Sobering article on recruiting: VT and 'blue chip ratio'

I thought about just putting this in the current recruiting thread, but I decided it was worthy of a separate more general discussion.

Great article about VT recruiting and their 'blue chip ratio' metric.

VT Blue Chip Ratio

The basic premise is that to be a contender, you have to have the blue chip talent. Specifically, any team that contends for a National Championship has to have a roster with >50% blue chippers.

(Spoiler, VT doesn't.)

Its a great read; and sobering.

A couple of my take aways: it is amazing that VT has had the success it has enjoyed in the past, using this metric as a comparison.
Also, we got some major work to do, but in Pry we trust.

One quote from the article:

This is probably Tech's least talented roster, at least based on Blue Chip Ratio, since before the Michael Vick era.

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


A couple of my take aways: it is amazing that VT has had the success it has enjoyed in the past, using this metric as a comparison.

The big difference between now and pre-2005 (or even pre-2011) is that recruiting services weren't as accurate as they are today. Hudl was founded in 2006, the iphone didn't have video recording until 2009, so there was a lot of really talented players (eg; Kam Chancellor) who couldn't promote themselves; their only option was to rely on their coach's network (which was often limited to schools inside the region).

That is no longer the case - high school coaches are no longer the middlemen - schools/coaches can scout players and contact players without ever meeting them before.

All this to say - I think some of those early-to-mid-2000s did have a 'modern day equivilent' bluechip ratio that was greater than original marker.

Twitter me

I dont think we have been close at all. You need 10 4*+ recruits a year. We've never came close. Even before hudl and such there has been the ESPN 300 which is older than Hudl, the ESPN 150 is from the late 90s.

We rarely get anywhere close to that. Even if you go back to 247 and boost the rankings to account most of our players were always really low ranked. We've never had a top 15 class, there is no way to get close to that ratio with out top 15 classes. I don't think we have been top 20 multiple years in a row.

That's not my argument - I'm suggesting that we had a lot of 2/3-stars over the years who were not properly evaluated by anyone other than our staff. Eg; Kam Chancellor would be a high 4-star today because other teams would be able to use technology to properly Scout him. If you account for all the kam chancellors who 'should've been' 4-stars, I think we might've had a blue hip ratio in the high 30% to mid 40%s.

That said, I'm speaking in generalities. It would be interesting to go back and rerank players.

Twitter me

Sorry I missed you point. I do have a question in this hypothetical situation you would have game tape of them? Cause Kam was a QB so its hard to believe he'd be a 4 star DB based of his film.

Foster had a quote about this recently, where he said in a modern offense, Kam would playing QB in a spread offense.

But the point still stands - there are 4-star athletes. There are players like Logan Thomas who are recruited as a position they've never played before. It's not uncommon for high schools to just put the best athlete at qb. It's not ridiculous to believe Kam would've been a 4-start today πŸ€·πŸ»β€β™‚οΈ

Twitter me

I an going to try and work backwards ....
From 2014 - 2021 there were
167 NFL draft picks that were 5*
570 NFL draft picks that were 4*

At ~30 5* each year from 2010-2017 (the years feeding the draft, sort of) that gives us 167/240 5 * players get drafted or about 21 a year.

Over that time frame 247 had around 310 4* recruits each year. 570/2480 or about 71 4 * a year.

So 92 Blue chip players were drafted out of 340 each year or 27%.

A roster is 85 players 43 makes 50% Blue-chip which at 27% means 12 players from the roster would be drafted.

So yup, you're probably right, We had more than that from the 1999 team. We had 9 drafter in 2006 alone and 8 in 2008. So we could have meet that.

The only issue that this over looks is the number of 3* in the draft 100+ a year. 3* and below still fill out most of the lower rounds which tends to be where our players were drafted. 5* recruits are typically the first 2 rounds, 4* recruits are more heavily the first 3, and shrink each round after, but have a good number in the layer rounds. 3* recruits are ever round fairly well spread across. Below there are 6th and 7th typically.

So it is possible that we never had the blue chip ratio and just coached them up. So this wasn't as good a method as I hoped, but you definitely have some backing to us having the ratio a long time ago. Either way, having 20 draft picks in a 3 year period seems to be something we should get back too.

Totally agree. Really interesting thought process.

Twitter me

I mean...why do you think our players were not being properly evaluated? Do you think the other schools who did qualify for this were also not being properly evaluated?

I disagree with your assumption


There are now more 4* than there were 15 years ago. People see more tape of players which should be more accurate rankings. Also if it was a,small school less colleges would see the player.

I'm saying that VT was properly evaluating these players (the 2-star Kam Chancellors, the unranked Vincent Fullers, etc), but recruiting sites (eg; 247) and other programs did not have the opportunity to correctly scout these players because (pre-hudl, pre-iphone, pre-social network) it was just really difficult to watch someone play enough snaps to properly rate them.

Now, recruits can create their own footage, upload it to hudl, and share it with coaches via social media.

Before, our coaches got tipped off about these players. For example, VT was Kam's only offer from a 'Power conference' (I know we weren't using this term then, but you know what I mean). Does this still happen today? Yes, but at a significantly lesser rate than it did 10-20 years ago.

