Theories abound as to what the Browns will do in the NFL draft, ranging from going with the same quarterback corps as last year, or signing a veteran to a multiyear contract or drafting a quarterback with the first overall pick.
My thinking is that Sashi and the Browns front office are very heavy into “analytics” or mathmatical valuation theory. I approve of that, by the way. It doesn't replace the need for good scouting, but it does help to make rational decisions about how to value players, and whether to trade up or down. All teams do that, and there needs to be some method to the madness. I'm okay with letting Harvard analysts figure it out.
If that sounds complicated, maybe we can summarize it this way: analytics tells you to avoid overpaying for something. You don't always get your first choice, but you don't make outrageous investments either.
For that reason, I don't think the Browns will throw draft picks and a huge contract at Jimmy Garoppolo of the Patriots. It doesn't sound like any kind of analytics would tell you to blow valuable draft picks and issue a Brock Osweiler sized deal ($18 M per year, say).
If the scouts are totally in love with a propspect (Mitch Trubisky? Pat Mahomes?), the Browns may draft a qb very early and perhaps even first overall. But even that seems to overvalue the quarterback position.
My own version of analytics say that historically teams have drafted quarterbacks too high, despite the fact that they have trouble distinguishing between prospects in the first and second round. The best chance to get a good quarterback is in Rounds 2 to 4.
This conclusion was reached by going back through recent history and listing every quarterback taken since 2012 and noting the draft position and assigning Schramm points to each pick (the so-called Schramm system assigns a point value to each pick, ranging from 3000 for the first pick overall, down to 2 for the seventh round). Then I regraded the draft to estimate how many Schramm points the guy should have been worth. For example in 2012, Andrew Luck was the first overall, Robert Griffin III was second overall, while Russel Wilson was a third round pick and Kirk Cousins was a fourth round pick. In regrading the draft, I put Wilson first overall, followed by Luck and Kirk Cousins moved up to the first round. I figured RG3 had a sensational year but so-so the rest of his career, so I put him in the middle of the second round. So RG turned out to be worth fewer Schramm points, whereas Russell Wilson was a huge bargain.
Anyway, I added up the results for each round since 2012 and graph the results below. The first chart shows the percentage of quarterbacks who were selected in each round that became decent starting quarterbacks, roughly in the Top 20 or better. The chart shows first of all that there were no quarterbacks in rounds 5, 6 and 7 that fit that description, with Trevor Semian being the closest. I don't think Trevor would be a first round pick if the 2015 draft were re-graded, so that's why I'm saying there were no "good' starting quarterbacks from the late rounds from 2012-2016. The results also show that guys taken in Round 1 really do have a higher chance of making it and becoming a successful quarterback. However, Round 2 picks are about as successful as Round 1 picks.
The second chart shows the return on draft capital (how many Schramm points were gained according to my re-grade, versus the draft cost of the pick). One thing that stands out is that the average return on investment is negative in Round 1. This means that the NFL teams are taking quarterbacks too high in Round 1. They are not actually better than second round picks so that shows up as a negative return on investment.
The second chart shows the return on draft capital (how many Schramm points were gained according to my re-grade, versus the draft cost of the pick). One thing that stands out is that the average return on investment is negative in Round 1. This means that the NFL teams are taking quarterbacks too high in Round 1. They are not actually better than second round picks so that shows up as a negative return on investment.
In addition, Rounds 3 and Rounds 4 have produced very good quarterbacks like Russel Wilson, Kirk Cousins and Dak Prescott. Even though many quarterbacks don't make it from those rounds, the data says you have a decent chance to score big in those rounds so the return on investment is very attractive.
Rounds 5, 6 and 7 have not produced superior starters recently, although going back a while a guy named Brady did okay. However, a large number of late round picks are still in the league as backups. I wasnt real sure how to rate these guys, but it's fair to say that usually teams are not willing to give up even late round draft picks for backup quarterbacks, except in rare cases like Jimmy Garoppolo whom I judge to be worth a number one pick.
 |
| Figure 1. Chances of drafting a Top 20 NFL quarterback by round based on results from 2012 to 2016. |
 |
| Figure 2. Return on draft capital (Schramm point system), as subjectively evaluated by the author in a hypothetical redraft. Quarterbacks in Round One often underachieve, meaning that the return on draft capital is negative (teams usually regret their pick in Round 1). Rounds 2,3 and 4 have positive returns meaning that they invest very little and hit it big often enough to make it worthwhile. |
The data shows that, though there are always exceptions, the best place to take a qb is usually Rounds 2-4. In those rounds the chances of getting a very good quarterback are about the same as Round 1, but much less draft capital is expended.
If you believe these numbers, perhaps a team should even be willing to draft more than one guy, and to keep an extra roster position for the extra qb. If it were me (or maybe Sashi), it could be worth taking two guys around Rounds 2-4 rather than one guy near the top of Round 1. The Redskins took two quarterbacks in 2012, and fourth rounder Kirk Cousins turned out better than first rounder RG3, if you recall.
