Recruiting Database: 1980-2022

If you're interested in playing around with recruiting data from as far back as 1980, there are currently spreadsheets available for download here:

EXCEL SPREADSHEET LINK

This fella's an NC State fan with one hell of a passion for barbeque, and an astounding collection of hikes on his page, too. Like, seriously, this guy hikes a lot and it's pretty awesome. (I've warned him in advance that a bunch of random Hokies might be poking around his site).

Long term, I have submitted this data to the redditor who runs this site as well. He's having a hell of a busy summer, but the plan is to upload all of this data for use with the optimized search and tool options on his site.

INTRODUCTION:

Long story short, I took a couple of years coinciding with the start of the pandemic to obtain, scan, edit and sort individual signee/recruiting data I found on eBay. I basically determined how the 247 Composite worked and then retroactively applied this format to create a database. In the following sections I'll describe what data there is in each spreadsheet, and I've added a few graphs to show how the data can be sorted and used. (I display the graphs showing Hokie curves where many are available- links are provided to the other graphs if you're interested in perusing them.)

PLAYER DATA

Fig 1.1: 114,062 Signee Scores plotted across 43 signing classes, highest to lowest

This spreadsheet has by far the most data, but there really aren't a ton of ways of graphically representing this. All of the Team Data comes from the individual signee data in this spreadsheet.

DATABASE STATISTICS

Fig 2.1: Signee threshold scores by season

Fig 2.2: Signee STAR score by season, per-centage

Fig 2.3: Database Teams & Graded signees by year

These graphs kind of show a blueprint of how the data looks before the internet era and after. Basically, before the internet, you'll typically only find evaluations on ~1/2 of the FBS signees. So with the older data, all of the unrated players end up having the same score, something I started calling the "Floor". I used this as the threshold for 2-3 star players. Starting in about 2006, most of the FBS signees started having some kind of individual score. From this point onward, there are far fewer "2-Star" signees in the database.

There are still signees without evaluations today; in fact COVID seems to have temporarily bumped up the number of players that weren't evaluated well enough to have a Composite score. These players simply don't have scores associated with them- and they don't affect a team's score at all. This most notably affects JMU's introductory FBS class (2022). They have 4 unranked signees. Even though they have 8 rated transfers, Transfers don't factor into team recruiting scores. So JMU's 2022 signing class effectively has no class score.

Fig 2.4: Total FBS Teams and FBS Signees by season

Fig 2.5: Average, Min & Max signees per class by season

These graphs show a bit of how college football has looked over the years, and how this has affected this database. I list MOST of the FBS programs, but essentially I've omitted all of the football programs that don't use scholarship rules and limits. This means that the Ivy League (1980-81) and the Service Academies (all seasons) don't appear at all in my database. The service academy's signing classes vary wildly, even into the internet era they didn't publicly release their signing classes. They aren't really competitive for Blue Chip recruits- basically they give full rides to a whole lot of well-rounded individuals- many of them are very good athletes, so their classes are full of respectable ** "signees" and some seasons they'll "sign" a couple of *** players, too.

In 1981, the NCAA had a purge of schools that didn't meet the requirements to compete in Division 1A. The Southern, Southland, Ivy League and (eventually) Missouri Valley conferences dropped down to FCS starting in 1982. In 1992 the scholarship limit for FBS (Div 1A) dropped from 95 to 85 per school. 1989 was the start of this transition, so basically an average signing class went from 25 signees to 22.5 over this period.

We're currently in a new adjustment period with the Transfer Portal (coupled with the Covid eligiblity queue from the 2020 season), so by 2025 or so a new average signing class size should be apparent. (Probably end up somewhere just shy of 20).

Fig 2.6: OVERALL signees by top producing States (VA is on the second graph)

Fig 2.7: BLUE CHIP Signees by top producing States (VA is on the first graph)

While it might seem like OVERALL signees by state is a useful metric to use to determine which states are producing the most talent, the truth is that it really isn't. When Richmond and William & Mary dropped down to the FCS in 1984, the number of signees from the state of Virginia plummeted. This was not a valid indictment of Virginia High School Talent. The recruitment of 2 and 3-Star players is simply based on geographic convenience. If North Dakota State were to jump to the FBS, there would suddenly be a half dozen signees or so from that state every year (which is about a half dozen more on average than there are now).

Blue Chip production is the measure to use to determine how productive the states are over time. Those players are getting signed by somebody somewhere, with little regard for the geographical appointment of available FBS schools across the nation.

Fig 2.8: P5/G5 split, Teams and signees

In 1980, there were similar numbers of Auto-Qualifying programs (P5 and equivalent Independents) and Non-AQ (G5) programs. Following the purge of 1982, the number of G5 programs dropped substantially, and over the course of the next four decades has essentially grown back again.

Over this time, the number of P5 programs has been pretty consistent, hanging out around 65.