We used to be able to build a competitive team with mostly 3-star players, year after year. You can't do that anymore. It's not because coaches aren't good enough to coach up 3-stars; it's because the players who used to be 2 and 3-stars but played like 4-stars are now rated as 4-stars by 247.

Twitter me

the players who used to be 2 and 3-stars but played like 4-stars are now rated as 4-stars by 247

And recruited as 4-stars by more programs accordingly

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

Here are two graphs of decades/annual data of the improvements in locating Blue Chip recruit talent vs being drafted or playing in the NFL. There has been improvement in evaluation accuracy over the years.

Nothing moves the needle for me on why Kam Chancellor would have had much of a different evaluation in 2006 vs now. Nobody hid him out and saved him for VT; he wasn't an unknown quantity. Other teams knew who he was and frankly weren't interested in him.

He was evaluated as a Quarterback, and his composite score is an 83.33 ***. His offer list includes both JMU and Kent State, and interest, but no offer from UVA.

When I re-ranked him based on 247's current grading curve, his score changed to 85.05- (his grade would be a little higher but he'd actually drop behind over 100 players from Rivals that 247 doesn't even have a composite ranking for in their database for 2006).

So he was tied for ~900th highest recruit out of high school (as a *** QB), and he improved in college to become the 133rd selection in the 5th round (as a DB). Sounds about right.

FYI; Vincent Fuller pretty close to a 4**** in 2000 (87.89) (He was graded solidly by all 3 post-season services I have data for that season - Alliance Sports, Tom Lemming and Prep Star).

I think we were recruiting better in the aughts as well, but I don't think it's from superior recruiting- it's probably just from opportunity. Here's a look at Virginia's output of FBS signees when you consider both High School and the golden era of the Virginia post-graduate Prep programs (since Richmond and William & Mary left the FBS after 83).

Also, that John Iezzi Sons of Saturday podcast was pretty eye-opening on the state of our recruiting efforts a couple years back.

[EDIT: If you right click "Open Image in New Tab" on desktop, these pictures show up a lot clearer. Sorry about that).

Wow, great post, thanks for all this info!

I keep using Kam as an example because (1) Bud Foster did, (2) Kam might've been more desirable as a QB in modern spread offenses, and (3) he left early for the draft, so I assume there's some level of talent vs 'coach em up'.

Twitter me

Kam is an outstanding athlete.
He obviously was taught how to take proper advantage of his athletic talents.
He was obviously taught position technique.

I do get the point that Frank/Bud et al were very able in identifying those talents that were not properly identified by other teams.

This is going to be great for the ACC.

Kam's still a great example of a lowly regarded prospect who turned into an NFL great. It's quite possible that Scout had him unranked or, in fact, a 2-star. I could see his potential being more recognizable in a newer-style offense.

I also think the world of Bud Foster- I truly believe he was great at all of it- seeing potential, coaching up players, a genius at scheming.

I see where you're coming from now on the "superior scouting" take. In 2006, we had about double the pool of in-state recruits, and in order to offer a guy like Kam (mid *** with little interest from other schools), we were definitely doing something well.

Unfortunately, that really went to the coaching staff's heads. Per Iezzi, we ended up being one of the last FBS programs to hire recruiting staff and pay for the digital services. Being really good at "sniffing" out athletes in the aughts put us way behind the curve when all of the other schools starting using their eyes to do the same thing even more effectively in the 2010s.

I guess I was just being salty there.

I'm just about finished up with the database- just sent over the spreadsheets to the guy who runs this site ( Probably won't be up any time soon- he just had a pretty big life event (a positive one), so he's focused on that right now. There are 117,245 player profiles, 4942 team scores, and 520 conference averages starting back at 1980. I probably seem like a prick jumping in here with this "proprietary" data from before the internet era. Hopefully it will be public soon.

Kam is also a great example of how VT benefited greatly from the lack of recruiting services available at the time, and just how great Beamer/Foster's talent identification and development was.

These advantages are long behind us.

I think we agree here - clearly the Beamer staff found a market inefficiency (VA being under-scouted/under-recruited) and capitalized on it. They failed to pivot when that market inefficiency disappeared.

double the pool of in-state recruits

I've heard/read some takes that talent (particularly in the 757) is down due to a population decrease caused by the 2008 financial crisis (less jobs in the 757 = people move = lower population = smaller pool to for talented football players to emerge from), although I haven't seen anyone prove correlation here. As someone who follows recruiting from a 30,000 ft view, it seems like we're targeting NoVA/DC way more than the 757, and I always assumed this was why.

Twitter me

757 is not as down as people would have you believe in terms of the volume of blue chip recruits being produced, just that there are more composite blue chip recruits elsewhere in the state (e.g. NOVA) and 757 is not as consistent year over year. Possibly there were a couple down years in the 757 and someone made that connection and the narrative stuck. And then it got extended to the entire commonwealth

I made a (very rudimentary) attempt to look at quantity of blue chippers in 757 as well as percentage of in-state blue chippers from the 757 somewhere else in a thread a few months back. I'll see if I can dig it out


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

"Round up the usual suspects!"

After reading the Blue Chip Ratio article, I was afraid to click on that link, so let me help some of you out:

Virginia Tech is NOT on that list.

If you're not sure if my comment warrants a "/s", it probably does.

