I just ran a project where I tried to learn as much about the scouting and training system and how they relate to both gems/busts and dev traits. Over the course of six seasons as a coordinator for the Georgia Bulldogs (force wins for the win), I scouted as many blue chip (4- and 5-star) quarterbacks as I had access to and followed their careers as closely as I could. Ultimately, I ended up with data on 177 quarterbacks over these six classes.
Here are some things I learned
1. Out of the 177 quarterbacks who were scouted (non-lockouts), 79 of them ended up transferring or declaring for the draft early. I checked on players' stats before the season, so I didn't have a way of differentiating between the two. Based on how early most of these quarterbacks transferred, though, I would say all but maybe five were transfers.
Also, two of these players just completely disappeared from the game. They weren't moved to a different position. I checked. They didn't end up on another roster. I checked. They just vanished sometime between national signing day and the start of the next season. Weird.
Of the 96 quarterbacks who did play their full careers at one school, 90 of them stayed for five years.
Why does any of this matter? Well, I figure it's more important to focus on a five-year development curve, due to the volume of five-year players compared to four-year players, when it comes to non-transfers. Why non-transfers? Well, the game doesn't give transfers any offseason training bonuses, which really ruins any point of bringing them in unless you want them to be a Day 1 starter.
2. Here's how the dev traits played out, by percentage, based on gem/normal/bust status:
Normal/Impact/Star/Elite
Gem: 0.0/37.5/40.0/22.5
Nor: 23.8/42.9/27.6/5.7
Bust: 83.3/13.3/3.3/0.0
22.9 percent of players scouted were gems. 17.1 percent were busts. The rest were normal. This rate held pretty steady between the 4-stars and the 5-stars, but I was only able to scout a total of 19 5-star prospects over six seasons. 5-stars are such a small sample size that I really urge you to focus on the data from the 4-stars.
3. There was not much of a variance between the players who transferred and the players I scouted overall. That is to say, the 90-man five-year non-transfers group is pretty representative of the 177-man total group of scouted quarterbacks.
Here's how the five-year players developed on the 4-star level, which is the only data I'd be comfortable using in this section because of the small 5-star sample.
4* elite: 74.1 overall star, 85.5 overall end (difference of 11.3)
4* impact: 73.9, 84.4 (10.5)
4* star: 74.5, 83.5 (9.0)
4* normal: 73.7, 80.6 (6.9)
Now we're getting somewhere. For the most part, star, impact and elite prospects end up in the same ballpark, within 2.3 points. That gulf between star and elite is roughly equal to the difference between star and normal, though.
Hint: AVOID NORMAL DEVS. Here's the cheat: None of the 40 gems I scouted had normal dev. 83 percent of the busts I scouted did, though. The rate for a normal dev player is 3.5 times higher for scouting busts (83 percent) than players who aren't gems or busts (24 percent). To avoid normal devs, avoid busts.
4. Development curves are kind of weird. Here's what the year-by-year overall increase looked like for a five-year curve (all player data was used up until their transfer) by dev trait:
Elite: 2.3/2.2/2.9/4.0
Star: 3.6/2.3/2.2/1.5
Impact: 2.9/2.6/2.2/2.8
Normal: 2.1/1.9/1.3/1.4
Mind you, this was all measured before the quarterbacks actually played that final season. These sample sizes aren't huge, but it seems like normal dev players hit a wall quickly after that freshman offseason boost. Impact dev players steadily grow throughout their careers. Star dev players quickly get to their ceiling but start to slow their progression in their final season. Elite dev players, though, are sort of like impact players, except that their stat ceiling is much higher. That means that they gain the most points at the end of their careers when they're more likely to be on the field earning XP to fill up that high ceiling. I thought that was interesting.
5. Here are some development standouts for you
Chad Glowinski, class of 2029, was scouted as a 4* gem. At USC, the star dev went from a 73 overall quarterback to a 94 overall over five years.
Pat Benn, class of 2025, is an honorable mention here, too. The 4* non-gem, non-bust was also a star dev and went from 73 overall to 90 in just four years. He declared early or otherwise he probably would have cracked around 96 overall his redshirt senior season.
The worst developmental curve goes to Isaac Cone, class of 2027. He was one of the few non-gem, non-busts who happened to be marked with a normal dev trait. He came in at 75 overall, gained one overall in the offseason between his true freshman and redshirt freshman season and simply never got an ounce better. As a redshirt senior, he was still just a 76.
I also wanted to mention that the lowest overall of any of these quarterbacks was actually a 5-star normal dev (non-gem, non-bust) by the name of Baraka Ihenacho. Clemson signed the 70 overall quarterback and he transferred out the next offseason.
tl;dr: You're pretty much good recruiting anyone but a bust, as they have an 83 percent chance of being a normal dev. You want to avoid normal devs because they increase their training points at about half the rate of elite, impact and star players.
