Which Quarterback in the 2015 NFL Draft Is Statistically Superior?
After the final snap of the Super Bowl, the lives of NFL enthusiasts hit a dark time. Every fan who doesn't root for the New England Patriots is forced to sit around and ask why their favorite team sucks.
Well, my woeful comrades, we can instead look to a much happier time: the NFL Draft. This is where sunshine and daisies overwhelm the demons of a 4-12 season and allow every team to pretend next year will suck a little less.
This is especially true for teams in search of a quarterback. In most instances, when you're in search of one of these dudes, it's probably because your team pooed the bed the previous year and is in need of some serious overhaul. But do any of the quarterbacks in this year's class provide such optimism?
Last year, we ran a little study to try to develop the formula for picking out a first-round quarterback bust based on their collegiate stats. The conclusion we came away with is that, although it is incredibly difficult to peg which quarterbacks will be busts based on their numbers in college, you can at the very least decide which guys may be riskier selections earlier on.
Prior to last year's draft, we used this data to conclude that Teddy Bridgewater was the statistically superior quarterback of the draft class. He finished the season as, far and away, the best rookie quarterback, and one of the top-10 rookie signal-callers since 2000.
The numbers weren't particularly keen on Blake Bortles and Derek Carr, who finished the season last and second-to-last respectively in numberFire's Net Expected Points (NEP) metric. This doesn't mean that this method necessarily works, but it may still be useful.
The statistics we used to evaluate these guys were the number of games in college in which they attempted at least 10 passes (which will be just "games played" from here on out), their Passer Efficiency Rating, Adjusted Yards per Attempt (a Sports Reference stat that factors in touchdowns and interceptions thrown with their regular yards per attempt) and ESPN's Total QBR, although that has been by far the least helpful predictor of professional success.
So, now, it's time to look at this year's crop of young pups to see which ones are "safer" picks and which ones will run a lot of risk if taken in the first round. We'll only be looking at the top-five quarterbacks in this year's draft class from Mel Kiper's positional rankings. This presents to us Jameis Winston, Marcus Mariota, Brett Hundley, Bryce Petty and Garrett Grayson. Let's see what the numbers say about each one.
1. Marcus Mariota, Oregon Ducks
Games Played: 41 | Passer Efficiency Rating: 181.7 | Adjusted Yards per Attempt: 11.5 | Total QBR: 90.8 | Top Statistical Comp: Cam Newton
Marcus Mariota is a statistical deity that thou shalt not slander. Not only is he the best quarterback of this class, he is one of the best of the past 20 years.
Of the 50 quarterbacks that have been taken in the first round since 1995, only two have had higher Passer Efficiency Ratings during their final collegiate seasons than Mariota had this year. Those guys are Robert Griffin III and Cam Newton. Despite RGIII's recent struggles, he still had the third-best season for a rookie quarterback since 2000, with Newton following in fifth place.
Mariota's Adjusted Yards per Attempt is the second best of these quarterbacks (behind Griffin) and the sixth best in NCAA history. Had he not thrown an interception on the final play of the National Championshp Game, he would have ranked fifth. This is largely because of Mariota's lack of interceptions -- only four this year and 14 through 41 career games. We'll revisit that in the section on Winston.
The thing that helps solidify Mariota as being on the safer side of things is the number of games he's played. The average quarterback in the "Top Tier" of last year's study (quarterbacks that have averaged a Total NEP of 20.0 in NFL seasons in which they have dropped back to pass at least 200 times) played in 37.37 collegiate games prior to entering the NFL. That number for the "busts" in the bottom tier (based on their average Total NEP being flat-out gross) shrinks to 31.58. First-round picks with loads of collegiate experience are far more likely to be successes in the NFL. Mariota fits that profile.
Once we look at all of this, Mariota ranks above average within the Top Tier of quarterbacks in each of the categories. The only other players to play in at least 35 games with Passer Efficiency Ratings above 170 and Adjusted Yards per Attempt above 10.0 are Griffin, Bridgewater, Philip Rivers, Chad Pennington and Jason Campbell. Rivers has finished in the top 10 in Total NEP seven times in the past nine seasons. Pennington did so three times and had better NEP numbers than you might expect. Campbell is the lone disappointment from this group, and even he averaged a Total NEP of 16.395 in the six seasons in which he dropped back at least 200 times, leaving him just short of the cut-off.
This is where the cries about Mariota being a system quarterback come in. If so, he's a pretty darn good one. The three main guys at Oregon since Chip Kelly's arrival in 2007 outside of Mariota have been Dennis Dixon, Jeremiah Masoli and Darron Thomas. Mariota's worst Passer Efficiency Rating (163.2 as a redshirt freshman) is better than the best season of any of those other guys (Dixon at 161.2 his final year). And it's not like these other quarterbacks taken in the first round were surrounded by dog-poo talent, either. It doesn't mean it's not a concern; it's just less so for me than people are making it out to be.