TEAM DATA:

Average Team Scores #1-10
Average Team Scores #11-20
Average Team Score #21-30
Average Team Score #31-40

Fig 3.1: Average Team Scores from #41-50 (VT is ranked #44 from 1980-2022)

Average Team Score #51-60
Average Team Score #61-70
Average Team Score #71-80

For the Team Data Spreadsheet, I calculate team scores 3 different ways. First, I basically made an "Estimator" of the 247 Composite Team score. I don't like this method, so I also include the Raw Average score of the signees, and an "Adjusted" average score, which is the method I prefer (and shows up on these graphs).

In truth, I'm still being deferential to 247 Sports, as this "Adjusted Average" team scoring method actually shows up in their Team Scoring explanation:
Fig 3.2: 247 Sports Team Score Formula

First off, 247 Sports doesn't even use this method anymore; they changed their methodology a season after they introduced the Team Score Calculator. Basically, this old method only really counted the top third to half of your class. But pay attention to that "C". (They stopped factoring in that "C" with their new method, BTW.)

Fig 3.3: A 3D image that 247 Sports provides to show their old methodology
Fig 3.4: A 2D representation of how that old method actually affects the tabulation of a team score

So they opted for a new methodology that would change the team score no matter where a new signee was added to a team (obviously a higher scoring recruit moves the needle much more).

Fig 3.5: A 2D approximation of the "New" 247 Sports team scoring

So back to my favorite- it's similar to taking the raw average. I just figure out what size the average signing class was each season (lately that had been 22 or 23 players a class), and simply taking the average of the top 22 or 23 players for a signing class.

That's the "C" from above.

There's a small advantage for larger classes & no penalty with small classes. Basically, most every program is pretty good at keeping 85 scholarship players on a roster. So class size from year to year is, for the most part, irrelevant. Teams that consistently have large classes are simply more actively managing, or churning their roster. So they get a little bump for this, but for the most part average recruit rank is still the best method of determining how programs are doing from year to year.

Fig 3.6: ACC Recruiting Trends

Fig 3.7: Big East/AAC Recruiting Trends

Fig 3.8: Middle & Minor Independent Recruiting Trends (VT was a MIDDLE independent)

These graphs show the team scores across the various conferences. Typically, where conferences change I showed the current conference in BOLD and the other conference affiliations in some manner of dotted or slashed lines.

When I was starting out this project, I tried to sort all of the early FBS independents into P5/G5 categories. I found that Independence was more of a spectrum, and I used recent schedules and opponents to determine "peers" for the Minor (G5), Middle and Major (P5) programs. Virginia Tech was a Middle Independent using this method; we were pretty split between playing other G5 and P5 teams (but less than half P5 opponents, or peers). Basically, for these teams, any conference shakeup could go either way. And in truth, it did.

When I developed team scores, I ended up grouping the Middle independents in with the G5 schools (they basically recruit in the middle, though- however with no conference affiliation in the TV era, they were always at a disadvantage.

In 1997, the last Middle independent (ECU) joined the CUSA, so that category went away. After this, you were either Notre Dame or you were tiny. Until BYU went independent, that is. I didn't reintroduce this category into the database, but I dissected this for a Reddit thread about a year ago and discovered that BYU would have essentially been a "Middle" independent using this same peer/schedule methodology.

Big 8/Big XII
B1G Ten
CUSA
MWC & MAC
Pac Ten/Pac 12
PCAA, Big West, Sun Belt
SEC
SWC & MAJOR Independent
WAC

EDIT: PS, I made this graph to show how the conference recruiting has changed historically, with a graph that shows how I would project the new conference alignments to change in the future:

Fig 4.1: Historical and Projected Conference Recruiting Trends

Forums: 
DISCLAIMER: Forum topics may not have been written or edited by The Key Play staff.

Comments

Take your up vote, I will finish reading this novel in a few days, looks like great work

I was proud of how short I kept the introduction... but then I undid all that with the graphs!

I guess my question to you after doing all this are:

- What was your biggest takeaways?
- Now that you have this, where do you go from here?

A few takeaways:

- Coaches have very little to do with a team's attractiveness to recruits. That's not only shocking, that's incredibly counter-intuitive. The biggest noticeable differences in a programs recruiting scores happens when a move between G5/P5 happens.

- there is less of a difference in how predictive (to the NFL) recruiting was from the 80's and 90's to the 00's, than the difference between the 00's and onward.

- there's a significant drop in production across the rust belt over 40+ years. Both Florida and Georgia have almost doubled their output over this time. (using blue chip signee numbers).

- I finished this project at the absolute weirdest time. I projected the composite scoring system backwards to see how past trends worked their way to the present. And now that I've done that I have no idea what "recruiting" will look like going forward. (Between NIL, Covid and the Transfer Portal, we're right in the middle of the first "major" sea change to cfb recruiting).

Going forward I want to spend as little time in front of a computer as I can. I bought a bike and I want to start training for some long runs.

Fair! Thanks for the consistent high-effort content you share with us!!!!!!!!!

"Why gobble gobble chumps asks such good questions, I will never know." - TheFifthFuller

First of all, amazing analysis, thank you for sharing. Have some reactions/questions/thoughts

Coaches have very little to do with a team's attractiveness to recruits. That's not only shocking, that's incredibly counter-intuitive. The biggest noticeable differences in a programs recruiting scores happens when a move between G5/P5 happens.

