Rookie QBs: To Play or Not To Play

[YoHoChecko]

I have long believed that the best way to maximize a rookie QB's chances for success is to sit that player on the bench for at least one year. A year's worth of practice, digesting the playbook, and acclimating yourself to the routine and pace of a yearlong schedule have invaluable benefits for a player at a position expected not just to participate in, but manage the entirety of the team's offense every snap.

In the past, I have done some fairly detailed analysis of rookie first-round QBs and found that QBs waiting at least a year is beneficial. While I no longer have that data (it would be outdated by now anyway), I can recall that the following bulletpoints were takeaways:

• Veteran QBs who started on teams that drafted 1st round QBs outplayed the 1st round QBs who started as rookies; this gave the teams who waited an on-field advtantage in year one

• The QBs who sat as rookies outplayed the QBs who played as rookies by a small amount in year two.

• The QBs who sat as rookies outplayed the QBs who played as rookies by a larger amount in year three.

Basically this means that the year on the bench allowed a better QB to take the field in year one and showed some evidence that it helped the QBs develop into better players themselves down the road.

I plan to repeat this analysis in time. But as I get there, I can start with some rudimentary data. I made a list of every QB drafted in the first round from 2000 through 2016. I categorized them into three groups: Hall of Famer(?)s (4), Starters (17), and busts (24).

I also looked at how many games each of those players started as rookies. In some ways, these groups are easy to categorize: rookies that starters 0 games and rookies that started almost all the games--12 or more. In between, it's a little less clear; is it really that different to start 0 games than 1 or 2? Is starting 3 games the same sort of group as starting 11? For that, I made some overlapping categories. I have 0 starts; 0,1,2 starts; 3-11 starts; 12 or more starts; any starts at all (>0); and at least 3 starts (3-16).

I have posted this below, with comments to follow:

image.png (35.72 KiB) Viewed 686 times

Of note:

• Players with 0 Rookie starts account for 16% of the players, yet 50% of the HoF?ers and 19% of the Starters+HoFers, yet only 12% of the busts. This indicates that this subgroup performed better than the group as a whole.

• Players who participated in 3 through 11 starts fare the worst. If I had to make a theory behind this, these players are very likely to be guys that were not thought to be ready to start initially, but were pressed into action before they were ready. This likely has a negative impact on their development, as playing too early can a) destroy confidence, and b) cement bad habits into a player's style. It is much harder to unlearn bad habits than it is to learn from scratch (that's a kinesiology and psychology thing, not just a random assertion)

• Players with 1 or 2 starts look a lot more like the players with 3-11 starts than they do like the players who started 0 games. This indicates there may be something to that thing about bad habits or confidence knocks.

When I went to SERE school in the Army, the theory of the education was that you train under stress so that when stress comes, the learned response is retained and instinctive. If you teach yourself the wrong things under pressure, they become instinct. That's why it's best to make sure you have all the techniques down correctly before stepping into a high pressure situation. And that is why performing under stress when unprepared may create long-term damage to your performance.

As I said, this is the beginning of a series of analyses I plan to do. Later, I will dig into the numbers, investigate individual circumstances, and try to establish some control and/or counterfactuals by looking at the players who started in place of the drafted QBs. I don't KNOW that the current players will continue to show the trends I discovered last time I did this (and listed above). But so far, with the caveat of small sample sizes, there does appear to be some correlation between QB performance and starting 0 games as a rookie.

I'll be excited to continue to research and discuss it with you all.

[bud fox]

There will be a bias in your analysis.

This is going to be a hard argument to make with stats given the changing philosophy over time and there are so many variables.

[salmar80]

Quoting myself below, since I'm lazy.

While I appreciate all analysis, sample sizes get real small here, and there's great variety between a prospects' pro-readiness, level of team (OL, weapons) they started or would've started as rookies with, changing nature of game, running QBs having the option of rushing more if not ready as throwers, et cetera. The highest picks who usually start most as rookies tend to be on suckiest teams and are picking high for a reason.