**Steps up on Soapbox**

This is why the playoff has been such a net negative for the fan experience of the sport. There's only 10-15 programs that have the talent to win a title in a given year, and only half of those have a manageable schedule. The sport used to be about surviving the regular season and enjoying the weirdness. Now it's all about just those 8-10 teams and there's little-to-no weirdness. Expanded the playoff won't help. I think promotion/relegation is the only thing that can bring meaning back to the regular season, but I'm not sure that would/could happen in the US.

That said, I'll still find ways to enjoy the sport

**Steps off Soapbox**

Twitter me

And absolutely nothing that the NCAA is doing right now is going to help this in either the short term or the long term, and in fact, appear to only be greasing the wheels of allowing it to get even worse than it already is.

Outside of VT football, I barely watch college football anymore. And with NIL rapidly turning rev sports into an in the open pay for play 'survival of the wealthiest' nightmare, I'm really not sure how much interest I'm going to have for even VT football going forward. I'll still watch on a weekly basis, but I can already tell you I'm far, far less vested and excited in the Pry era than I was when Fuente was hired. I can't shake the feeling that it really doesn't matter what the new coaches and all the positive fundraising ventures we've been doing because the gap between everyone else and us is growing exponentially year over year. When I see the B1G asking for a $1 billion a year tv deal and we're stuck on $240m a year until 2036.... You can't compete with that, you just can't, so why bother getting vested for something that you have no hope of attaining. The ACC is a mid-major conference, and nobody is willing to admit it, and if you are a mid-major, the sport has already left you behind.

Money and greed have ruined college football. I'd rather follow professional sports where they at least try to promote parity where there is some semblance of hope for every team through the draft or free agency where wealthy teams can't just consolidate talent because of a salary cap or luxury tax. (and no, I do not watch the NBA, I think outside of college sports, ESPN has ruined that league the most, and they're probably set up for a hell of a fall in popularity within the next 10 years or so when fans of the teams that players don't force their way into just turn their back on the sport for good, which is very much already happening)

This is my school
This is home

We'll see.

This is probably the 4th or 5th time in college football history that the sport has undergone large scale changes. Every time, there were people saying the sport was ruined, and it wasn't.

Things are changing, and eventually the dust will settle. The sport may be very different from what we know and love, but I'm not convinced that's a bad thing.

Twitter me

Promotion/Relegation would be the best thing to return to the weirdness that made many of us fall in love with college football.

I'm imagining a scenario where you have two columns of 15 teams; each column picking a team by snake draft. You play a 14 game schedule; top 2 teams advance to the championship, the remaining teams of the tiers play each other in bowls. The bottom 2 get relegated. Snake draft is entirely random draw; similar to the FIFA World Cup draw.

So Championship level is 30 teams in 2 tiers. relegated teams cannot return for one year.
Tier A is another 30 teams in 2 tiers. Top 2 from each tier advance and are safe from relegation for 2 years. Top 4 play in bowl games
Tier B is the next 30 teams, same promotion/relegation scenarios, Top 4 play in bowl games
Tier C is the next 30 teams, same promotion/relegation scenario, Top 4 play in bowl games
Tier D is the bottom 11+ teams, they play each other and can schedule non-FBS teams, Top 2 play in a bowl game

21 post season games. Every team gets a 14 game season. Any team can go from Tier D to Championship Tier in a player's career.

Never Forget #1 Overall Seed UVA 54, #64 UMBC 74

I disagree that this is "sobering" because:

  1. I'm pretty sure most VT fans were already aware of this. It's been discussed quite a few times.
  2. If anything this causes me to drink more, not less

How I Learned to Stop Worrying and Love the Jet Sweep

To everyone pissing on college football because of how it is evolving, I got news for ya it was already there.

You think we had a real chance to win the whole thing 10 years ago? No we sure didn't. Sure, the gap is wider now, but it was always just as much of a long shot.

I just sit on my couch and b*tch. - HokieChemE2016

Your point?

Everyone has their breaking point. Not sure going hipster 'I realized it was broken before everyone else' accomplishes what you think.

This is my school
This is home

Did you enjoy watching college football 5 years ago? Or 10 years ago? What and 30 years ago? If so then you'd probably still enjoy watching it now because the difference are minor, players are getting paid in the open and they can transfer with out going down a level or losing a year of playing time.

I watched as many games as I could 10 years ago. 5 years ago I couldn't be bothered to watch the playoffs. Last year I only watched VT games.

This is my school
This is home

10 years ago, ignorance was bliss.

VT was nationally relevant, recruiting wasn't covered by mainstream media, no one knew what the bluechip ratio was, and in 14 years of the BCS, we saw 10 different teams win a title.

The playing field was no more/less level then than it is now, but no one new how stacked the deck was against 90% of the sport.

Now we know. And now, the media coverage has changed in a way that reminds us how irrelevant we are every week.

Twitter me

It's common for peoples interests to change as they age. New work, new hobbies, family, whatever. It would be more surprising if things hadn't changed than That they did

10 years ago the top 10 looked a hell of a lot like it did last year, so I'm not sure you can say that things have actually changed that much

I just sit on my couch and b*tch. - HokieChemE2016

10 years ago the top 10 looked a hell of a lot like it did last year

And maybe that's a big part of the problem. A problem that isn't getting any better because new developments within the NCAA are doing a very good job of eliminating what little parity might have remained.

This is my school
This is home

Only half the top ten then is a top 15 today. Boise, Stanford, and Tech itself sitting at 13 so there have been minor shifts the last ten years.

Wet stuff on the red stuff.

Join us in the Key Players Club

Same. But I also got a lot more into tech football after college as a way to hold on to some of the community I missed by being out in the working world. Changing jobs after 2.5 years and moving towns hit the reset button on my social life so it made fandom that much more important. Not saying the landscape isn't different but our particular timeline does lend itself to some biases.

In the last 6 years, I've built community and the need to be obsessively involved in hokie fandom has waned. I've got kids, a house, dog, job. All sorts of things that pull on my time. I'm not sure it's fair to look at it in a vacuum and say that CFB simply isn't good enough to hold my attention anymore. It's more that I don't need it like I once did and am more likely to say "screw it" when it involves aspects that annoy me.

Warning: this post occasionally contains strong language (which may be unsuitable for children), unusual humor (which may be unsuitable for adults), and advanced mathematics (which may be unsuitable for liberal-arts majors)..

Also could just be some of priorities changing as time passes? I know I have other things I am (forced to be?) putting my focus on instead of football on Saturday afternoons.

Clemson is the model. No fucking excuses.. Go out and sign fucking Sammy Watkins, DeShawn Watson, Mike Williams.. Win a bunch of fucking games in the medicore ACC. Not rocket science. Get some fucking studs... win...

Many moving parts of course, but if I am a recruit and I watched the weak small team get their ass kicked by Maryland... the 12th best program in the B1G, I'm not excited. When I see QBs leaving and not developing, and a RB room full of "I hope THIS guy can run the ball" I'm not excited. You can talk about facilities, budget, relationships all you want, but the product on the field has been shitty. Weak, small, no identity, poor inconsistent offense, mediocre defense, few top 3 round NFL picks, etc. We must play better. We must beat the BC's of the world and a shitty Kentucky team with no passing game in a bowl. Play better. first and foremost.

So I'm not sure I entirely agree with the general premise. I think to have continued success this is probably more accurate, but I think in any given season, there's a lot more possibilities. One team in a conference that might not typically be the normal champ may have retained and developed more talent and some of the other top 2-3 schools may have a down year because a number of their players transferred, graduated, or got drafted. So that school rises to the top. I think this can happen in almost any conference (see Clemson not winning last year and Wake and Pitt being at the top).

I'd further argue that if the premise is true, us scoring 35 points on Clemson, the eventual national champions, and then Ohio State getting shut out by the same team SHOULD. NOT. HAVE. HAPPENED. Hell, I'd argue if the premise is true, those 2 things both happening probably shouldn't have even been possible.

Yes, yes, I know we had a lot of talent on that team, but both teams are in the era of talent having been properly scouted and Ohio State had a lot more recruited talent on their roster (not all highly rated recruits are as successful as their ratings indicate they will be). Don't forget that Ohio State had won their conference and were 11-1 going into the game with Clemson. Even given that we may have had a lot of talent on our team, here are other teams that year that did not get shut out by Clemson: South Carolina (6-7 that year), Wake Forest (7-6), Pitt (scored 42 and was their only loss that year 8-5), FSU (10-3), NC State (7-6), BC (7-6), Louisville (9-4), GT (9-4), and even M-effing Troy scored 24 on them (you tell me Troy had a higher blue chip ratio than Ohio State, even though they were 10-3), and Auburn (8-5). That's a LOT of schools that theoretically would/should have had higher blue chip ratios than 11-1 conference champ Ohio State (arguing that ability to score is related to talent on your team).

For illustrative purposes, Ohio State's recruiting rankings (according to 247) for 2016 and previous 4 years (with 5 star and 4 star numbers in parentheses for that year):
2016 - 4 (1, 17)
2015 - 7 (1, 14)
2014 - 3 (1, 15)
2013 - 2 (1, 19)
2012 - 5 (2, 14)

As Miami is an indication, simply collecting as many 4-5 stars as you can is not entirely what makes Bama and OSU successful. It's recruiting related to your roster composition. UNC sucked last year because they had no depth at the skill positions to help Howell. They have a ton of highly rated recruits, but not in the right positions at the right times. The pervious staff was awful at roster management .. and for the well "you can't predict players bolting" crowd..1. Most times you can. 2. See my point above. You have 1 prima donna safety you trust to play? better get some depth behind him... good coaches get this.

I kind of have 2 points, maybe 3. I think teams can make it to the playoffs without having the blue chip ratio the article describes (just mostly not likely to be able to repeat that without improving recruiting). To your point, blue chip ratio by no means guarantees success, and I think FSU is probably an even better example over the past 5 or more years (they are usually within top 5 recruiting and weren't even bowl eligible one year recently). I guess the last point is - just how much talent is enough to have success?

I don't think Clemson was dominating recruiting landscape, and over the years I listed, they weren't a LOT better than Tech, but they certainly seem to have capitalized on that 2016 championship with their recruiting classes being fairly drastically better. Maybe that's the model we try to emulate. I know our championship game appearance didn't do a lot of good to our rankings, but how much of that was because a) we lost, b) we were primarily only recruiting VA, and c) the technology/evaluation issue we've discussed on accuracy of recruits' ratings?

Unfortunately, even if I'm right and we could potentially succeed and get to the playoffs, our roster is so hosed right now, it may very well take us 4-5 years to get the overall roster talent we need. Maybe less if we win games and our staff are great recruiters, we can get some good talent from the portal, and we actually develop the talent we DO have.

Unfortunately, even if I'm right and we could potentially succeed and get to the playoffs, our roster is so hosed right now, it may very well take us 4-5 years to get the overall roster talent we need. Maybe less if we win games and our staff are great recruiters, we can get some good talent from the portal, and we actually develop the talent we DO have.

Our roster is so hosed that its going to take us probably 3 recruiting cycles just to get the baseline established from which we can begin to try and legitimately chase the blue chippers. And that's before you consider the damage control within our own recruiting footprint that we need to do coming out of the Fuente regime. We have our work cut out for us, just to get back to where we think our base line should be.

This is my school
This is home

We can chase the blue chippers. If Mike London could get 5 stars to go to UVA in the glory years of their suckiness, it's possible to get them to go anywhere. I'm assuming it's a matter of the right relationships and figuring out the right way to sell the program. Hell, apparently, to some extent, winning isn't even important (apparently - again with Mike London).

I think teams can make it to the playoffs without having the blue chip ratio the article describes

This reminds me of that scene in Moneyball, where the scouts are living in some anecdotal nirvana and Billy Beane is in data-driven reality.

Scouts: "I like Geronimo. The guy's an athlete. Big, fast, talented. Clean cut. Good face. Good jaw."

Beane: "But can he hit?"

Scouts: "He's got a beautiful swing. The ball explodes off his bat. He throws the club head at the ball and when he connects he drives it."

Beane: "If he's a good hitter, why doesn't he hit good?"

Beane's theory is cool. But he doesn't win championships. Why? because your pitching "collective" that has the same WHIP as Scherzer in may doesn't work in the playoffs when facing other good hitters. That's where moneyball is flawed. Yeah, you can use 3 right field guys to get you 35 home runs, but Barry Bonds at the plate in the playoffs is a different animal. The Rays are brilliant at it, but they don't win world series with it. Fu tried it with our running back room... hmm if 4 guys combine for 150 yards in a game, we are good right? lol. doesn't work. Need studs to win.

Moneyball works. It absolutely works.

The reason the A's never got over the hump is because Boston, after Billy Beane turned them down, went out and hired Bill James, the person who literally wrote the book on baseball analytics, and Theo Epstein, the man that Beane recommended, and went all in on it. And because there's no cap in baseball, their deeper pockets were able to go out and literally create the best analytical teams that money could buy. And they very much are still operating as one of the more analytically driven front offices in baseball. They took the A's model and just applied a bigger budget to it.

4 World Series titles later and its hard to argue it doesn't work.

This is my school
This is home

The concept of moneyball is literally to not "buy a team".. thus the name. So when Boston's pay roll is 200 million dollars, regardless of analytics, its no longer money ball. Beane can get you competitive with what looks like a double A roster, I give him that. But Boston signing mega free agents is the antithesis of the concept. The Rays are "moneyball" Boston is not.

Moneyball is just an analytical concept to find inefficiencies in the current landscape and exploit them. It was about having a collection of players, and knowing how to identify those players, in order to create a winner. Yes, in the specific case of the A's, they did it to create a team that could play like a contender on a bargain budget, but when it started up they were able to do that because they had data that nobody else was using and they could exploit it to their heart's content.

The second that one of the big budget teams who could outspend them started using their same analytics and started bidding against those same players they wanted, just with a bigger budget? Well, that's when it was game over for the A's. But in terms of analytics as a whole, which is very much what the whole concept of Moneyball was about, its a slam dunk success that cannot be argued.

Bill James wrote the playbook, Billy Beane changed the game by introducing it to the sport, and John Henry and the Red Sox saw it to its natural bloom by applying a big budget to it. Once Boston went all in on the concept, it completely neutralized any competitive advantage that the A's once had.

This is my school
This is home

Moneyball started as Scott Hattiberg's production over time would equal Pudge Rodriguez at a 30th of the price. That was the ENTIRE concept. Numbers vs. specific players. That's the premise. It did not start out as simply synonymous with "analytics". it was using analytics to find similar "production" on the roster comparable to high priced players. The moment you have a top 3 payroll in the league, that concept goes out the window. You simply pay the best players.

Focused on just 1 thing that was measurable and not opinionable.
On base.

Can't do it all.
Do 1 thing. Identify it. measure it. execute on that 1 thing.

This is going to be great for the ACC.

Astros won the WS on it as well. Orioles have best farm system in baseball because of it. The Padres aren't on fire anymore because of it. Even outside of the Red Sox, it objectively is a good way to evaluate players.

Hell it even got the Cubs to win one

This is my school
This is home

Yeah, basically too many examples now.

You're missing the point of Moneyball. It isn't EXACTLY what the A's did. The point is applying math to solve problems. The A's had a different problem than Boston did. The As wanted to win, with a very limited budget. So they applied math to what was available and found a way to win. For Boston, you take off the "with a very limited budget" part of the problem. So you can apply the same theories and numbers, just with a higher budget. They could evaluate all the higher rated talent the same way (so essentially more upside, but with fewer flaws). It's a better way of looking at, of the available free agents out there, who is actually worth the money? It's also about challenging traditional metrics as what are the best predictors of success.

I think most people think of Moneyball as a baseball movie that also happened to talk about metrics. It's really a movie about applying analytics to problems (rather than gut feel), using baseball as an example.

So the Padres signing Manny Machado for 200 million dollars away from the orioles is what? superior analytics? Were the orioles not trying to win with Machado? Did the padres figure out how good he was or did they just pay him insane money? Like the yankees did in the late 90's? Was Randy Johnson a money ball thing? or did he throw 105 and want 100 million? His advanced numbers weren't that great. If you are saying that guys get paid based on analytics now vs. star power..? ok. maybe. How many low payroll teams have actually WON the world series? Like I said, Beane is great at taking a double A roster and going over .500. Winning championships? nah.

You're missing the point. There is a reason OPS is more important than batting average and RBIs now.

I don't think he's missing the point, I think he's trying to intentionally blur it

This is my school
This is home

All I'm saying is that you wouldn't know Billy Beanes name if the 90's As had the highest payroll in baseball and lost in the first round to the twins. Who would care? He is not famous for doing less with more. No, he is famous for trying to win in MLB with a basement payroll. His approach? - Moneyball- i.e. find production for less value. That is the entire concept. Now we can play semantics with that all day. But Beane;s theory is not based on a top 3 payroll justified by advanced numbers- it's not. If it was it wouldn't be "moneyball"- it would be the New York Yankees. All I'm saying. I'll be quiet now.

Beane's "theory" wasn't about payroll. It was all about how to gain an edge over how other clubs run things. The low payroll forced his hand to try it to gain some form of an advantage. Once everyone else caught on, the playing field was no longer even. It's how any team outperforms expectations before everyone else inevitably catches up. It's like Beamer with special teams. He found an edge to compete with bigger programs and then the bigger programs caught on and adopted what he was doing to remove that schematic edge he had. Beamer's schematic edge was necessitated by the financial disadvantage compared to bigger programs, but using starters on special teams isn't inherently because of the money.

"That move was slicker than a peeled onion in a bowl of snot." -Mike Burnop

His approach? - Moneyball- i.e. find production for less value.

You're confusing 'what' for 'how'

Winning with less is what Beane was tasked to do.

Moneyball aka Sabermetrics aka baseball analytics is how he accomplished it.

Once teams that weren't concerned with winning with less started adopting the methods that he used to do it, it neutralized the competitive advantage they had.

This is my school
This is home

Moneyball works. It absolutely works.

Moneyball plus money works. The analytical ability to identify the personnel you need, and the financial wherewithal to go out and acquire them. You need both.

It applies to college football as well, except replace "acquire" with "attract".

Edit: NM, I just responded to your first line. You get it.

If you're not sure if my comment warrants a "/s", it probably does.

Fu didn't try anything at any position other than throwing jello against the wall and seeing what stuck. QB and RB seemed to be a higher volume required to get something to stick problem.

To not make that post any longer, here are Clemson's and VA Techs recruiting rankings over that same time period:

2016 - 11 (1, 11)
2015 - 9 (3, 9)
2014 - 16 (0, 10)
2013 - 15 (1, 10)
2012 - 20 (0, 10)

VA Tech
2016 - 42 (0, 1)
2015 - 29 (0, 4)
2014 - 28 (0, 4)
2013 - 21 (1, 4)
2012 - 21 (0, 7)

Dang, those are some sad numbers. If Pry could pull 7 four-star recruits in a class today we would think that was amazing lol.

I think what we are dealing with right now is the fact that HS coaches and other private developmental organizations are wanting to see how well Pry (and any other coach) develops talent.

For the last 10 years, what we have dealt with has been 5 years of Beamer's age being used against him (which was a legit concern) followed by Fuente being unable to display the ability to develop players at any position.

My guess is that Pry is going with the long game, perhaps even borrowing a page from Wisconsin, by showing the ability to develop 3* recruits and punching above his weight class, this proving to those HS coaches that he can and will do great things to the higher touted recruits instead of that talent getting wasted.

I mean look at what a talent Isiah Ford was as a Freshman....a good coaching staff would have turned him into at least a 5th round draft pick.

IF only someone with 36 (all orphan) hours of Calc (Dr. Crittenden) and Stats (Dr. Mann) would Mcbryde Hall write all of this O&M, up...

...someone with a 24th in The Sporting News Championship bling; and a National Championship bling behind that in this very sport... ('thx' @Coach God, twice!!!)

Eye wonder... who???

35th and 57th... that's your legit boundary and your near kiss-of-death boundary, historically.


God Bless!

I'm technically proficient in both Stats and Calculus, although I'm confident in my abilities in neither. Still, I love math and I love to run statistics projects. (And when I'm questioning what ever it is that I'm doing, I run stuff by the wife- she teaches Stats in college and can usually unpack whatever bad ideas I throw at her).

That said, I'm not challenging your math here, but I don't necessarily with the premise of the two "ceilings" you've discovered. I believe your math and want to point out the culprit before it seems like I'm coming after you. We college football fans have been subjected to some bad data when it comes to recruiting.

While 247 and Rivals are the heavy hitters of college football recruiting and I respect the work they do, they aren't without flaws. And frankly, both services do team rankings wrong. IMO, They're both arbitrary, ineffective systems designed to add randomness and excitement into the (fairly boring and predictable) act of tabulating team success.

Rivals is more obvious. All the random-ass point bonuses look conspicuously like somebody set up a shiny booth at a carnival to scam people trying to play an unwinnable game.

I mean, at least they're not really hiding their intention.

247 is a little more devious. And because their over the top ("LOOK YOU BIG DUMMY- WE DID A CALCULUS!")

explanation is designed to make people like me feel really stupid and be super cautious in questioning it, it took longer to figure out.

First off, the entire explanation is bullshit- they completely shelved the method in they're describing a few years back. When they added their "Ratings Calculator", their methodology was exposed for a season. So they just scrapped it and went with something even dumber.

But let's go through that initial methodology first. When they first came up with that method, they must have agreed to use a bell curve.

Then one rogue guy left the office, stalked out a bell curve in the wild, shot it, chopped the carcass in half and dragged the damn thing back in the office for the next morning. And that's what they went with.

So not only is this not how bell curves work (the middle part is for the normals, not the freaks, guys- and hey, ya'll just chopped it in half like that did something sciency), the actual 2D image that this monstrosity provides some clue as to why it was so unpopular:

Only the top quarter of your class actually adds points. So if you add a new recruit, and he's say, right smack in the middle of your class, your team score changes from 214.56 to 214.58 points or something like that.

So to make their class calculator appear more "functiony", they decided to keep the basic core principle (FRANKEN-BELL CURVE LIVES!) and go back to something that puts out something more like an actual bell curve:

And so now only the very bottomest of recruits add almost nothing to a team's score. So what's the trick this time, and why do I claim it's even dumber than that first thing? (Full disclosure, the reproduction I created is only an approximation- it doesn't always nail the team score exactly. Which is good enough for me)-

Here's Clemson's 2020 class. While my low effort reproduction is a few tenths of a point off on occasion, it was derived from working backwards from their class calculator.

The key is the red column- they use that multiplier (or something close to it) and simply plug in each team's recruit scores from top to bottom to get a team score. I simply took the equation that excel spit out and used it cold- works pretty good for a quick job.

So what's actually a reliable indicator for class rankings?

Well, let me point out one more bad assumption before we get there. Rivals actually takes the time to explain their thought process in this article. Where it starts to go south is in this assumption (here they're trying to push you off of the correct answer; basically average recruit rank):

An argument a few college football fans make every year is that average star rating is a better indicator of class value. In the Rivals team rankings formula, however, average star rating is not factored in, precisely because it is not a good indicator of class value. Here is an example that makes this crystal clear:

You have a choice between two classes. Class A consists of 10 four-star prospects and average star rating of 4.0. Class B consists of 10 four-star prospects and 10 three-star prospects with an average star rating of 3.5. Which one do you choose?

The point they evidently completely fail to recognize is that college football has a limit on scholarships awarded. And every FBS school (I mean, maybe not Temple) is working hard at filling out that 85 player limit. So, no, if you're specifically a college football coach, you wouldn't just sign as many people as quickly as possible- you're going to be strategic about getting the best athletes.

So Rivals example is completely irrelevant- if you have room for 20 recruits, you take the best twenty you can. If you have room for ten, you take ten. You can't even compare those two situations. It's two completely different situations.

Joe alluded to it in this article here (spoiler- average recruit ranking)

And newcomer ON3 has about the best public methodology of the major services (although they cap the median class size, which I like, but they penalize small classes, which I don't. They also don't factor in JUCO's or Transfers- 247 only started counting transfers in 2022. ESPN's is proprietary and you have to pay to get their results. I haven't.)

So the best overall method was actually hidden right there in 247's explanation all along:

the average number of players recruited per team.

In the calculus equation from their original method, this is the lower case sigma in the denominator above e.

Simple is better, so the average recruit ranking is sufficient. And if you cap the average recruit ranking, you reward teams either cheating, or actively overhauling their roster- they're going out and getting extra guys each year to hit 85 schollys. So if the class average is 23 signees (it typically was before the portal), and Ole Miss signs 28 to a class, only their top 23 recruits count towards the average. That basically just means that they don't have to count the kicker and the long snapper, because they're actively managing their roster to fill out 85.

If USC takes 10 recruits, and they average 94.23, that's a legit score for the situation they're in. They clearly would be able to attract more recruits, but they could only take ten, and they got some hella good recruits this cycle. Apparently they're pretty happy with their roster, too, since they can only take ten. Who cares if it's a small class- USC is managing their talent well- just like their average score says.

All legitimately competitive programs operate on an 85 scholarship roster, it's important to minimize the significance of a single class size. (I mean REALLY minimize it- 247 make this claim, but larger classes have significantly higher scores. Rivals only counts the top 20). Recruiting is really a constantly rolling cycle of the last 4 classes. And average recruit score (capped at the average class size) is about the best way to do this, IMO.


Helluva an effort!!!

Literally, the best I've ever seen as a response goes.

...I did see some of that when I went old-school MegaStat and did my own scatter diagrams off of variance analysis. Line of best Fit, 4-squares, and what not... The average number thingy moving by 1/10ths don't impress me and Coach Shania Twain, much.

Good catch on that.

Though-- I went in and did my very in indicie. And then ran it (crosstabs) vs. wins. What Blue Chip count thingy do you need to have a chance to be really legit? (that's the 35th part). After your team dipped below 35th in my indicie, they have nearly no real shot at anything big. Think conference title or bowl win or 9'ish wins.

The soberstreet part was thus... once you dip below 57th in my indice, it is very very very difficult to tell the oranges from the tangerines. (get the attempted old-school bowl humor?) As they clustered or clumped in nearly overlapping Venn Diagram terms.

This means... it is not just the haves vs. have-nots... or the rich getting richer; the proletariat is what is thickening. And it is getting challenging to objectify any distinction(s) regarding them.

Hence, the Blue Chip Ratio, or just that one true shock-power-player, staying 10-12 games fully healthy is a Nadia Comăneci gold medalist in the vault. Think... our star transfer from KU as our only real Rb1 in a decade. Had k.Herbert remained healthy, Fu' could very well still be our big whistle.

It is that dang difficult -trending on umpossible- to parse the bottom-60 or so teams most seasons.

That was what science fact boggled my mind the most.
I had no expectation of that going in. Which is why you run the maths...


God Bless!

This might be the greatest nerd fight I've ever seen online, bravo!

"Call me anything-- just call me often."



God Bless!

I am not one bit surprised that these guys want to make you think they have advanced metrics that are better predictors than simple averages , gotta sell those subscriptions. Where they need to make their hay is in the evaluation side, improving the source of the data and plugging it into the simple composite.

I am not sure what to do with my hands now

@Brockman - Have you ever considered using upper quartile ranking to score recruiting results? I did some (significantly simpler) 'analysis' a few years ago and noticed that, while Fuente classes always had a higher floor than Beamers', Beamer's classes always had higher ceilings.

My theory is that, although Beamers' teams had a lower average recruit rating, they almost always had a few 'game breakers' (eg; high 4/5-star players) who could change the outcome of a game. Fuente almost never had this at VT, despite having a higher average recruit rank for the team.

TL;DR - I would rather have one .8500 and one .9200 than two 0.885, because (as I see it) the difference between a .9200 and a .8900 is greater than the difference between a .89 and a .85.

Thoughts? Am I making sense or no?

Twitter me

Honestly, that blue chip graph up top was one of my early efforts into something like that. I've been just generally using data to get a feel for how their scale is built (since I'm using it to retroactively apply it to older data). But I really haven't been testing it out a lot. I do generally trust their player ranking curves (ON3 uses a slightly steeper descent, but almost the same numbers across their scale).

I do think that this general shape (in theory) performs that function that you describe. If the curve works accurately, there should be the same difference everywhere on this curve with basic arithmetic (there's less players the higher you go, thus the steeper ascent rate from 100).

That's more like how a bell curve should work. (I don't necessarily agree with the steep descent on the bottom side- we're not ranking half the HS football players in the nation, just like 1% or so. I'm not really sure why this shape needs to resemble an entire half of a bell curve. If anything I'd think it would go asymptotic with the x-axis for the level of prospects we're dealing with).

Ranking the players in this way, basic averages should get you to a sound team score.

unplanned but high quality #content like this is what makes the TKP community so great. thanks Bar, Brockman, b.street for the high effort posts and discussion in this thread, you deserve all the turkey legs for it

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

Since I've gone pretty off the rails already, I figure there's a couple more interesting visualizations on the consensus that this player ranking curve (or something like it) is a pretty well established concept.

When I was just hitting the home stretch of my pandemic project, I suddenly discovered upstart ON3 (a couple of guys from 247 branched out and started it). I contemplated adding their scoring to the recruiting database, but in the end (No JUCO's, and ON3 seems more obsessed with the Transfer Portal and becoming the NLI stock market to really worry to much about just yet).

But I did break down the FBS signee rankings for the 2020 247, ON3 and ESPN scoring to see how they compare:

All three services seem to really like working with this basic curve idea (or something like it). Kinda neat.

Also, for some reason I thought it would be neat to track out the FBS signees in the 2022 composite from summer to signing day. I decided to learn how to make a gif for it:

The GIF at the bottom basically says more kids are ranked by signing day each year?

Twitter me

Well, it's just for 2022, but ESD makes for earlier decisions for the bulk of recruits. I hadn't done this before, so I couldn't say for sure how much.

I picked a weird year to do this. Normally between HS and JUCOs, average class size was right around 22.5 signees. With the Covid bottleneck, coupled with the Transfer portal floodgates open (NLI probably stratified P5 and G5 even more than before), this doesn't show much of anything typical- more like a huge system in flux.

You think it's bad for a High School kid last year- I think the number of JUCOs taken has gone from a fairly steady 300/yr to only 50 in 2022.

The most interesting thing to watch for me is the floor. Typically the lowest score for a signee is 70. You can see the scores start to drop to 75 in August. And after that, nobody gets graded below a 75. I think the average player ranking was almost 2 points higher than a normal year. Makes sense; there have been close to 2700 High School signees a year for a while now, last year there were more like 2100.

2019 was a typical year for signees, but the influx from the portal started. 2020 is when the number of typical signees started to drop.

I think we're still a year or two away from seeing what "typical" looks like going forward.