This definitely goes counter to the narrative around the cult-of-personality coaching figures, but I think anyone who follows the recruiting knows this - there's a reason why Tennessee, Texas, etc has always recruited well despite being a dumpster fire for at least a decade.

Playing Devil's Advocate, I imagine that a good coach can turn recruiting around, but good coaches typically don't stick around - Is this reflected in the data anywhere? For example, let's look at Mack Brown at UNC: If recruiting improved when Mack Brown came to UNC, then he left and recruiting regressed towards the mean, then he returned and recruiting improved again, that's gotta say something about that coach, right?

I'm curious if you're able to isolate any other data/landscape changes? For Example:

  • Oregon's recruiting performance relative to Nike cash infusions
  • Mizzou/TAMU leaving the B12 going to the SEC
  • Colorado leaving the SEC for the P12

Finally, have you done any analysis on what impacts a recruiting class? I've always hypothesized that the three major indicators of recruiting success are (1) athletic dept funding, (2) quality, quantity, and distance to recruits (average recruiting rating per mile from the university?) and (3) quality & quantity of nearby rivals (ie; who is recruiting the same places you are)?

Curious if there's the data that confirms my suspision.

there is less of a difference in how predictive (to the NFL) recruiting was from the 80's and 90's to the 00's, than the difference between the 00's and onward.

So essentially recruiting has gotten more accurate in the 2000s then at any other point in time?

Twitter me

If recruiting improved when Mack Brown came to UNC, then he left and recruiting regressed towards the mean, then he returned and recruiting improved again, that's gotta say something about that coach, right?

Typically with coaches who have seen recruiting scores jump under their tenure, there's a programmatic explanation behind it. Bowden at FSU (the jump to the ACC is visible), Beamer (large but shake jump up in the Big East, smaller jump up to the ACC, and that's more or less where we are now).

Mack Brown has seen 3 dramatic rises, and frankly I'd even like to know more about why. Recruiting actually dropped when he was at Tulane (they failed to catch on in the SEC, remaining independent until they eventually jumped to CUSA. This likely just coincided with the end of his tenure there). Then he had 3 tenures where recruiting scores rose (UNC/UT/UNC).

While something along the lines of Jordan money could explain the first and Texas expanding recruitment beyond the state once the SWC folded could possibly explain the first two, the latest jump at UNC simply being a return to glory from the threat of severe sanctions (following the latest hire) doesn't seem like the whole story there, either.

I will say if Mack Brown has the power to defy conventional logic and lure kids to programs like nobody else, I'd emphasize the "like nobody else" part.

It's possible he's some kind of recruiting genius, I still think it's far more probable that bigger things were happening while he was at those schools. (Mainly because I don't see that happening with anyone else's career).

I'm curious if you're able to isolate any other data/landscape changes? For Example:

Oregon's recruiting performance relative to Nike cash infusions
Mizzou/TAMU leaving the B12 going to the SEC
Colorado leaving the SEC for the P12

Oregon used to recruit closer to Oregon State and Wazzu. Their recruiting rose trailed the Phil Knight infusion by about 5 years in the 90s.

Mizzou, USCe and Texas A&M were all helped by their jumps to the SEC (Arkansas really didn't change much at all).

Big Eight/XII teams seem to be hurt by moving conferences (neither Colorado nor Nebraska saw any recruiting advantage by leaving).

Finally, have you done any analysis on what impacts a recruiting class?

Not really, I feel like I'd end up using the wrong financial numbers or interpreting them wrong. I do think the SWC was overly obsessed with fighting each other off recruiting Texas, and that was likely holding back the bigger schools there.

always wondered if average recruit rating was less important than median, since that also brings quantity into play in a slightly different way than average does

"Why gobble gobble chumps asks such good questions, I will never know." - TheFifthFuller

I think median is better at disregarding the negative effect of signing specialists (you could sign the nation's best K, P & LS and the effect on your team score is having 3 guys ranked ~85-86 bringing down your average. That's dumb, but that's how they do it).

To fix that problem long term, you probably just shouldn't factor specialists into your team average (or median).

There's nothing wrong with using median; my priorities were a) modeling whatever the composite does, and then b) keeping things simple and consistent. An "adjusted" average actually kind of does both, so that's what I went with.

You left of the link to the ACC in your conference graphs at the end! Freudian slip for your desires of the conference's future?

I am not sure what to do with my hands now

Edit: I posted the Big East, ACC and Independent graphs with VT a little above the links.

I saw those, was looking for the link to see it a bit bigger. Thanks.

I am not sure what to do with my hands now

ACC
Big East
Independent

On desktop, the best way to see close up is to right click and open in a new tab (you might have to do something similar to zoom in on Imgur as well).

This is an amazing data set. I'd love to see the results of a multivariate regression on it. You could look at school, head coach, conference, national average recruit score, and some location metric to attempt to numerically tease out how each of those things actually contributes. Then you really could find outlier coaches who have a real effect, and how much of a bump being in the SEC gives vs. being in TX or FL.