When do you KNOW a QB will be good? Solving the Wentz and Prescott question

USA TODAY Sports, Getty Images

Carson Wentz and Dak Prescott look awesome this season, you might have heard. Neither Prescott nor Wentz would have expected to start the year under center for his team when the preseason started, but the injury to Tony Romo and the trade of Sam Bradford opened up starting spots for the two rookie quarterbacks, and here we are. Neither NFC East starter has thrown an interception through his first three starts. Prescott ranks fourth in the league in Total QBR with an 82.5 mark, while Wentz comes in at 11th at 66.0. The two passers' teams have combined to start 5-1.

Nobody can argue that Prescott and Wentz have been disappointing. Quite the opposite, actually: Draftniks and fans who believed in the two quarterbacks before this year's draft have been patting themselves on the back for backing the right horses. Eagles general manager Howie Roseman has garnered credit for trading a pile of draft value to grab Wentz, while the hapless 0-3 Browns are already being dragged through the ringer after trading away the opportunity to draft him with the second overall pick even with all the draft picks now in hand.

History also tells us, though, that there are examples of touted quarterbacks who lived up to the hype before fading, some fast.

Most recently, Robert Griffin III looked for all the world to be a superstar after his first season, not just first three games, and the Rams took flak for their decision to trade away the second pick and allow Washington to grab him. That trade turned out to be prescient. Mark Sanchez won his first three starts, including a 16-9 victory over Tom Brady. Ryan Leaf won his first two starts, albeit with middling numbers, before throwing one touchdown against 13 interceptions in his eight ensuing contests as a rookie, in which San Diego went 1-7. You get the idea. There are many others you could point to.

We look back now and recognize the obvious flaws in their cases, but at the time, those issues seemed like they were small problems in an otherwise impressive package. Maybe we'll look back at Prescott and see how he collapsed when he wasn't supported by a dominant offensive line and a great running game, or rue how Wentz leaned into too many big hits and failed to adapt when teams took away his short passes. It's way too early to say.

Or is it?

Players will always take weird career paths, but on the whole, when can we say with any confidence that a quarterback who is playing at a high level to start his career will continue to play that way? How long does it take for a quarterback to prove his mettle -- or, alternately, how quickly can we identify the ones who don't have it? It would be fair to say that answering the question is an inexact science, but let's take a stab at that exact problem here.

There's no correct way to do this, because one person's definition of success is different from another's. The methodology I landed on was to sort every quarterback who has thrown 10 passes or more during his first four seasons since entering the league in 1990 into one of six groups, which I'll explain here. The choices are arbitrary, but they're designed to make it easier to identify successful quarterbacks, and they were made before looking at any of the data on hand.

Group 1: The superstars. The no-doubt perennial Pro Bowlers, annual MVP candidates and plausible future Hall of Famers. This group ranges from Tom Brady and Aaron Rodgers on the high end to Eli Manning and Carson Palmer on the low end. Obviously, it would be great if Prescott and Wentz ended up here.

Group 2: The stars. These are the quarterbacks who put together (or are putting together) lengthy careers as typically above-average regulars, with a year or two where they leaped into the upper echelon. Your typical quarterbacks in this group are guys like Drew Bledsoe, Joe Flacco and Michael Vick. It would be very exciting for Prescott to end up here given his draft status, and it would be a fair return for Wentz as the second overall pick.

Group 3: The flashes. These are passers who have had one or even two big seasons over the course of their careers but were too inconsistent or injury-prone to stay as above-average regulars at the position. The group includes quarterbacks like Colin Kaepernick and Griffin, along with players like Jeff George and Marc Bulger. This is about the lowest Prescott could fall and still be a successful quarterback given his draft return. The Eagles expect more of Wentz after trading so much draft capital for the former North Dakota State star.

Group 4: The first-round disappointments. Can't be much clearer than this. The word "bust" is generally unfair to use given how many variables go into developing a quarterback, but these are the guys who did not become their team's (or any team's) franchise quarterback after being taken in the first round. It's everyone from JaMarcus Russell to Leaf to Johnny Manziel. Obviously, Prescott wasn't a first-round pick, but neither he nor Wentz wants to end up looking similar to the players in this group.

Group 5: The to-be-determineds. These are the recent quarterbacks who are still either on rookie deals (such as Blake Bortles) or in the first year of their second contracts (such as Brock Osweiler). It's too early to put them in any of the other groups. It's important to separate them out from everyone else, but they're mostly irrelevant in terms of projecting the long-term performance of Prescott and Wentz; by the time we know what we have with those two, everyone in this group will have his own future resolved, too.

Group 6: Everyone else. This is a big bucket of backup quarterbacks and replacement-level journeymen who weren't taken in the first round of the draft. There's no shame in that, of course, but it's not the group that either of these promising rookies wants to end up occupying.