I highly doubt one can definitely conclude either "start all QBs as rookies" nor "sit all QBs as rookies".

The military analogy is interesting. There's nothing that can offer combat experience except for combat, but enter the fray with poor training and the dice are loaded in an ill way... Vets not bothering to remember names of hastily-trained replacement guys before they happen to survive a battle or two in WW2 comes to mind...

While @bud fox is right that real game snaps are usually the best learning ground, but I still think it depends on the prospect. Some come out of college more pro-ready than others. Throw a QB who needs development to the wolves, and I doubt being overwhelmed, sucking and getting booed is a great learning experience.

This works for some, some could use a lesson or two before attempting, but both can become good swimmers:



Rodgers needed to overhaul some throwing mechanics + add strength and has said himself he benefited from time to work without having to start. His first preseasons and spot duty snaps in real games weren't exactly stellar. I remember the Pro-Favre and Fire TT -crowds being elated when AR suuuucked early on...

The scouts are in agreement Love has ample tools, but needs development. If he had been more pro-ready, he would've gone in top 10. I think he especially needs to work on reading zone coverages.

[lupedafiasco]

I don’t have a problem with one year sitting. Get your mechanics right. Learn the playbook. Get some work in with receivers to get down timing and tendencies.

Any more time than that and you’re wasting your rookie contract window IMO. Gotta get your 1st round picks in the field to play.

[British]

Thanks Yoho, good bit of research this. Looks like a pretty good sample size to me. By definition it's going to be quite small but considering it's almost all relevant QBs over the last few years it's plenty big enough to draw conclusions.

Already suggests that sitting provides a long term benefit. For a guy like Love with his upside/rawness combo it seems he's a prime candidate to site at least 1 probably 2 years.

[YoHoChecko]

Before I move on, anyone have issues with the classifications of the QBs? Jameis and Mariota are questionable. I'm not sure Cam Newton and Matt Ryan, both league MVP winners, are right to be grouped with "starters" but they're also a notch below the HoF(?)ers. I dunno.

[Packfntk]

Before I move on, anyone have issues with the classifications of the QBs? Jameis and Mariota are questionable. I'm not sure Cam Newton and Matt Ryan, both league MVP winners, are right to be grouped with "starters" but they're also a notch below the HoF(?)ers. I dunno.

Well, if you want to get nitpicky, Andrew Luck could have been in the next tier, had he played his career. But that is very debatable.

Ryan is an interesting one, especially if you are just looking at stats.
-10th all time in yardage
-11th all time in TD's
-9th all time in completions
-NFL MVP
-1 time All-Pro

His yardage, TD, and Completion rankings will all rise after this year. By the end of his career he could be top 5-7 in all categories.

[YoHoChecko]

Well, if you want to get nitpicky, Andrew Luck could have been in the next tier, had he played his career. But that is very debatable.

Ryan is an interesting one, especially if you are just looking at stats.
-10th all time in yardage
-11th all time in TD's
-9th all time in completions
-NFL MVP
-1 time All-Pro

His yardage, TD, and Completion rankings will all rise after this year. By the end of his career he could be top 5-7 in all categories.

Yeah, agreed on both counts; I'm a bit more concerned with classifying between busts and starters than between starters and home runs, since either of those count as a hit.

Grossman, Winston, Mariota I guess are my main questions there.

[Pckfn23]

Before I move on, anyone have issues with the classifications of the QBs? Jameis and Mariota are questionable. I'm not sure Cam Newton and Matt Ryan, both league MVP winners, are right to be grouped with "starters" but they're also a notch below the HoF(?)ers. I dunno.

Create a 4th group with Joe Flacco, Matt Ryan, Matt Stafford, Cam Newton, and Andrew Luck. Call it quality starters.

