Class of 2016 Interesting read for those anti-star ppl

[OttoinGrotto]

I agree 1000x. I stay away until Wed if we lose. Too many yahoos use the site as a place to vent.

Eh, for the standard poster I don't think it's quote the cess pool we joke about it being. There's still good discussion to be had, they just get surrounded by venting.

For anyone directly associated with a player or coach though... yeesh.

 

[CUSU]

That's really pushing it though. When you have a handful of cases on 1 side, and everyone else on the other, I think it's pretty clear they are the exception. Plus if you dig deeper with those cases you can see why. Like I said: All very good coaches, very good systems, and very smart scheduling. If you are a math guy it's pretty hard to argue.

I actually am a Math guy and I would be interested in looking at the data further. If someone could actually get the data I would look at it and break it down.

One thing I think you are forgetting is that a lot of these 5 and 4 star teams also have very good coaches and systems, it isn't just the exceptions. If you ranked the coaches and ran the same tests would that also correlate the same way over time? I would think yes.

Does the 5 star team have better coaching than the 3 star team they beat? Most likely, it isn't only the teams star rating. Some of the middling teams coaches and schemes are not as good as the perennial powers. There are many factors to this equation and I don't think this breakdown of the numbers is definitive by any means. It tells a story at the team level, but not at the individual level.

I think that's the thing a few people are disagreeing on is team ranking vs. individual ranking. Do the team rankings correlate ? Yes. The bigger question I have is it that top rated 5 star in their class that contributed most to that or the 3 star guy? If I could see that data I think it would make or break the analysis for me.

Bam and Bcubs are looking at it more at the individual level which I would say that this article fails to address. This is purely the overall team ranking.

 

[OttoinGrotto]

I actually am a Math guy and I would be interested in looking at the data further. If someone could actually get the data I would look at it and break it down.

One thing I think you are forgetting is that a lot of these 5 and 4 star teams also have very good coaches and systems, it isn't just the exceptions. If you ranked the coaches and ran the same tests would that also correlate the same way over time? I would think yes.

Does the 5 star team have better coaching than the 3 star team they beat? Most likely, it isn't only the teams star rating. Some of the middling teams coaches and schemes are not as good as the perennial powers. There are many factors to this equation and I don't think this breakdown of the numbers is definitive by any means. It tells a story at the team level, but not at the individual level.

I think that's the thing a few people are disagreeing on is team ranking vs. individual ranking. Do the team rankings correlate ? Yes. The bigger question I have is it that top rated 5 star in their class that contributed most to that or the 3 star guy? If I could see that data I think it would make or break the analysis for me.

Bam and Bcubs are looking at it more at the individual level which I would say that this article fails to address. This is purely the overall team ranking.

The article is pretty clear stating that individual rankings are subject to a lot of error.

 

[anomander]

I actually am a Math guy and I would be interested in looking at the data further. If someone could actually get the data I would look at it and break it down.

One thing I think you are forgetting is that a lot of these 5 and 4 star teams also have very good coaches and systems, it isn't just the exceptions. If you ranked the coaches and ran the same tests would that also correlate the same way over time? I would think yes.

Does the 5 star team have better coaching than the 3 star team they beat? Most likely, it isn't only the teams star rating. Some of the middling teams coaches and schemes are not as good as the perennial powers. There are many factors to this equation and I don't think this breakdown of the numbers is definitive by any means. It tells a story at the team level, but not at the individual level.

I think that's the thing a few people are disagreeing on is team ranking vs. individual ranking. Do the team rankings correlate ? Yes. The bigger question I have is it that top rated 5 star in their class that contributed most to that or the 3 star guy? If I could see that data I think it would make or break the analysis for me.

Bam and Bcubs are looking at it more at the individual level which I would say that this article fails to address. This is purely the overall team ranking.

Hey i'll be the first to admit there are mistakes at the individual level. That's inevitable, but all I care about is team performance. But even if your disregard teams 1 - 28 (group 1 +2) in the national recruiting rankings, the numbers still work with teams 29-47 having a convincing record vs teams 48-75, and so on. That basically takes most of the 4 and 5 stars out of play, since most of those, except for a scattered few would be on the top 25 teams.

