I've Come to Realize that Stars Mean Everything

[xc84]

As a general rule, I would take a 3 star kid that Florida State wants/gets over a 4 star kid who goes to UConn all day.

 

[Lou_C]

On this page there are links to a series by Matt Hinton that break down the star system incredibly well, and why they definitely matter.

http://www.tomahawknation.com/2011/1/31/1965917/recruiting-rankings-do-matter

 

[Crusty]

On this page there are links to a series by Matt Hinton that break down the star system incredibly well, and why they definitely matter.

http://www.tomahawknation.com/2011/1/31/1965917/recruiting-rankings-do-matter

We have covered those points already. They do not speak directly to the hypothesis which is - what impact do 4 and 5 star players have on winning. Total team rank does not answer that question. Nor does the number of 4 and 5 stars that become All-Americans or get drafted. It is really nothing more than anecdotal information as it does not speak to W-L record or national ranking.

We really don't know things such as does it matter what position the 4 or 5 star plays? Do they leave earlier? Etc., etc.

Look, I think there is a correlation but I have no idea how strong it really is or is not and what subsets of data might be revealing.

 

[Lou_C]

Really?

From http://sports.yahoo.com/blogs/ncaaf...er-defense-recruiting-rankings-214251813.html

"2011 was not a banner year for the usual suspects — Ohio State turned in a losing record for the first time in well over a decade; Florida and Texas barely struggled past .500 — and a great one for middle-of-the-pack recruiters like Arkansas, Michigan State, Oklahoma State, Stanford and Wisconsin. Still, on the final count, the higher-ranked team according to the recruiting rankings won more than two-thirds of the time (68.7 percent of the time, to be exact), and every "class" as a whole had a winning record against every class ranked below it. The gap on the field also widened with the gap in the recruiting scores: At the extremes, "one-star" and "two-star" recruiting teams managed just five wins over "five-star" recruiters — four of them coming at the expense of Florida State and Texas — in 31 tries.
It's a simple equation: The better your recruiting rankings by the gurus, the better your chances of winning games, against all classes. Emphasis on the word chances — the counterexamples are obvious and legion in both directions. But as far as forming a reasonable basis for predictions, well, it probably goes without saying that you never want to count on being one of the anomalies."

That doesn't answer that hypothesis? Either you didn't read it, or I don't understand your hypothesis. I did go to FSU, so that could be.

There's plenty of data support for the idea that teams with higher rated recruits win more often. There's no data other than the anectdotal or occasional anomaly to suggest otherwise.

 

[Crusty]

Really?

From http://sports.yahoo.com/blogs/ncaaf...er-defense-recruiting-rankings-214251813.html

"2011 was not a banner year for the usual suspects — Ohio State turned in a losing record for the first time in well over a decade; Florida and Texas barely struggled past .500 — and a great one for middle-of-the-pack recruiters like Arkansas, Michigan State, Oklahoma State, Stanford and Wisconsin. Still, on the final count, the higher-ranked team according to the recruiting rankings won more than two-thirds of the time (68.7 percent of the time, to be exact), and every "class" as a whole had a winning record against every class ranked below it. The gap on the field also widened with the gap in the recruiting scores: At the extremes, "one-star" and "two-star" recruiting teams managed just five wins over "five-star" recruiters — four of them coming at the expense of Florida State and Texas — in 31 tries.
It's a simple equation: The better your recruiting rankings by the gurus, the better your chances of winning games, against all classes. Emphasis on the word chances — the counterexamples are obvious and legion in both directions. But as far as forming a reasonable basis for predictions, well, it probably goes without saying that you never want to count on being one of the anomalies."

That doesn't answer that hypothesis? Either you didn't read it, or I don't understand your hypothesis. I did go to FSU, so that could be.

There's plenty of data support for the idea that teams with higher rated recruits win more often. There's no data other than the anectdotal or occasional anomaly to suggest otherwise.

Lou, the hypothesis is not higher ranked recruiting classes producing higher ranked teams, The hypothesis is limited to the impact of only 4 and 5 star players and nothing else - i.e. can't count 3 stars etc in recruiting class rankings.

 

[anomander]

Lou, the hypothesis is not higher ranked recruiting classes producing higher ranked teams, The hypothesis is limited to the impact of only 4 and 5 star players and nothing else - i.e. can't count 3 stars etc in recruiting class rankings.

there is no possible way to figure this out

 

[RF2044]

Individually the star system is highly flawed. But in the aggregate it tends to be somewhat predictive.

Ding, ding, ding.

Astute.

 

[Crusty]

there is no possible way to figure this out

Probably so. If you have the 4 or 5 stars for each school for say, 5 years or so, you could get a number. I just don't know how good it would be. There are just too many variables. It would also be specific to the rating system chosen.

I have a hunch that the number would show a higher correlation than overall class ranking.

 

[richard levy]

I am not sure that is the metric to use. What I think is important is the number of 4 and 5 stars on ranked teams not the overall recruiting rank.

the number of stars, or the average star per recruit or a number, is what determines recruiting rank

 

[Crusty]

the number of stars, or the average star per recruit or a number, is what determines recruiting rank

I know, I know.

 

[sufandu]

It's easier when you have a roster filled with 4 and 5 star recruits and if a couple turn out to be busts so be it. Easier to win that way then having a roster of 1,2, and 3 star recruits and you hope one or two are diamonds in the rough and end up being 5 star type level players.

While the "STAR" system isn't 100% accurate its a pretty damn good judge of talent at the end of the day.

Very true. The "star" system isn't the be all end all, but if you're loaded with 4 and 5 star players the ones that bust are just pushed to the background as the ones that meet expectations take over. When you only get 1 or 2 of those guys, you NEED them to meet expectations, and when they don't it's much more glaring. FSU doesn't just have 4 and 5 stars on the field, they have 4 and 5 stars that will never play in anything but mop up time, but you'll never hear about those guys being a bust because the guys that excel just overshadow them.

 

[sufandu]

the number of stars, or the average star per recruit or a number, is what determines recruiting rank

Actually it's the number of points assigned by the recruiting service for each guy added together. That's why a class with 25 recruits can have a higher ranking than a class with 20 recruits with a higher average star rating. That's why Tennessee is ahead of Georgia, Texas A&M, Ohio State, and Notre Dame in r*vals rankings despite having a lower average star rating.

 

[PhatOrange]

Rvls currently has the Syracuse class rated #62 with only 13 commitments. But if you sort it by average star rating (to which Syracuse checks in at 2.77) Syracuse is rated #47

FSU by the way is #6 (total) and #18 (average 3.28) – (25 commitments)
0 – 5 stars
10 – 4 stars
12 – 3 stars