FreddieMercuryHayes

January 23rd, 2013 at 1:31 PM ^

Is it?  Putting aside the small sample size (53 by my count), can this really point to star rankings being predictive?  And predictive of what?  Super Bowl appearences?  Putting aside the 13 players who are N/A (before Rivals), we've got 40 players.  21 (52%) were ranked at 3* or lower.  Now, you don't have to convince me that star ratings are more predictive of college success, wins, and chances at being drafted.  But trying to extrapolate that to NFL success seems too thin at this point.  It's hard enough trying to predict how a high schooler will succeed at the college level.  Now trying to extrapolate that to what happens when an NFL player starts a career with tons of money playing professionally?  Maybe in 10 years when all NFL players would have gone through the Rivals age we can look at the entire NFL two deep and try and draw conclusions, but from what I can tell just the above data isn't really showing much except star rankings become less important to succeeding in the NFL.

FreddieMercuryHayes

January 23rd, 2013 at 2:08 PM ^

I mean I don't know the answer to that.  I know the chances of getting drafted gets higher with your star rating, but I have seen no data about starting in the NFL based on star data.  I suppose we could look at the entire roster of each team and at least get a data point by how many "stars" are on the 2-deep, roster in general, and practice squad.  That would be interesting to look at.  But I'm really too lazy to do that.  I'm sure someone will sometime.  One minor confounder in this is that we still have many players in the NFL playing that came before the huge easily available recruiting databases and rankings we have now.  I think we could find out a lot in 10 years when everybody was evaluated by the ranking services.

There's actually a lot of interesting analyses one could do.  I think it's already shown that star rankings are predictive of getting drafted.  But what about making an NFL roster?  Does that change to percentages?  Or once you've made an NFL roster, do the star rankings lose their predictability of actually making the 2-deep?  Of starting?  Does the correlation strenghten or weaken?

saveferris

January 23rd, 2013 at 2:57 PM ^

When you think about it,  there are about 10,000 players playing at 120 FBS schools.  Given that each recruiting class is made up of about 20 5 stars, and 100 4 stars extrapolated over 4 years, you're talking about 1% of the college football population being 5 stars and another 5-6% being 4 stars with the rest being 3 star and below.  So we're talking about 500-600 players.  Since the annual manpower requirements of the NFL is 1700 people, your roster will consist of a large percentage of players who were rated 3 stars or lower.

The fact that almost 40% of the starting line-ups in the Super Bowl are drawn from what was rated the top 5% of the high school talent pool provides very strong correlation of star rating and future NFL potential.

turtleboy

January 23rd, 2013 at 2:24 PM ^

Only one star rating shows that they are predictive in Haloti Ngata's. He was a 5 star in high school, college, and the NFL. Most of the remaining best players were 3 and 2 star players like Kaepernick, Iupati, Staley, Willis, Aldon Smith, and basically the entire Ravens offense.

turtleboy

January 23rd, 2013 at 8:13 PM ^

I suppose it depends on what conclusions you are looking for in the data, but there are more unranked players starting than 5 stars. 5 to 4. My conclusion was that there are 4 and 5 star players on both rosters, but many of the lower rated players coming out of high school are better than many of the 5 star players. Iupati (2) is a much better Guard than Boone (5,) and Staley (2) is a much better Tackle than Davis (4.) Smith (3 ) is a much better Linebacker than Upshaw (4,) and Willis (3) is the best linebacker in the NFL. 

Star ratings determining entrance into the NFL do correlate more often than not, but my conclusion was that there is little correlation between being one of the top 20 players in high school (5star), being one of the top 20 players in the draft or All American, and then being one of the top 20 players in the NFL in the Pro Bowl and such. Haloti Ngata is one of those rare exceptions that did all 3, while most of the other "best" players in the NFL were 3 star players.

JeepinBen

January 23rd, 2013 at 1:13 PM ^

Some guys peak in high school. They're so much better than everyone that they get the 5th star and go on to the great college, and don't end up the best anymore compared to the competition. Some guys peak in college, they go on to be high draft picks. Some guys peak after being in the NFL for a bit, they go on to have long NFL careers.

It's pretty rare that a guy will stay at his relative athletic peak for the entire time. One that jumps out is Peyton Manning. I don't know his star ranking, but he was a top draft pick who has performed at a high peak for many years. There aren't a ton of those guys.

Dutch Ferbert

January 23rd, 2013 at 1:09 PM ^

Alabama and Ohio State have two starters each. So Rutgers and Maryland clearly are at the same level as Bama and OSU.

Forget the tv ratings and footprints. The B1G just picked up a couple of football juggernauts.

elm

January 23rd, 2013 at 1:15 PM ^

I was going to post a similar comment, noticing that Michigan, Iowa, Penn State, and Purdue each had one.  Wisonsin and MSU, notably, have none.

 

Obviously, this is two teams and the picture would look different if NE and ATL had made it instead, but Rutgers, at least, has become a pretty good pipeline to the NFL in recent years.  Their problem seems to be more about depth and stability (especially along the lines and at QB) rather than lack of star players.

elm

January 23rd, 2013 at 1:22 PM ^

Quickly looking at what would have happened if the conference championships had gone the other way, no Big10 team would have more than one starter, with Michigan, Rutgers, Illinois, Purdue, Nebraska, Wisonsin, and Iowa each having 1 starter.

