and less running than prior Malzahn QBs like Newton and Marshall did. That's why I'm focusing just on passing here.
Backup quarterbacks who sit behind entrenched starters generally get their work in garbage time. I figured that getting work in garbage time is kind of like the work guys get in spring games: if you do well then, it may or may not mean something, but if you can't do well then, you're probably not going to do well in regular game time.
I looked through a lot of recent SEC quarterbacks to find guys who, like Johnson, spent entire seasons as only a backup prior to starting. I only found one quarterback whose best full-season AY/A was more than about two yards per attempt better than his final year as a backup getting mostly garbage time work. That guy was Connor Shaw, whose best season as a starter (10.1 AY/A as a senior in 2013) blew away his freshman year garbage time work (4.6 AY/A). That low number was highly influenced by two interceptions in just 33 pass attempts, though, without which the gap between that year and his senior year drops from 5.5 to 2.3 AY/A.
Johnson managed 13.4 AY/A on 37 pass attempts as a backup in 2014. I feel comfortable in asserting that 13.4 AY/A is unsustainably high over the course of an entire season as a starter. Newton's 2010 Heisman winning campaign had the highest AY/A for an SEC starter that I saw, and it was 11.2. Unfortunately, Newton never had enough pass attempts as a backup for me to feel comfortable in comparing his backup time to his starting time.
I did see two quarterbacks who managed over 10.0 AY/A as backups though, and they happened to be former teammates of Newton's: Tim Tebow and John Brantley. The former had 12.5 AY/A on 33 attempts in 2006, while the latter had 11.5 AY/A on 48 attempts in 2009. Tebow's best as a starter was 10.6 in 2008, for a drop off of 1.9. Brantley's best was 8.1 in 2011, for a drop off of 3.4. Those were the only two to have a drop off of more than 0.6, by the way.
Brantley's decline had a lot of extenuating circumstances that don't apply to Johnson. He dealt with the turmoil of Urban Meyer's last year in Gainesville, Steve Addazio's inability to craft an offense to suit him, and an astonishing drop in skill position talent from 2009 to 2011. I don't think we learn much about Johnson's chances from Brantley.
Tebow is a different story. He had a better collection of skill position players around him in 2007 and especially 2008 than he did as a freshman. He also ran an offense expertly developed to fit his skills. While he was more of a runner than Johnson will be, his situation is far more comparable than Brantley's is.
Even with all of those things in his favor, Tebow's AY/A still dropped by 1.9 from his time as a backup to his best starting year. If Johnson's AY/A drops by the same amount from last year to this year, he'd post 11.5 AY/A.
If you remember from a few paragraphs ago, Newton posted 11.2 in his Heisman year. That 11.5 mark would match Marcus Mariota's and Jameis Winston's Heisman years and be comfortably ahead of Tebow's (10.4) and Johnny Manziel's (8.8)—though those last two used their legs in a big way in their campaigns. Only Robert Griffin III's 11.8 is ahead of the 11.5 mark. Getting up there doesn't guarantee a Heisman win—Bryce Petty hit 11.7 in 2013, Russell Wilson matched RG3's 11.8 in 2011—but it will almost certainly make you a finalist. Every full-season starter who put up at least 11 AY/A in the past decade has been a finalist except for Colt Brennan (11.0) in 2006, and he was a relative unknown that year playing for a mid-major school.
Hopefully all of that wasn't too dense. The TL;DR summary is this: by one pretty good measure, Jeremy Johnson had the best season of any pure backup quarterback in the SEC in the past 15 years. He won't keep up that same level of play as a starter, but if his decline isn't too dramatic and is consistent with other recent SEC signal callers, he'll be at a level of play in line with several recent Heisman finalists.
I wouldn't agree with Alex Gray's somewhat hyperbolic assertion that Johnson is the SEC's only Heisman hope, but he does appear for right now to be legitimately one of the best ones.