I mean look at that bottom group (teams 63-75 national ranking). Boise St, Boston College, UCF, UConn, Duke, Iowa St, Kansas St, Memphis, SU, SMU, Temple, and Wake Forest. Really only Kansas St, and Boise St played well above their recruiting ranking, and both of those programs have a well established identity, and really soft scheduling. When it comes to outliers you literally can identify why each and everyone would be considered unique. What I can't understand is just what people are trying to argue? Is it that they make mistakes on an individual level? Sure I admit that. But when it comes to overall team ranking I don't see an argument? Certain players are rated higher because they get an offer from an elite? Ok, well maybe that player should be rated higher then. But if we are trying to argue that programs like Uconn, Iowa St, SMU, and even Syracuse should have players rated higher tell me why? I know nobody is going to try and make the argument that coaching held them back, which could be a very reasonable reason that teams perform under there ranking.

 

[anomander]

The article is pretty clear stating that individual rankings are subject to a lot of error.

Exactly. Nobody is disputing that. But if we are comparing winning % to overall team recruiting ranking, this article is pretty bulletproof.

 

[007]

I actually am a Math guy and I would be interested in looking at the data further. If someone could actually get the data I would look at it and break it down.

One thing I think you are forgetting is that a lot of these 5 and 4 star teams also have very good coaches and systems, it isn't just the exceptions. If you ranked the coaches and ran the same tests would that also correlate the same way over time? I would think yes.

Does the 5 star team have better coaching than the 3 star team they beat? Most likely, it isn't only the teams star rating. Some of the middling teams coaches and schemes are not as good as the perennial powers. There are many factors to this equation and I don't think this breakdown of the numbers is definitive by any means. It tells a story at the team level, but not at the individual level.

I think that's the thing a few people are disagreeing on is team ranking vs. individual ranking. Do the team rankings correlate ? Yes. The bigger question I have is it that top rated 5 star in their class that contributed most to that or the 3 star guy? If I could see that data I think it would make or break the analysis for me.

Bam and Bcubs are looking at it more at the individual level which I would say that this article fails to address. This is purely the overall team ranking.

If you are a "math guy" and you also understand the concept of statistical significance, standard deviation, and the law of large numbers, here is a thesis research study that recorded every team in the FBS and their complete recruiting class for 12 years. Specifically, they recorded the number of 5, 4, 3, and 2 star recruits for each year for every FBS school. They also recorded wins, losses, conference championships, bowl appearances, conference standings at the end of the season, and conference wins and losses for each team. The data was obtained from The ESPN Encyclopedia of College Football and online resources such as athletic department websites. Even when controlling for between school heterogeneity within the same conference (too complex to explain here, but they controlled for some of the variables you mentioned), the results were still statistically and economically significant.

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

 

[OttoinGrotto]

If you are a "math guy" and you also understand the concept of statistical significance, standard deviation, and the law of large numbers, here is a thesis research study that recorded every team in the FBS and their complete recruiting class for 12 years. Specifically, they recorded the number of 5, 4, 3, and 2 star recruits for each year for every FBS school. They also recorded wins, losses, conference championships, bowl appearances, conference standings at the end of the season, and conference wins and losses for each team. The data was obtained from The ESPN Encyclopedia of College Football and online resources such as athletic department websites. Even when controlling for between school heterogeneity within the same conference (too complex to explain here, but they controlled for some of the variables you mentioned), the results were still statistically and economically significant.

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

LIAR!

 

[TheCusian]

Eh, for the standard poster I don't think it's quote the cess pool we joke about it being. There's still good discussion to be had, they just get surrounded by venting. For anyone directly associated with a player or coach though... yeesh.

Stockholm Syndrome?

 

[007]

LIAR!

Huh...I am a liar, I always lie, in fact, I'm lying now.

 

[CUSU]

If you are a "math guy" and you also understand the concept of statistical significance, standard deviation, and the law of large numbers, here is a thesis research study that recorded every team in the FBS and their complete recruiting class for 12 years. Specifically, they recorded the number of 5, 4, 3, and 2 star recruits for each year for every FBS school. They also recorded wins, losses, conference championships, bowl appearances, conference standings at the end of the season, and conference wins and losses for each team. The data was obtained from The ESPN Encyclopedia of College Football and online resources such as athletic department websites. Even when controlling for between school heterogeneity within the same conference (too complex to explain here, but they controlled for some of the variables you mentioned), the results were still statistically and economically significant.