So, Rutgers had the most starters in the conference championship and MSU ( along with only Minn. and Indiana) was shut out.

elm

January 23rd, 2013 at 1:44 PM ^

Since I'm bored at work, I looked at starters for all playoff teams:

Wisconsin/Ohio State: 6

Maryland: 5

Michigan: 4

Ill/Purdue/MSU/Rutgers/Iowa: 3

Nebraska/Penn State: 2

Minn/NW/Slippery Rock: 1

Indiana: 0

Surprised that Maryland was so high and Penn State so low.  And poor, poor Indiana.

 

Ron Utah

January 23rd, 2013 at 1:17 PM ^

This would seem to add credence to the theory that ratings matter much more on defense than they do on offense.  Both defenses are loaded with top tier players, and there is only one two star (Paul Kruger) and he was an athlete who played QB in high school.

Data:

  • 12 of the 21 rankable defensive starters either had four or five stars
  • 6 of the 19 rankable offensive starters either had four or five stars--the same number of rankable offensive starters were two stars or unranked
  • Only 3 defensive starters were two stars or unranked

Recruiting matters, especially on defense.

ChiBlueBoy

January 23rd, 2013 at 1:21 PM ^

If the writer wanted to make a point about rankings, how about giving the percentage of players who enter college as 5*, 4*, 3* and 2* and then compare that to the percentage on the SB-bound teams? That would actually give some meaning to the numbers. Notwithstanding the write-up, I couldn't help but notice the goodly percentage of 3- and 2*s on the teams who, if you believe the ranking services, should not have made it into the NFL at all.

I'll point out that I've seen much better analyses by the "mere" bloggers and posters on this site than the drivel in the article regarding the correlation between *s and success in college/pros.

MileHighAnnArborite

January 23rd, 2013 at 1:59 PM ^

A guy on the Colorado Rivals board did this a few days ago.  Don't want to repost his numbers, but the takeaway was that you are much more likely to be on one of these two teams if you were a 5* than a 4*, and more likely if a 4* than a 3*, etc.  With the diffference in %s being about double (i.e. a little over 2% of the 5*s, 1% of the 4*s, etc.).

bluebrains98

January 23rd, 2013 at 1:25 PM ^

Keep in mind this is only for the Super Bowl teams. We could do this next year, and SF and BAL will have very similar rosters but likely won't be in the Super Bowl. Unless you do this league wide, or at least for all playoff teams, it doesn't seem prudent to jump to any real conclusions.

copacetic

January 23rd, 2013 at 2:27 PM ^

 

Unforunately, more than a dozen starters in the Super Bowl were recruited prior to 2002 [when Rivals started doing stars], including the celebrated likes of Justin Smith, Terell Suggs, Anquan Blodin and a quartet of former Miami Hurricanes, Ray Lewis, Ed Reed, Frank Gore and Bryant McKinnie

 

Also Randy Moss.

It would be interesting to see what a list like this would like for super bowl teams 5-10 years from now. I imagine many more highly ranked players. As other people mentioned, percentages would be more useful as well

gwkrlghl

January 23rd, 2013 at 2:37 PM ^

the number of 5*'s there versus the number of 5*'s ever is pretty amazing. There's probably 25-30 a year over only 8-9 classes of players who've made it to NFL age and there's 4 of those guys on just two teams plus a bunch of other 4*'s. The rankings guys definitely know what they're doing for the most part

LandryHD

January 23rd, 2013 at 2:41 PM ^

Though Alex Boone(5*) made the Pro Bowl, he is probably the worst lineman of those 5. He got into a lot of problems early in his career and was on the bench up until this year. He was moved to RG and it got him a Pro Bowl.

FreddieMercuryHayes

January 23rd, 2013 at 3:53 PM ^

I have found Alex Boone as a very funny example lately.  In the usual battle of back-and-forth between rivals where you both try and point out any weakness another may have and why it's the worst ever, OSU fans have been using Alex Boone as an example why our loaded OL classes won't mean crap when they get on the field.  Apparently having a bad game in the NC game as a true sophomore against a 5-star senior means all OL who are highly ranked don't pan out.  Super Bowl Alex Boone disagrees.

Seth

January 23rd, 2013 at 3:49 PM ^

I followed recruiting back in the '90s probably more than I do now and while star ratings weren't around yet I can tell you many of the pre-2002 dudes on there were huge HUGE recruits.

Randy Moss and Anguan Boldin were among Lemming's Top Five players OF ALL TIME. McKinnie, Terrell Suggs, Ed Reed, and Frank Gore were all very highly sought after, and Gore and Suggs were said to have been paid to go where they did (I never heard it for McKinnie and Reed and Ray Lewis but they were there for some of the really dirty Canes teams so...). Goodwin was a nobody as a freshman but then his brother became a big deal and Goodwin was All-MAC as a true freshman before transferring to Michigan.

 

befuggled

January 23rd, 2013 at 6:30 PM ^

They wouldn't admit, though, because of a little run-in he had with the law (I believe he put somebody in the hospital in a fight). So he transferred to Florida State, where he sat out for a year before getting booted off the team for violating his parole (by smoking pot). He transferred to Marshall the last year it was I-AA, so he didn't have to sit out another year.

I saw him play a game when he was at Marshall. They nuked one of the poorer MAC teams, since Marshall was in the MAC at the time. I think it was Akron. Chad Pennington was the QB on that team.