Here are some things I learned
1. Out of the 177 quarterbacks who were scouted (non-lockouts), 79 of them ended up transferring or declaring for the draft early. I checked on players' stats before the season, so I didn't have a way of differentiating between the two. Based on how early most of these quarterbacks transferred, though, I would say all but maybe five were transfers.
Also, two of these players just completely disappeared from the game. They weren't moved to a different position. I checked. They didn't end up on another roster. I checked. They just vanished sometime between national signing day and the start of the next season. Weird.
Of the 96 quarterbacks who did play their full careers at one school, 90 of them stayed for five years.
Why does any of this matter? Well, I figure it's more important to focus on a five-year development curve, due to the volume of five-year players compared to four-year players, when it comes to non-transfers. Why non-transfers? Well, the game doesn't give transfers any offseason training bonuses, which really ruins any point of bringing them in unless you want them to be a Day 1 starter.
2. Here's how the dev traits played out, by percentage, based on gem/normal/bust status:
Normal/Impact/Star/Elite
Gem: 0.0/37.5/40.0/22.5
Nor: 23.8/42.9/27.6/5.7
Bust: 83.3/13.3/3.3/0.0
22.9 percent of players scouted were gems. 17.1 percent were busts. The rest were normal. This rate held pretty steady between the 4-stars and the 5-stars, but I was only able to scout a total of 19 5-star prospects over six seasons. 5-stars are such a small sample size that I really urge you to focus on the data from the 4-stars.
3. There was not much of a variance between the players who transferred and the players I scouted overall. That is to say, the 90-man five-year non-transfers group is pretty representative of the 177-man total group of scouted quarterbacks.
Here's how the five-year players developed on the 4-star level, which is the only data I'd be comfortable using in this section because of the small 5-star sample.
4* elite: 74.1 overall star, 85.5 overall end (difference of 11.3)
4* impact: 73.9, 84.4 (10.5)
4* star: 74.5, 83.5 (9.0)
4* normal: 73.7, 80.6 (6.9)
Now we're getting somewhere. For the most part, star, impact and elite prospects end up in the same ballpark, within 2.3 points. That gulf between star and elite is roughly equal to the difference between star and normal, though.
Hint: AVOID NORMAL DEVS. Here's the cheat: None of the 40 gems I scouted had normal dev. 83 percent of the busts I scouted did, though. The rate for a normal dev player is 3.5 times higher for scouting busts (83 percent) than players who aren't gems or busts (24 percent). To avoid normal devs, avoid busts.
4. Development curves are kind of weird. Here's what the year-by-year overall increase looked like for a five-year curve (all player data was used up until their transfer) by dev trait:
Elite: 2.3/2.2/2.9/4.0
Star: 3.6/2.3/2.2/1.5
Impact: 2.9/2.6/2.2/2.8
Normal: 2.1/1.9/1.3/1.4
Mind you, this was all measured before the quarterbacks actually played that final season. These sample sizes aren't huge, but it seems like normal dev players hit a wall quickly after that freshman offseason boost. Impact dev players steadily grow throughout their careers. Star dev players quickly get to their ceiling but start to slow their progression in their final season. Elite dev players, though, are sort of like impact players, except that their stat ceiling is much higher. That means that they gain the most points at the end of their careers when they're more likely to be on the field earning XP to fill up that high ceiling. I thought that was interesting.
5. Here are some development standouts for you
Chad Glowinski, class of 2029, was scouted as a 4* gem. At USC, the star dev went from a 73 overall quarterback to a 94 overall over five years.
Pat Benn, class of 2025, is an honorable mention here, too. The 4* non-gem, non-bust was also a star dev and went from 73 overall to 90 in just four years. He declared early or otherwise he probably would have cracked around 96 overall his redshirt senior season.
The worst developmental curve goes to Isaac Cone, class of 2027. He was one of the few non-gem, non-busts who happened to be marked with a normal dev trait. He came in at 75 overall, gained one overall in the offseason between his true freshman and redshirt freshman season and simply never got an ounce better. As a redshirt senior, he was still just a 76.
I also wanted to mention that the lowest overall of any of these quarterbacks was actually a 5-star normal dev (non-gem, non-bust) by the name of Baraka Ihenacho. Clemson signed the 70 overall quarterback and he transferred out the next offseason.
tl;dr: You're pretty much good recruiting anyone but a bust, as they have an 83 percent chance of being a normal dev. You want to avoid normal devs because they increase their training points at about half the rate of elite, impact and star players.
Comment