Landing Moe today was one of those moments...

[FrancoPizza]

Agreed. Like I said last week, give me a dozen 4 stars and if half are huge flameouts, I still have a half dozen 4 stars.

Furthermore, it's less likely that a 2-star player will outperform relative to expectations than, say, a 4-star player will under-perform. Will there always be exceptions? Yes. But on the whole, stars matter in an absolute sense.

 

[bcubs9497]

Furthermore, it's less likely that a 2-star player will outperform relative to expectations than, say, a 4-star player will under-perform. Will there always be exceptions? Yes. But on the whole, stars matter in an absolute sense.

At best, the ratings measure probability. Probability being calculated in ways that are proprietary, and private. We don't know what kind of biases are built into these calculations.

 

[TheCusian]

Of those 4 schools? Probably ours. But what's your point? There are 2 ways to sort the site databases. Total points which benefits a larger class or average per player which ignores class size. In some cases out average is worse than the total so the excuse we have a small class doesn't always hold water.

That's my point. I'd take our class over the next 4-5 schools this year. Obviously the best case is high average and lots of them.

But it's trending up. And that should make everyone happy. Good enough? Of course not. But moving in the right direction is good news coming off of 3-9.

 

[007]

Avg rating should be the primary metric. It's not like some schools have a higher scholarship limit. Every school, over 4 years, will have doled out the same number of schollies.

I'd love to see one of our resident statisticians compile a weighted, 4-year ranking based on average stars and see how that measures up with the current top 25-50 in the polls or whatnot. Compile the 2011, 2012, 2013, and 2014 classes. To calculate overall impact on the field for this year, multiply the averages by seniority which factors in projected playing time by class. Current seniors and juniors (2011 and 2010 recruiting classes) would be weighted at 1.0... then sophs at .75... and RFs at .25. Just an idea, but I'd be surprised if there was not a strong correlation to the star-system when all is said and done.

Those studies exist, and have been talked about here many times. There is overwhelming evidence to support a significant correlation between rankings and winning across all variables. Here are a few:

http://www.econ.ohio-state.edu/trevon/pdf/Bergmen_Logan.pdf

http://www.cbssports.com/collegefootball/eye-on-college-football/21641769

http://www.footballstudyhall.com/20...-matters-why-the-sites-get-the-rankings-right

 

[reedny]

Those studies exist, and have been talked about here many times. There is overwhelming evidence to support a significant correlation between rankings and winning across all variables. Here are a few:

http://www.econ.ohio-state.edu/trevon/pdf/Bergmen_Logan.pdf

http://www.cbssports.com/collegefootball/eye-on-college-football/21641769

http://www.footballstudyhall.com/20...-matters-why-the-sites-get-the-rankings-right

These studies are interesting, and they do show that, at the team level, star rankings correlate. At the individual level, however, it's still a crap-shoot. As stated in the OSU study (2013, ex ante look at team ratings):

There is little empirical evidence that individual recruit ratings are strongly related to performance, but there are studies which show that particular aspects of team performance are related to recruit quality (Meers 2013).

And in the CBS sports article, they admit the same thing. In fact, they illustrate the point with a picture of a Heisman Trophy Winner who's a 3 star (Manziel).

 

[TheCusian]

I'd love to see the correlation between measurables and stars. I'd bet most of the recruiting sites use size/speed/strength/agility as a big filter. And that filter largely works (big + fast = good) as intended.

So it's not a shock to see the stars work out on a macro level. Team x has bigger and faster players, so they win more.

It's the gap between "not quite big enough" and "has great leadership" or "just knows how to play" where things get muddy. And that's why coaches don't rely on recruiting services for more than size/speed etc - and even then they want to see them in person.

 

[Scooch]

These studies are interesting, and they do show that, at the team level, star rankings correlate. At the individual level, however, it's still a crap-shoot.

I think most people are in agreement with that, and that's what we've been discussing.

 

[reedny]

I think most people are in agreement with that, and that's what we've been discussing.

I saw that you posted something to that affect (as did Bleed about the "macro level"). Just making sure we're keeping stars in proper perspective .. since the OP was about one kid (MN).

 

[rrlbees]

That's my point. I'd take our class over the next 4-5 schools this year. Obviously the best case is high average and lots of them. But it's trending up. And that should make everyone happy. Good enough? Of course not. But moving in the right direction is good news coming off of 3-9.

You're making a totally different argument now than the original one I was commenting on.

 

[rrlbees]

Those studies exist, and have been talked about here many times. There is overwhelming evidence to support a significant correlation between rankings and winning across all variables. Here are a few: http://www.econ.ohio-state.edu/trevon/pdf/Bergmen_Logan.pdf http://www.cbssports.com/collegefootball/eye-on-college-football/21641769 http://www.footballstudyhall.com/20...-matters-why-the-sites-get-the-rankings-right

Those studies or others like them have been pointed out before by others. A lot of people don't get it though.

 

[007]

Those studies or others like them have been pointed out before by others. A lot of people don't get it though.