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

The "math guy" comment was a response to ano, wasn't trying to come off any certain way. I appreciate the link, was an interesting read. It is much more telling than the article from SB nation IMO.

 

[anomander]

The "math guy" comment was a response to ano, wasn't trying to come off any certain way. I appreciate the link, was an interesting read. It is much more telling than the article from SB nation IMO.

But basically the same thesis.

 

[CUSU]

But basically the same thesis.

Similar yes, but this one is more in depth and addresses some of my questions.

 

[longtimefan]

If you really get our coaches in a locked room with truth serum they would want almost an entirely better recruiting class down to the man than we are currently getting.

The good news is that the trend appears to be a positive one.

 

[tep624]

Rivals describes a player having a numerical rating of 5.0-5.4 as the following : Division I prospect; considered a mid-major prospect; deemed to have limited pro potential but definite Division I prospect; may be more of a role player

Now here is a list of players that were recruited into the program in the 2008-2010 classes that were "mid major prospects with limited pro potential". And I only picked players that were really good for us. There are others that you could make the argument for. And granted there are stinkers too like 4 star Romale Tucker who was the beneficiary of Rivals bizarre way of overrating Prep prospects.

2008
Antwon Bailey - 5.1 - All BE 1st team, 1,000 yard rusher as a senior
Chandler Jones - 5.2 - 1st round draft pick
Ryan Nassib - 5.2 - 4th round draft pick

2009
Alec Lemon - 5.3 - All BE 1st team, 1,000 yard receiver and Sr Bowl invitee as a senior
Justin Pugh - 5.2 - 1st round draft pick
Shamarko Thomas - 5.1 - 4th round draft pick

2010
Sean Hickey - 5.2 - All ACC 3rd team as a senior, likely would have been drafted if not for injuries
Macky Macpherson - 5.3 - All ACC 2nd team as a senior
Dyshawn Davis - 5.3 - 3 yr starter
Tyson Gulley - 5.4 - 2,000 career rushing yards, 7th in BE as a Soph

That's 4 NFL draft picks including 2 first rounders they completely whiffed on. That's enough for me to question the accuracy of the system when it comes to specifically evaluating our recruits.

My argument has always been very simple - Rivals does a poor job of evaluating our recruits because they don't pay any attention to our recruits.

 

[JrodCuse]

How is our recruiting trending over the past 15 years?

 

[PhatOrange]

Offers are often BS.

i agree. But i think it's more clear these days who actually wants a kid. mutliple visits, officials, coaches in schools, golf cart treatments etc.

 

[PhatOrange]

We have a whole bunch of data that says you are dead wrong.

Everyone chimes in saying how stars are meaningless, but nobody has any data to prove otherwise, other then 1 or 2 examples of the extreme.

Here’s your data. I’ve done this exercise before and I don’t have time today to go through it all again. But look at these 3 teams. Record the last 5 years followed by recruiting ranking (rovals). Star rankings don’t jive with their expected records.

"All that matters is those final team rankings are a pretty solid indicator of win %, and that's indisputable."

Really, indisputable?

Texas
2014 / 6-7 - #20
8-5 - #2
9-4 - #3
8-5 - #3
2009 / 5-7 - #5

Florida
2014 / 7-5 - #8
4-8 - #4
11-2 - #3
7-6 - #12
2010 / 8-5 - #2

Michigan
2014 / 5-7 - #31
7-6 - #5
8-5 - #7
11-2 - #21
7-6 - #20
2009 5-7 - #8


The above teams are not outliers either. Check these teams and there are a lot more

Tennessee
Miami
Arkansas
Oklahoma
Mississippi

 

[anomander]