[Drj820]

History can tell alot. But history as stats sometimes miss monumental shifts in progressions in the short term. I do think traditionally sitting a year is the best move. But im not sure just stats of who succeeded and who didnt can tell a full story.

1. when a first round pick gets to sit a year, he probably came into a more stable environment that didnt need him day one. He was able to sit back and get nurtured, that obviously would help him succeed.

2. On the contrary if it was a horrible system, like browns under Hue with Baker Mayfield, sitting that first half of the season probably benefited him very little. They were just sitting him out of principle, similar to the kind of principle you are trying to show, and it just didnt help because he wasnt getting tutored, nurtured, and learning.

3. If you go into a stable environment where you can sit, most likely the org that created a stable environment has also built an entire stable team, and the pieces around you are probably more prepared to help you succeed.

4. Finally, the big shift is this...The NFL used to be totally different than the college game. There were college systems and NFL systems. The two systems look alot closer to each other now. QBs are running things they did in college and thats helping them be ready to transition faster than any time in history. Some positions still take a year to develop because the position has remained very different from in college, that would be most obviously TE and then even RB if you are expected to pass block.

So what demanded a year of learning in years past, might not demand it now. It would probably be helpful, but factors like system stability and coaching are more important.

Sitting a year wouldnt have saved Trubisky, and sitting a year would probably still have Darnold about where he is now. Its just not as necessary as it used to be.

[Packfntk]

Yeah, agreed on both counts; I'm a bit more concerned with classifying between busts and starters than between starters and home runs, since either of those count as a hit.

Grossman, Winston, Mariota I guess are my main questions there.

You mean between starters and busts? Winston is in the situation that he is in now, but we have to remember, he is (right now) the reigning passing yardage leader of 2019, (INT's as well). He is so young yet, and has a long career to fix his mistakes. I would have him at starter.

Gross man, total bust.

Mariota is a real difficult one. Where he was drafted? Certainly bust as of now. Do I think he can turn it around? He has a hurdle about a truck wide to jump. If I were ranking it as of right now, I would say bust. He started just over 4 years, never got to a big game, never led the league in a major passing stat, never been to the big one, benched, ec.

[Yoop]

there isn't anything a rookie can do that he wont do even better year 2, that includes ingraining all the bad habits and tech he's developed so far, thats why it's best to sit them for a year or maybe even two, it takes time to unlearn that stuff and develop better tech and habits needed to succeed here, if possible a coach will never play a rookie at any position prior to his readiness to play it, once you break a kids confidence then you just increased his learning curve.

soooooo, if you just spent valuable resources to aquire one why would you even take that chance, that is sooooo counter productive, sure one or two may do OK year one, but your gambling on those small % that do.

[Drj820]

there isn't anything a rookie can do that he wont do even better year 2, that includes ingraining all the bad habits and tech he's developed so far, thats why it's best to sit them for a year or maybe even two, it takes time to unlearn that stuff and develop better tech and habits needed to succeed here, if possible a coach will never play a rookie at any position prior to his readiness to play it, once you break a kids confidence then you just increased his learning curve.

soooooo, if you just spent valuable resources to aquire one why would you even take that chance, that is sooooo counter productive, sure one or two may do OK year one, but your gambling on those small % that do.

by this i assume you think #12 pick Gary was handled correctly and we shouldnt have drafted someone who would have had to help immediately last season?

I do not support high draft picks learning by "watching" their entire rookie year generally. Except for potentially at QB. But I do not support sitting for 2 years unless it is a very special Favre Rodgers type circumstance. Outside of a very rare case like that, if you cant earn your way on the field before year 3 after being a high draft pick, you are probably looking like a career backup. (josh Jackson?)

[YoHoChecko]

4. Finally, the big shift is this...The NFL used to be totally different than the college game. There were college systems and NFL systems. The two systems look alot closer to each other now. QBs are running things they did in college and thats helping them be ready to transition faster than any time in history. Some positions still take a year to develop because the position has remained very different from in college, that would be most obviously TE and then even RB if you are expected to pass block.