I went through the game logs of each quarterback who made his debut after 1990, a 25-year sample, and tracked how they all performed, appearance by appearance. I included only games in which they threw 10 passes or more and contests that took place during the first four years of their career, the length of the typical rookie contract. To measure each player's performance, I used adjusted yards per attempt, which is a souped-up passer rating; to adjust it for era, I compared the AY/A to the league average in the year in which it was recorded. So if Aaron Rodgers posted 8.0 AY/A last year and the league average was 7.0, Drew Brees was 14.2 percent above average.

Let's test different lengths of time and see how our groups differed in terms of their average AY/A versus the league baseline to see where the stars broke away from the pack. For reference, through three games, Wentz's 8.5 AY/A is 21.4 percent better than the league average of 7.0, while Prescott's 7.9 AY/A figure is 12.9 percent ahead of the mean.

What we see after four games: Let's start with what qualifies as a round number in football. A quarter of a season seems way too early to declare anything, but even after four games and as few as 40 pass attempts, we can already see a pretty notable difference between the absolute top-level stars and the vast majority of other rookie quarterbacks. That's not to say there aren't exceptions. Both Eli and Peyton Manning were messes as rookies. Palmer sat the entirety of his rookie year (zero throws) and still struggled during his first four starts.

On the other hand, Jake Locker was a force of nature in his first four appearances, posting 8.5 AY/A with five touchdowns against one pick. Tim Tebow was effective in his own weird way by throwing downfield without even considering his rushing value. Dave Brown of the Giants went 3-1 with a completion percentage and level of efficiency that he would fail to match the rest of the way. There's a gap, and plenty of examples suggest that we would struggle to tell the transcendent players from the eventual disappointments. And the second tier of useful long-term starters is actually still bad.

What we see after eight games: With eight games in the books, everyone improves notably besides the busts, who never really get any better. There are still effective first-round busts like Patrick Ramsey and Andre Ware, but just nine of the 31 players in the group post an AY/A better than league average over their first eight appearances. Twelve of the 17 superstars are on the positive side of the ledger, with Donovan McNabb the low man on the totem pole. We also still really can't separate the pretty good long-term contributors from the guys who are merely off to hot starts.

What we see after 16 games: With a full 16 games of data to look at, the useful quarterbacks in the group really begin to pull away. Perhaps more accurately, the disappointing picks have revealed themselves to be mistakes. The only players in Group 4 to post AY/A figures 10 percent better than league average over their first 16 games are Ramsey and Tim Couch, each of whom had his career waylaid by coaching changes, pass-protection issues and injuries. Leaf, meanwhile, has seven touchdowns against 24 picks in the qualifying games with a AY/A 43.3 percent below league average. Figured something out there.

What we see after 32 games: It looks like it really takes two full seasons' worth of data to get a real sense of separation between useful, productive quarterbacks and those who failed to live up to expectations. Unsurprisingly given their name, the flashes in the pan have regressed back toward league average, while the useful contributors have showed the consistency that separates them from those inconsistent options. The best first-round bust is David Carr, whose AY/A is 6.7 percent above league average. The worst superstar is Brees, who is 3.6 percent below league average. It's worth remembering that Brees struggled enough during his first two seasons as a starter that the Chargers handed the job over for a time to Doug Flutie before acquiring Philip Rivers in a draft-day trade during the 2004 offseason.

It wasn't until his fourth season in the league that Brees played like a Pro Bowl quarterback, and he's been excellent ever since.

The lesson to take away from all of this, as best I can tell from history, is that the excitement around Prescott and Wentz is justified, in part because they've managed to avoid failing immediately. The washout rate for players who struggle at the very beginning of their professional careers, even first-round picks, is higher than I expected. Whether by a lack of opportunity or an inability to adapt, cases like that of Brees are few and far between.

At the same time, it's way too early to crown Prescott and Wentz as future superstars and come up with trade packages for Romo while laughing at the Browns. They're not going to go the rest of their respective careers without throwing an interception. They're going to lose key pieces, as Prescott is likely to experience this weekend with Dez Bryant unlikely to play. They'll go down early in games and have to throw against teams teeing off on passes to catch up. They'll take hits and have to play at 80 percent, and even worse, they might develop bad habits with relation to the pass rush to avoid getting hit again. Defenses will make adjustments, and Prescott and Wentz and the ex-quarterbacks who coach them will make adjustments, and we may very well end up in a different place from where we all started.

As tempting as it is to crow about a future secured, Cowboys and Eagles fans should probably wait at least until the end of the season before they start naming their children Dak and Carson. Put it this way: If everything works out right, they'll have plenty of time to brag about their new franchise quarterbacks.