Mariota's top comp is Newton, but he may actually be a safer pick than Newton was. While their passing stats are incredibly similar, Mariota actually played 15 more games than Newton did, even when you include the games Newton played in junior college. If I'm a team that needs a quarterback in this class, give me Mariota, or give me the box of tissues so I can cry away my sorrows.
2. Garrett Grayson, Colorado State
Games Played: 35 | Passer Efficiency Rating: 166.2 | Adjusted Yards per Attempt: 10.3 | Total QBR: 71.7 | Top Statistical Comp: Ben Roethlisberger
This is where things start to get interesting. Mariota was an easy selection for the top spot. Although things tightened up after that, Garrett Grayson still appears to have an advantage over the rest.
In each and every one of the categories that we examined, Mariota was first in the class. Grayson ranked third in games played and Total QBR, but then kicked Brett Hundley's behind in Passer Efficiency Rating and Adjusted Yards per Attempt. Thus, the least-known hombre of the pack ends up here.
The biggest question about Grayson is whether or not his numbers are valid because Colorado State is the lone school not in a Power 5 conference which we are investigating. So, we go back to other guys in similar situations to see how they fared.
Let's take this to the complete extreme and look at a pair of guys that went in the first round that faced even lesser competition than Grayson - Joe Flacco at Delaware and the late Steve McNair at Alcorn State.
Delaware is a Division-I school, but it's in the FCS. In his final year with the Fightin' Blue Hens, Flacco had a Passer Efficiency Rating of 144.9 and an Adjusted Yards per Attempt of 8.6. This would have put him below average in both statistics, yet he still ended up being a good enough quarterback at the NFL level.
McNair's numbers there weren't bad with a 155.4 Passer Efficiency Rating and 9.0 Adjusted Yards per Attempt. When you adjust for the era in which he was playing, the numbers look pretty solid.
What these two guys show is that you don't need to post disproportionately good numbers at lower levels to be a starter in the NFL. And it's not as if Grayson was at some obscure school. That said, Grayson is far from a sure thing.
Colorado State played nine games against non-Power 5 opponents this year. In those games, Grayson had a 175.42 Passer Efficiency Rating. Good! But, in their three games against Power 5 programs, that number dropped to 119.11. Not as bueno.
But you also have to think about it this way. It's not just Grayson going against higher competition during these games, but his teammates as well. If the offensive line can't hold up or his receivers can't get open, then there's not a whole lot Grayson can do. Because quarterback can be so dependent on the performance of others on the team, I am hesitant to discount Grayson as a prospect simply because his numbers sagged against tougher competition.
3. Bryce Petty, Baylor
Games Played: 25 | Passer Efficiency Rating: 157.8 | Adjusted Yards per Attempt: 9.6 | Total QBR: 75.1 | Top Statistical Comp: Ryan Leaf
I was so ready to slobber all over Petty at the end of last season. Remember how Mariota finished sixth all time in single-season Adjusted Passing Yards per Attempt? The guy right in front of him was Petty in 2013. Filthiness.
And yet here we are, ranking Petty third statistically in a class of throughly mediocre quarterbacks. Welp City, yo. How did we get here?
The obvious answer with Petty is injuries. He sustained the same injury as both Newton and Tony Romo did. He then went on to sustain a concussion later in the year, right when things were starting to get right.
Let's break Petty's season into two halves. In the first six games, which included the game in which he hurt his back and the five following, Petty's Adjusted Yards per Attempt was 9.17, and he had a Passer Efficiency Rating greater than 160 just once. In the final six games, Petty's Adjusted Yards per Attempt leapt to 10.08, and his Passer Efficiency Rating topped 160 in every game except for one.
Then we get to Petty's top comp in Ryan Leaf. The Grand Puba of quarterback suckitude. The biggest red flag similarity between the two is the number of games played. The fewer games played, the higher the unpredictability. Leaf also bested Petty in both Passer Efficiency Rating and Adjusted Yards per Attempt, which is concerning. But, obviously, Leaf had other problems (read: he was hella balls crazy, yo), so let's hope Petty doesn't meander down that route.
The biggest competition Petty would have for this spot is Hundley. Hundley played 14 more collegiate games than Petty. Petty, however, showed last year what he can do when he's healthy. He led the nation in Adjusted Yards per Attempt, and then turned around and still posted better numbers than Hundley while playing hurt this year. For that reason, I have no problem with placing Petty third in these rankings.
4. Brett Hundley, UCLA
Games Played: 39 | Passer Efficiency Rating: 152.7 | Adjusted Yards per Attempt: 8.6 | Total QBR: 78.6 | Top Statistical Comp: Eli Manning
Hundley finished second among these five in both games played and Total QBR. However, his other, more significant rate statistics leave Hundley as a guy who appears to have a fairly low ceiling in the NFL.