So what demanded a year of learning in years past, might not demand it now. It would probably be helpful, but factors like system stability and coaching are more important.

Sitting a year wouldnt have saved Trubisky, and sitting a year would probably still have Darnold about where he is now. Its just not as necessary as it used to be.

I plan to look into the changes over time...

but I wonder, what sort of evidence you might have to say things like that? Are rookie starters performing better now than they have in the past? I'll look into that, but anecdotally, I don't see it. The best time for rookie starters was when Flacco and Ryan both succeeded immediately and maintained that.

One thing I remember from my last look at this is that rookie starters improved more slowly and often regress, hence the second year being close to the same but the third year giving an edge. The list of draft busts includes a couple rookies of the year (Young and RGIII). Sitting in year one may (like I said, I have to look again now) lead to a bigger 2nd-3rd-year jump than the rookie starters experience. There are probably going to be tons of iterations of research, but hey, I've got all summer.

[YoHoChecko]

Before I move on, anyone have issues with the classifications of the QBs? Jameis and Mariota are questionable. I'm not sure Cam Newton and Matt Ryan, both league MVP winners, are right to be grouped with "starters" but they're also a notch below the HoF(?)ers. I dunno.

Create a 4th group with Joe Flacco, Matt Ryan, Matt Stafford, Cam Newton, and Andrew Luck. Call it quality starters.

I don't like this because now you're really asking me to parse hairs. Is Matt Stafford better than Carson Palmer? Is Joe Flacco a better pick than Mike Vick? Why are we giving Andrew Luck credit for a career he never quite had (6 years as a starter over 7 seasons, never quite hit a 100 passer rating in the current era). Isn't it too soon to determine what "quality" of starter Wentz or Goff will be (which is why I only went through 2016).

[Pckfn23]

Before I move on, anyone have issues with the classifications of the QBs? Jameis and Mariota are questionable. I'm not sure Cam Newton and Matt Ryan, both league MVP winners, are right to be grouped with "starters" but they're also a notch below the HoF(?)ers. I dunno.

Create a 4th group with Joe Flacco, Matt Ryan, Matt Stafford, Cam Newton, and Andrew Luck. Call it quality starters.

I don't like this because now you're really asking me to parse hairs. Is Matt Stafford better than Carson Palmer? Is Joe Flacco a better pick than Mike Vick? Why are we giving Andrew Luck credit for a career he never quite had (6 years as a starter over 7 seasons, never quite hit a 100 passer rating in the current era). Isn't it too soon to determine what "quality" of starter Wentz or Goff will be (which is why I only went through 2016).

Sure, add Vick and Palmer in there! I think it solves more problems than it creates. Like you said, there is a huge gap between Matt Ryan/Cam Newton and say Rex Grossman/Chad Pennington.

[YoHoChecko]

I was playing around with my filters to try to get a feel for whether/where there are any apparent groupings or cutoffs in terms of the # of games played.

I found, interestingly, that the BEST performing group is if you look at guys with either 0 or 16 starts. This group has 16 players in it, with only four busts, 10 starters, and 2 HoF(?)ers. In other words, not starting at all? Appears to correlate with good results. Starting every game? Appears to be correlated with good results. Everything in between? Appears, at the very least, not to correlate with good results.

This may indicate that teams who start their rookies from Day 1 and those rookies stay healthy and remain in the starting line-up have been properly-evaluated by their teams to be Pro-Ready. Where as guys who take over midway through or only pay partial seasons may be accurately gauged to be unready initially, but circumstances push them into the fire and either expose that inability or actually have a causal relationship to stunting the development.