If the quarterback position is really so important, why not consider having four quarterbacks on the 53 man roster? Nobody in the NFL does that, but that is exactly the unorthodox conclusion that analytics might lead you to consider, if your scouts tell you that all four guys have a chance to develop.
If you believe these numbers, perhaps a team should even be willing to draft more than one guy, and to keep an extra roster position for the extra qb. If it were me (or maybe Sashi), it could be worth taking two guys around Rounds 2-4 rather than one guy near the top of Round 1. The Redskins took two quarterbacks in 2012, and fourth rounder Kirk Cousins turned out better than first rounder RG3, if you recall.
If the quarterback position is really so important, why not consider having four quarterbacks on the 53 man roster? Nobody in the NFL does that, but that is exactly the unorthodox conclusion that analytics might lead you to consider, if your scouts tell you that all four guys have a chance to develop.
I'll be the first to advocate that the 53rd guy is an important member of the team, even if he is a third string guy. But another roster spot to develop another quarterback would be much more important if it increases your chances of getting the franchise guy, and I believe it does.
In the case of the Browns, if they don't feel comfortable long term with either Cody Kessler or RG3, I would cut RG3 immediately on the grounds that although he appears to be good enough (in my opinion) to win some games, his skills are best suited for pistol read option style plays, which is not the long term right answer. If RG3 is not a threat to make the 2018 Pro Bowl, it's time to cut ties.
I would install Cody Kessler as the quarterback of the present (like the Broncos did with Trevor Semian, even though they had a hotshot first round pick on the bench in Paxton Lynch). Cody is good enough that he is not going to kill the team, and he might develop further. I would draft one or even two guys in this upcoming draft, given that it is a deep draft for quarterbacks this year. There are several intriguing prospects that have size, speed, production and arm strength, though few have four years of college these days. It's worth it to gamble in a middle round; there could be another Dak Prescott or Russell Wilson in there. Find him.
It's time to cut ties and go back to the draft as soon as the Coach believes the quarterback is not developing beyond the Seneca Wallace level (i.e. even if he is a valuable backup player but not a threat to become a "franchise"). The numbers say that any given pick is not more than 40% likely to be the right guy, so there are going to be some casualties along the way. But the data says the odds are in your favor by taking guys in the middle rounds. Blowing it on a middle round pick is less painful than blowing it on a high first round pick or costly free agent. Unless that quarterback is an Andrew Luck (big, strong, fast, accurate, athletic and sings in the choir at church), I would resist that temptation to draft that hotshot guy near the top of Round One. Later in the round or Rounds 2-4 are better investments.
Certainly different people would re-grade the drafts differently than I did, though it's usually a no-brainer (Russell Wilson near the top of a re-draft, RG3 not). I suspect that Sashi and the boys are doing something similar, though in a lot more detail and with much greater care. But I believe in the basic sensibility of the approach. My guess is that the Browns will not throw huge dollars and high draft choices at the quarterback position, but instead try to grow a home-made quarterback from the draft with one or two Kessler-type draft picks per year for while.
.
I would install Cody Kessler as the quarterback of the present (like the Broncos did with Trevor Semian, even though they had a hotshot first round pick on the bench in Paxton Lynch). Cody is good enough that he is not going to kill the team, and he might develop further. I would draft one or even two guys in this upcoming draft, given that it is a deep draft for quarterbacks this year. There are several intriguing prospects that have size, speed, production and arm strength, though few have four years of college these days. It's worth it to gamble in a middle round; there could be another Dak Prescott or Russell Wilson in there. Find him.
It's time to cut ties and go back to the draft as soon as the Coach believes the quarterback is not developing beyond the Seneca Wallace level (i.e. even if he is a valuable backup player but not a threat to become a "franchise"). The numbers say that any given pick is not more than 40% likely to be the right guy, so there are going to be some casualties along the way. But the data says the odds are in your favor by taking guys in the middle rounds. Blowing it on a middle round pick is less painful than blowing it on a high first round pick or costly free agent. Unless that quarterback is an Andrew Luck (big, strong, fast, accurate, athletic and sings in the choir at church), I would resist that temptation to draft that hotshot guy near the top of Round One. Later in the round or Rounds 2-4 are better investments.
Certainly different people would re-grade the drafts differently than I did, though it's usually a no-brainer (Russell Wilson near the top of a re-draft, RG3 not). I suspect that Sashi and the boys are doing something similar, though in a lot more detail and with much greater care. But I believe in the basic sensibility of the approach. My guess is that the Browns will not throw huge dollars and high draft choices at the quarterback position, but instead try to grow a home-made quarterback from the draft with one or two Kessler-type draft picks per year for while.
.
No comments:
Post a Comment