Recruiting and the NFL Draft (Random Stats)

[Consigliere]

CollegeFootballScoop ‏@CFScoop 27 Apr
2013 #NFLDraft by Home State (Final):

California: 28
Florida: 28
Texas: 24
Georgia: 19
S Carolina: 13
Louisiana: 11
Ohio: 11

CollegeFootballScoop ‏@CFScoop 27 Apr
2013 #NFLDraft Conference Draftees (Final):

SEC: 63
ACC: 31
Pac-12: 28
Big 12 22
Big Ten: 22

SEC East: 32
SEC West: 31

CollegeFootballScoop ‏@CFScoop 27 Apr
2013 #NFLDraft Recruiting Rankings (Final):

5-Star: 19
4-Star: 84
3-Star: 74
2-Star: 47
Not Rated: 30

 

[umhurricano]

Interesting data point about the stars. If you assume that there are 50 5*, 250 4*, 500 3*, and 750 2* (rough estimates from Rivals, and assumes early departures come out in the wash between years), then a player has the following probabilities of getting drafted based on ranking:

5* - 38% (probably higher since 50*s a year is generous)
4* - 33.6%
3* - 14.8%
2* - 6.3%

It would be interesting to track Golden's success against these numbers.

 

[FreePawn]

Interesting data point about the stars. If you assume that there are 50 5*, 250 4*, 500 3*, and 750 2* (rough estimates from Rivals, and assumes early departures come out in the wash between years), then a player has the following probabilities of getting drafted based on ranking:

5* - 38% (probably higher since 50*s a year is generous)
4* - 33.6%
3* - 14.8%
2* - 6.3%

It would be interesting to track Golden's success against these numbers.

Good point. Depends on who´s rankings they´re using, but I think rivals for instance rarely has more than 35 5*s, so you would be looking at a 54% chance.

 

[westcoastcanes]

Interesting data point about the stars. If you assume that there are 50 5*, 250 4*, 500 3*, and 750 2* (rough estimates from Rivals, and assumes early departures come out in the wash between years), then a player has the following probabilities of getting drafted based on ranking:

5* - 38% (probably higher since 50*s a year is generous)
4* - 33.6%
3* - 14.8%
2* - 6.3%

It would be interesting to track Golden's success against these numbers.

Good point. Depends on who´s rankings they´re using, but I think rivals for instance rarely has more than 35 5*s, so you would be looking at a 54% chance.

taking a few random years as sample from anything but scout who hands out 5 stars like they are free, closer averages are in the

35 5*
370 4*
1050 3*
and 2k+ 2* and lower

putting your percentages closer to:
54% for 5*
23% for 4*
7% for 3*
<4% for 2* and less

stats aren't perfect though, as kids could come from 3 different high school classes. a year high in 5 stars drafted one year probably points to a low year either before or after. none the less, it is pretty remarkable how good of a job some of the rating services are doing projecting 17 year old kids 4 years later.

 

[umhurricano]

Agreed that this information is merely directional. I think it points to the conclusion that recruiting is becoming increasingly commoditized (i.e., it is much harder to find prospects that everyone doesn't know about). If Golden is as strategic as he appears to be, he is identifying the areas where we can be most competitive and focusing on those.

 

[Port Knox]

 

[Fzac75]

Those percentages are bit skewed when you take into account that you should add up all the eligible "5 star" players in each draft, which would include all redshirt sophomores to seniors, as NFL rule is 3 years removed from high school, and the same throughout every "star ranking". In the end the simple fact is, though star system has its merits, it is as ridiculously unpredictable and inexact as the NFL draft is, in fact even more so. 254 total guys drafted, of that 103 were 4 or 5 star while 251 were not ranked or 2 or 3 star guys. Which is why, though I will follow the recruiting of so called highly ranked players, in the end it is the evaluation from the coaches to find and get the right guy to fit their system. And when all is said and done, if a good draft in the NFL is to find 3 solid, not great, but solid starters out of at least 7 picks, meaning less than 50%, what does that say about College recruiting where they have a much larger pool to sift through and don't even come close to spending or having as many resources as the NFL?

 

[umhurricano]

As I mentioned earlier, I would expect the "all eligible" five stars to be a wash since you have some leaving early each year, but some who have already left previously. The number would average out to be around the number of 5 stars in each high school class with a little attrition.

I completely agree that coach evaluation drives the system, but because of technology they can sift through a lot more players, and recruiting services are starting to get really good at talent identification (per westcoastcanes comment). My point is really that unpredictability now almost exclusively comes from the fact that these are still kids, and develop significantly physically and make a variety of life choices from now until the draft. Even a decade ago, there were many more kids flying under the radar because film and communication was less available. Therefore there were more "finds" that would probably be highly rated prospects today.