This really plays into Salmar's point that finding an "all" solution is highly unlikely. The 16-game starters seem to be players that are very unlikely to bust. I'll have to look more into which guys started the season and didn't finish it (indicating poor play, but less common--but helps explain this specific grouping) and which guys were inserted into the lineup later and kept the job, which shouldn't impact this specific group

[Yoop]

4. Finally, the big shift is this...The NFL used to be totally different than the college game. There were college systems and NFL systems. The two systems look alot closer to each other now. QBs are running things they did in college and thats helping them be ready to transition faster than any time in history. Some positions still take a year to develop because the position has remained very different from in college, that would be most obviously TE and then even RB if you are expected to pass block.

So what demanded a year of learning in years past, might not demand it now. It would probably be helpful, but factors like system stability and coaching are more important.

Sitting a year wouldnt have saved Trubisky, and sitting a year would probably still have Darnold about where he is now. Its just not as necessary as it used to be.

I plan to look into the changes over time...

but I wonder, what sort of evidence you might have to say things like that? Are rookie starters performing better now than they have in the past? I'll look into that, but anecdotally, I don't see it. The best time for rookie starters was when Flacco and Ryan both succeeded immediately and maintained that.

One thing I remember from my last look at this is that rookie starters improved more slowly and often regress, hence the second year being close to the same but the third year giving an edge. The list of draft busts includes a couple rookies of the year (Young and RGIII). Sitting in year one may (like I said, I have to look again now) lead to a bigger 2nd-3rd-year jump than the rookie starters experience. There are probably going to be tons of iterations of research, but hey, I've got all summer.

well imo it's the table ware, first season, keep the plate small, don't give the youngster to much to eat, 2nd season give him the adult size plate, with all the thanksgiving trimmings, so naturally his learning curve continues, and we see them struggle with play comprehension at times, year 3 he has a better grasp of everything so naturally we notice improvement.

QB is a whole different level of learning, there mistakes are often much more costly as well, thats why they tend to have much more to learn and coaches don't start them a s rookies.

[YoHoChecko]

well imo it's the table ware, first season, keep the plate small, don't give the youngster to much to eat, 2nd season give him the adult size plate, with all the thanksgiving trimmings, so naturally his learning curve continues, and we see them struggle with play comprehension at times, year 3 he has a better grasp of everything so naturally we notice improvement.

QB is a whole different level of learning, there mistakes are often much more costly as well, thats why they tend to have much more to learn and coaches don't start them a s rookies.

That reminds me of another piece I did wayyy back in the day examining the development of Tom Brady, Rich Gannon, and Kurt Warner from little-regarded back-ups to MVP-caliber players. Consistency of system was big, and so was the air yards--first season was a lot of shorter passes and more of a "just run the system." Gradually increased what was asked of them in terms of longer, more NFL-level throws. Eventually the comfort, familiarity, and variety of tools developed breaks them out of being a "system" guy to someone good in their own right.

[Yoop]

there isn't anything a rookie can do that he wont do even better year 2, that includes ingraining all the bad habits and tech he's developed so far, thats why it's best to sit them for a year or maybe even two, it takes time to unlearn that stuff and develop better tech and habits needed to succeed here, if possible a coach will never play a rookie at any position prior to his readiness to play it, once you break a kids confidence then you just increased his learning curve.

soooooo, if you just spent valuable resources to aquire one why would you even take that chance, that is sooooo counter productive, sure one or two may do OK year one, but your gambling on those small % that do.

by this i assume you think #12 pick Gary was handled correctly and we shouldnt have drafted someone who would have had to help immediately last season?

I do not support high draft picks learning by "watching" their entire rookie year generally. Except for potentially at QB. But I do not support sitting for 2 years unless it is a very special Favre Rodgers type circumstance. Outside of a very rare case like that, if you cant earn your way on the field before year 3 after being a high draft pick, you are probably looking like a career backup. (josh Jackson?)

well actually my points pertain mostly to QB, and it doesn't matter where ya picked them as even top slotters are rarely ready to play QB at the NFL level, as to Gary or other top half of round 1 rookies, ya, you pick them there to start quickly, QB is vastly different then every other position, why anyone would think differently doesn't even make sense.

seriously, what is your point?