Yeah, tell me about it. I have referenced those studies and articles repeatedly over the years. To be fair, I think it's tough to wrap your head around concepts like statistical significance, correlation coefficients, and the law of big numbers without some background knowledge in data analysis.

 

[TheCusian]

You're making a totally different argument now than the original one I was commenting on.

No I'm not. The point is: you'd take our class over the others this year and so would I. No more, no less.

 

[Scooch]

Yeah, tell me about it. I have referenced those studies and articles repeatedly over the years. To be fair, I think it's tough to wrap your head around concepts like statistical significance, correlation coefficients, and the law of big numbers without some background knowledge in data analysis.

Agreed, but people should be able to understand the underlying concept of sample sizes.

I mean, a bad analogy is that you don't judge a baseball batters overall effectiveness from one game where he had 4 at bats. That's the equivalent of one kid's star rating. You judge the batter over the 600 at bats he had all season, and the multiple seasons he's played. That's the equivalent of the total star rating of a class, and multiple classes over time.

 

[rrlbees]

No I'm not. The point is: you'd take our class over the others this year and so would I. No more, no less.

You've changed your argument. First it was we haven't been in this "range" in a decade. False. Then you went the route of how many total recruits. Average team ranking isn't much different. Then you show me 5 teams and ask which is better. Awful small sample size. You've been all over moving your goalposts.

 

[TheCusian]

You've changed your argument. First it was we haven't been in this "range" in a decade. False. Then you went the route of how many total recruits. Average team ranking isn't much different. Then you show me 5 teams and ask which is better. Awful small sample size. You've been all over moving your goalposts.

You act like I haven't acknowledged those arguments and where I was mistaken (a complete rarity on this board and esp from you).

- 8 years still proves the original point in so much as it's been a long time. And again - I acknowledged my error.

- The players we've gotten DO show that though we have a smaller class, they are of a higher quality than those 5 teams. Which is good. Not a moving goal post. (Also when I looked 3 months ago there was a significant difference between avg and class size in the ACC -like 5 spots). I acknowledged that it wasn't as large a gap as I thought.

- Normally, if someone says "huh, I was mistaken" the other person says "okay, cool" not "these arguments when taken as a whole don't make sense" ... Because yeah - they weren't meant to be read that way.

- Finally. I am not being confrontational. I'm not starting a fight. I believe we would like to see the same results as fans.

 

[rrlbees]

You act like I haven't acknowledged those arguments and where I was mistaken (a complete rarity on this board and esp from you). - 8 years still proves the original point in so much as it's been a long time. And again - I acknowledged my error. - The players we've gotten DO show that though we have a smaller class, they are of a higher quality than those 5 teams. Which is good. Not a moving goal post. (Also when I looked 3 months ago there was a significant difference between avg and class size in the ACC -like 5 spots). I acknowledged that it wasn't as large a gap as I thought. - Normally, if someone says "huh, I was mistaken" the other person says "okay, cool" not "these arguments when taken as a whole don't make sense" ... Because yeah - they weren't meant to be read that way. - Finally. I am not being confrontational. I'm not starting a fight. I believe we would like to see the same results as fans.

The 8 years should be 2 years.

 

[TheCusian]

The 8 years should be 2 years.

So in your world 51 is less than 50?

Don't answer - tried to build a bridge, try to acknowledge when I'm wrong, try to be a fair person. I can live without knowing or agreeing with your opinion.

 

[rrlbees]

So in your world 51 is less than 50? Don't answer - tried to build a bridge, try to acknowledge when I'm wrong, try to be a fair person. I can live without knowing or agreeing with your opinion.

No it's not less. But it's in the same "range".

 

[Toga]

So in your world 51 is less than 50?

Don't answer - tried to build a bridge, try to acknowledge when I'm wrong, try to be a fair person. I can live without knowing or agreeing with your opinion.

 

[TheCusian]

No it's not less. But it's in the same "range".

Okay - let's rephrase so you can go to bed. How does "Shafer and staff have accrued the 3 best classes in the last 8 seasons. With this class not being done yet, 2016 may end up as the best in the last 8 years."

Problem solved. Thanks for doggedly pursuing that to its fitful end.

 

[TheCusian]

I was trying so shame him into acting like a human, but it didn't work.

 

[rrlbees]

Okay - let's rephrase so you can go to bed. How does "Shafer and staff have accrued the 3 best classes in the last 8 seasons. With this class not being done yet, 2016 may end up as the best in the last 8 years." Problem solved. Thanks for doggedly pursuing that to its fitful end.

Bed? I have voting for Floyd to do.

 

[rrlbees]

I was trying so shame him into acting like a human, but it didn't work.

Don't move goal posts and you won't get push back.

 

[TheCusian]

Don't move goal posts and you won't get push back.

if you acknowledge that you're moving them and say you were mistaken - I don't see why you'd be held to the original line. Plus the intent still holds.

Things are better than they have been. And that's good.

 

[IthacaMatt]

I've seen his videos... I came away thinking he's a good talent but that the competition was like watching t
CNY high school athletes. DC2 was rated very high but weak competition means it's a lot harder to evaluate players that stand out.

So was Mike Hart.