The crazy thing about Hundley is how similar he was in each of his three seasons leading the Bruins. His Passer Efficiency Rating was never above 152.9 nor below 147.7. His Adjusted Yards per Attempt was never above 8.6 nor below 8.0. You can either use this as a positive mark for consistency or a negative for a lack of progression.
Despite playing in the same conference as Mariota over the same time frame as his Oregon counterpart, not even Hundley's best statistical season touched Mariota's worst. That's more of an endorsement of Mariota than an indictment of Hundley, but it should show that this quarterback class, as many have stated, is not particularly strong.
This brings us to Hundley's top comp of Eli Manning. Manning finished the 2003 season behind prolific passers such as Rivers, Ben Roethlisberger, Bruce Gradkowski and Kevin Kolb. Honestly, Eli's numbers were frustratingly mediocre for a No. 1 overall pick, especially when stacked up with those of Rivers and Roethlisberger. Considering Manning has finished in the top 10 in Total NEP three times while Rivers and Roethlisberger have done it seven times apiece, maybe, in retrospect, that wasn't the right selection, anyway.
Hundley certainly has the build of what many are looking for in an NFL quarterback at 6'3", 227 pounds. But, as we have seen with guys taken after the first that have had success, that's a poor method of evaluation. Hundley isn't a terrible prospect, but he certainly doesn't offer anything inspiring statistically.
5. Jameis Winston, Florida State
Games Played: 27 | Passer Efficiency Rating: 145.5 | Adjusted Yards per Attempt: 7.7 | Total QBR: 74.3 | Top Statistical Comp: Tim Couch
You could very easily make an argument to rank Winston here on reasons totally unrelated to football. But, when looking at the stats, you could make a similar argument without even venturing off the field.
Like Petty, Winston was off-the-charts awesome last year. He led the league in Passer Efficiency Rating at 184.8 and was second behind Petty in Adjusted Yards per Attempt at 11.7. If he had posted those numbers again this year, he would have given Mariota a serious run for his money. Instead, he's last on this list. Womp.
Of these five quarterbacks, Winston ranks last in Passer Efficiency Rating and Adjusted Yards per Attempt and second to last in games played and Total QBR. He was the only quarterback to finish last in more than one category as Petty was last in games played and Grayson was last in Total QBR. Based on 2014, Winston is clearly the fifth-ranked statistical quarterback.
It isn't just that Winston stacks up poorly to this class. He is in the danger zone for every stat across the board. Of the 55 quarterbacks involved in this (50 first round picks since 1995 and this year's top five guys), Winston ranked 40th in Passer Efficiency Rating, cozy up behind Brady Quinn. His Adjusted Yards per Attempt was the 44th best, between Josh Freeman and J.P. Losman. His 27 games played are the 42nd most, planting him right behind Brandon Weeden (Petty was 49th).
A large portion of Winston's struggles statistically were because of his interceptions. His 18 interceptions were second in the nation behind only Tyler Rogers of New Mexico State. Winston threw four more interceptions in 467 attempts this year than Mariota threw in 1,167 career attempts.
In Florida State's victory over Florida, Winston managed to throw as many interceptions in 24 attempts as Mariota threw the entire season (four). He threw multiple interceptions in four other games this year, a feat Mariota achieved three times in is career, with none coming this year.
This all doesn't mean that Winston isn't going to succeed. His redshirt freshman season showed that he can do exactly that. But it does make him a risky selection to say the least.
Winston is not the first top prospect that has struggled in his final year at school. Michael Vick had a startlingly similar path to the number-one pick, if that's where Winston ends up.
In his redshirt freshman season, Vick led the nation in Adjusted Yards per Attempt at 11.5 and finished third in Passer Efficiency Rating at 171.1. He then plummeted further than Winston to marks of 7.0 and 127.4, respectively, before declaring for the draft as a redshirt sophomore. Vick just barely reached the top tier of NFL quarterbacks, averaging a Total NEP of 27.71 in the four seasons in which he has recorded at least 200 drop backs.
There are successful quarterbacks that have posted numbers as bad as or worse than Winston's. The problem is that they are far less plentiful at this end of the spectrum than the opposite.
The only top-tier quarterbacks that have finished their final collegiate seasons with Adjusted Yards per Attempt totals less than 8.0 are Vick, Jay Cutler and Matt Ryan. Ryan is the only one to finish in the top 10 in Total NEP more than twice, having done so five times.
Winston's top comp is Tim Couch. In 26 games, Couch had a Passer Efficiency Rating of 153.3 and an Adjusted Yards per Attempt of 7.8. More glowing comparisons!
Again, this all doesn't mean that Winston won't succeed in the NFL. It just means that selecting him, especially in the early portion of the first round, carries an extreme amount of risk. That same risk is not necessarily apparent in a guy like Mariota. Risk can be a good thing when your team is struggling, but not when there are assets with similar upside that are a considerable amount less dangerous on the board.