Sunday, April 24, 2016

Should Rookie QBs Start Right Away?

Anyone remember an Eagles off-season where they didn't make headlines?

Howie Roseman and company just set off a massive trade bomb, exchanging a bounty of current and future draft picks with the Cleveland Browns in exchange for the second overall pick in the draft -- which the team will use on a quarterback.

Now let's ignore which of the top quarterback prospects the Eagles are going to take -- and let's focus on the choice the Eagles will have to make after they draft in the highest slot since they took that guy from Syracuse.

Should they start the rookie?

Now we know what Roseman has indicated they'll do. It appears the Eagles will keep Sam Bradford and Chase Daniel and let them be the primary quarterbacks this year (Sam Bradford's rumored trade demand notwithstanding).

But that's not our question -- our question is what SHOULD the Eagles do with their newly drafted first round QB?

I was having this debate over email with friends -- is it better to start a rookie QB right away or to sit him for a year. We were talking pros and cons of the two approaches, but kept reaching for standard QB narrative tropes -- that sitting would let him learn, avoid bad habits, etc.

We've heard all these things before. There are people who would say that rookies really benefit from time to 'learn the speed of the NFL game.' There are others who would argue the best way to learn the game is to actually play it. Both the sit and wait camp (look what happened to David Carr) and the start them right away (Andrew Luck is good!) have tons on anecdotes to throw around.

But no one really looked at the data -- and some brief google searches didn't scratch my itch -- so I thought we should fix that.

I pulled together data from all first round quarterbacks drafted since 2000, and looked across each of them to identify the first year they truly became 'starting quarterback.' That distinction is based on playing time (if you started most of your team's games) and media reports on when quarterbacks were named starter going into the season.

We have have two groups -- guys who started immediately and guys who had to wait. With that distinction, we can compare how they performed and see if either group is distinctly outperforming the other. The implications could suggest how the Eagles SHOULD handle their new QB of the future.


We have 42 quarterbacks in our sample -- and the split is roughly 60/40 -- 24 of those quarterbacks started immediately and the remaining 18 took over the job some year(s) after their rookie season. Below is a table of our QBs and how they were defined:

I pulled in a whole slew of performance metrics compared the two groups. And what's key in comparing the two groups of QBs to each other is lining up their first starting years with each other. It makes no sense to compare the rookie years of both groups when one didn't play. So in our analysis, you'll see I aligned all the quarterbacks around their respective First Starting Year - the first season where they were their team's starting QB.

If you'd argue that sitting a rookie quarterback for a year or more to acclimate to the NFL would be beneficial, then we should see some outperformance either in that First Starting Year or in their career performance trajectory.

Below are illustrations of average performance of those QBs who started immediately after being drafted (orange) and those who had to wait (blue). There are three charts, the first is using QB rating, the second is using Passing Yards per Attempt, and the third is using Pro-Football-Reference's Approximate Value metric. See if you notice a pattern.

In looking across all these metrics -- I wasn't able to identify any clear outperformance by quarterbacks who waited to get their job. Across all the variables, the guys who wait extra year(s) perform pretty much just as well as the guys who start right away.

Also - you don't see any significant trajectory changes over time - by that I mean the guys who wait don't get any kind of long-term performance advantage in doing so.

The only difference I could really see was in their ages (obviously). QBs starting immediately are on average, 22.8 years old. QBs who wait, they're 24.1 when they get the job. It just doesn't appear as though there's any difference beyond a slight age gap.


This data would suggest that all that time riding the pine isn't really worth it for the QB's ultimate performance. They sit, they go to meetings and practices, but once they actually start they do just about as well as the guys who are thrown right into the fire.

Now, if we were in college and debating the merit of red-shirting freshmen quarterbacks, there's really not much downside to having the prospect wait. College players have a defined window of eligibility and cannot play beyond it. But NFL rookie QBs have much longer potential tenures, and they start their careers on fixed contract terms with clocks that tick even when they don't play.

If you're a college coach, redshirting a quarterback means you'll still get four years of performance, you just need to wait one year.

If you're an NFL GM/coach and you have a first round rookie quarterback, that quarterback is on a fixed four-year contract (five with a team option). If you bench that quarterback for a year, that contact becomes a three-year deal with an option. Benching him does come at a cost - because that's one less year on his rookie deal.

If you keep a quarterback on the bench for a year and he turns out to be a superstar, well then you've wasted a year of his term. If he's a bust, then it's kind of a moot point because you're probably going to lose your job regardless.

What's particularly interesting is that the NFL may have figured this out.

Below is an illustration of all the QBs in our sample by draft year. They're split across our two buckets, those who waited to start and those who started immediately. See if anything jumps out at you.

The practice of benching a rookie first round quarterback for a year or more to learn the ropes has declined noticeably since 2000. Before 2008, only one-third of first round QBs started right away. Since 2008? About 80% of them have started immediately.

It's now much more rare to see a QB drafted in the first round and benched for a year. The only one drafted in the last few years who wasn't made starter on Day 1?

Johnny Manziel.

While it's hard to identify a specific driver for the shift in team behavior -- it may be that teams have looked at the data and come to the same conclusion -- that you don't gain very much by waiting.

So what does that mean for the Eagles?

Well, data suggests there's likely no performance benefit to sitting the new quarterback on the bench a year. We could debate that point if that quarterback is Carson Wentz, who lacks playing experience -- but I doubt he's a special exception to the overall trend.

If there's no real benefit to benching him, then what you SHOULD do is look to move Sam Bradford. While there are some sunk financial costs from the new deal you just signed him to -- moving him might be a way to recoup some of the draft capital you burnt to move up in the draft. That's probably what I'd do.

Some might argue that Bradford gives you the best chance to win now and that in a weakened NFC East you should definitely take that chance. My counter to that would be as follows:

- You've just spent a ton of capital to acquire a QB, and the data suggests keeping him on the bench won't make him that much better
- If the new quarterback can't be ready to play right away -- why is he worth giving up so much in draft assets
- Sam Bradford is not a Super Bowl caliber quarterback - demonstrated by his entire career to date - so what's the point in continuing to run him out there if the ceiling is limited?

It doesn't appear that the Eagles are going to go this route -- all indications are Bradford will be the starter and the Eagles new quarterback will sit and wait. Just know that the evidence suggests an extra year of waiting won't make much of a difference.

No comments: