This confuses me about 247's composite rankings.

More weight is given to higher rated recruits, it looks funny there because we have 15 4-star to their 11, but if you look at their top top recruits I bet they overall are higher.

The way total points is calculated is not based on an average at all of your recruits.

For example if you top recruit it the #1 overall player and they are worth 30 points, you get like 29.998 points for them. Go down to your #10 recruit in the class, you get a lot less of their value. Go down to the #20 kid and you get very little for them.

Basically your best recruits matter the most and receive the majority of their weight in points. A good way to look at this is open up class calculator.

A Recruit worth 20 points may get 19.98 points in 1 class vs. 10.12 in another because of position in the class itself.

The point he's trying to make is that our average recruit is rated 92.61 and Ohio St's is 92.14. It is the numbers in grey, second from the right. He's right in saying it makes no sense. (Sum of ratings/#of recruits) = Avg Rating, so then wtf are points? Seems like a way for 247 to rank people above others.

The avg points per recruit is just that, a straight average of each individual recruit's RANKING value (not points). #1 Bresee is .9993, which will show up as a 99.93 on that average you are looking at.

The points are not an average, but a weighted mean of assigned points (see poster above example with his #1 recruit = 29.9998 points). The point system is the weighting.

Player rankings, averaged or otherwise, are only 1 factor in points.

From 247's site:

quote:

---

Each recruit is weighted in the rankings according to a Gaussian distribution formula (a bell curve), where a team's best recruit is worth the most points. You can think of a team's point score as being the sum of ratings of all the team's commits where the best recruit is worth 100% of his rating value, the second best recruit is worth nearly 100% of his rating value, down to the last recruit who is worth a small fraction of his rating value. This formula ensures that all commits contribute at least some value to the team's score without heavily rewarding teams that have several more commitments than others.

---

quote:

---

A Recruit worth 20 points may get 19.98 points in 1 class vs. 10.12 in another because of position in the class itself.

---

As a guy that looks at and develops statistics every day for living, I just don't think this is the best model.

Recruits should all get a value and should retain it. That is a much cleaner way to do this. They want to weight them based on individual school classes because it brings everyone closer to together that have different #'s of recruits.

They grade on a curve.

ETA:

This is fine for one view, but the main view, IMO should be average recruit ranking with some weight for % of class filled.

This would quantify how well each school filled their respective open slots with what level of talent.

Their 4 starts are also ranked a lot higher too. 11 in the top 250

we have 15 4* but 6 are outside 250.

(this is just to explain why we are were we are, not my feelings on the rankings)

quote:

---

This confuses me about 247's composite rankings.

---

It's just a way for them to control which teams have the higher rated recruits. You'll see an alabama recruit get a big bump soon as they commit. Same thing goes with Clemson. These rankings are biased and only useful as a reference. You want to know how good a kid is, go watch their videos yourself. Evaluation should be unbiased. I tend to trust O and staff. I think they are proving to be very good evaluators.

quote:

---

What it means is most of the recruiting ratings are nonsense. Rating a high school kid 99.93 is absurd. What ever. General rankings of 4 star and five star I guess are ok. Even that is not very accurate.

The NFL is half 4 and 5 stars and half 3 stars and below.

---

And Burrow should be a 6 star IMO.

quote:

---

As a guy that looks at and develops statistics every day for living, I just don't think this is the best model.

Recruits should all get a value and should retain it. That is a much cleaner way to do this. They want to weight them based on individual school classes because it brings everyone closer to together that have different #'s of recruits.

---

Here’s a post in another thread on this subject, but it’s worth mentioning again:

It’s not just about balancing based on number of recruits in each class. Their method is also related to how the recruits are expected to contribute to on-the-field performance. Remember that teams get 25 scholarships a year, and each recruit gets 4 years of eligibility. Contrast that with the fact that there are only 22 starters on offense and defense.

Consider two teams:

Team A - 25 recruits, each with a 0.89 rating.
Team B - 5 recruits with a 0.95 rating, 15 recruits with a 0.89 rating, and 5 recruits with a 0.83 rating.

Now assume they maintain the same class ratings every year (for the sake of argument). When you compare the two squads:
- Both teams have an average rating of 0.89
- 25% of Team B’s players are rated higher than Team A
- 50% of Team B’s players are rated equal to Team A
- 25% of Team B’a players are rated lower than Team A

That means the best 22 players for Team B (the offensive and defensive starters) will have a higher average rating than Team A’s best 22. In other words, the top recruits in a class are expected to contribute more than the bottom recruits, regardless of the overall quality of that class.

The other factor is that the bottom the recruiting class is more likely to be processed out to stay under the 85 scholarship limit. I’d imagine this part is a wash, though, when you consider that the top recruits in a class are more likely to declare for the draft early.

TL;DR - the 247 system does make sense when you’re trying to correlate to on-the-field performance of the team.

quote:

---

The NFL is half 4 and 5 stars and half 3 stars and below

---

Oh boy...

There are 300-400 recruits every year with a 4- or 5-star rating. There are roughly 300,000 players every year with a rating of 3-stars and below. About 6,400 sign division 1 (FBS or FCS) scholarships each year.

So 4- and 5-star recruits make up roughly 0.1% of high school football players and roughly 5% of college football players. But they account for 50% of NFL rosters.

If you do the research you will find that a 5-star recruit is, generally speaking, 10x more likely to be drafted in the NFL than a 3-star recruit. So yes, the ratings do mean something.