Here’s your data. I’ve done this exercise before and I don’t have time today to go through it all again. But look at these 3 teams. Record the last 5 years followed by recruiting ranking (rovals). Star rankings don’t jive with their expected records. Schools like Syracuse don’t sniff the top 20 recruiting classes but their records are about the same except the two 11-2 anomalies. "All that matters is those final team rankings are a pretty solid indicator of win %, and that's indisputable." Really? Texas 2014 / 6-7 - #20 8-5 - #2 9-4 - #3 8-5 - #3 2009 / 5-7 - #5 Florida 2014 / 7-5 - #8 4-8 - #4 11-2 - #3 7-6 - #12 2010 / 8-5 - #2 Michigan 2014 / 5-7 - #31 7-6 - #5 8-5 - #7 11-2 - #21 7-6 - #20 2009 5-7 - #8 The above teams are not outliers either. Check these teams and there are a lot more Tennessee Miami Arkansas Oklahoma Mississippi

Ya and I think everyone can pick out the reason those 3 teams have struggled for that time period. You make it sound like Syracuse has more then 3 winning seasons in the last 15 years and are getting screwed in their recruiting ranking, when in fact they are probably finishing right where they should.

 

[Louie and Bouie]

Ya and I think everyone can pick out the reason those 3 teams have struggled for that time period. You make it sound like Syracuse has more then 3 winning seasons in the last 15 years and are getting screwed in their recruiting ranking, when in fact they are probably finishing right where they should.

For the sake of factual accuracy Syracuse has had 5 winning seasons (2000, 2001, 2010, 2012 and 2013) in the last 15 seasons and 2 years at .500 during that same time span.

 

[007]

Here’s your data. I’ve done this exercise before and I don’t have time today to go through it all again. But look at these 3 teams. Record the last 5 years followed by recruiting ranking (rovals). Star rankings don’t jive with their expected records.

"All that matters is those final team rankings are a pretty solid indicator of win %, and that's indisputable."

Really, indisputable?

Texas
2014 / 6-7 - #20
8-5 - #2
9-4 - #3
8-5 - #3
2009 / 5-7 - #5

Florida
2014 / 7-5 - #8
4-8 - #4
11-2 - #3
7-6 - #12
2010 / 8-5 - #2

Michigan
2014 / 5-7 - #31
7-6 - #5
8-5 - #7
11-2 - #21
7-6 - #20
2009 5-7 - #8


The above teams are not outliers either. Check these teams and there are a lot more

Tennessee
Miami
Arkansas
Oklahoma
Mississippi

You looked at those 8, 9, 10 (whatever the number) teams over the course of 5 years and conclude that the correlation between rankings and winning percentage does not "jive."

The study I linked looked at EVERY FBS team (including the 8 you listed) over the course of 12 years and concluded that the correlation between rankings and winning is both meaningful and statistically significant regardless of the comparison group (they even correlated the "economic" impact of having higher rated classes). This was true even when comparisons were made within conferences - meaning the study analyzed the effect within any given conference for teams who were recruiting the same players. The teams with the higher rated classes (regardless of whether you agree with the rankings or not) consistently had a higher head to head winning percentage over time.

This is not an opinion. It is a statistical conclusion of significance based on analysis of 12 years of data. The other thing to keep in mind is that, although statistical significance does prove that there is a strong correlation between any two variables, it is not an absolute. That means that the occurrence of outcomes that don't "jive" with the correlation are not particularly meaningful. In fact, they are expected to occur. For example, the fact that there are thousands of people who smoke more than a pack of cigarettes a day and never get cancer does NOT disprove the correlation between smoking and cancer.

 

[longtimefan]

How is our recruiting trending over the past 15 years?

If you're asking me, I was referring to the Shaffer Era.

 

[TheCusian]

Seems like the takeaway is that there is a hierarchy based on talent (this study), coaches, money, facilities, conference affiliation, focus, and fan support (shocker!). Plucking the recruiting numbers and comparing them against w/l is one component in the mix of relative success or failure - but it's not the only one.

So then the questions become: How important is talent in the equation (60%, 80%, etc.)? And is there a correlation between the outliers (both positive and negative examples) and the other variables (like Hoke as your coach)?

And as it applies to us: how do we rise up a tier? How do we continue the upward trend in recruiting?

 

[007]

Seems like the takeaway is that there is a hierarchy based on talent (this study), coaches, money, facilities, conference affiliation, focus, and fan support (shocker!). Plucking the recruiting numbers and comparing them against w/l is one component in the mix of relative success or failure - but it's not the only one.

So then the questions become: How important is talent in the equation (60%, 80%, etc.)? And is there a correlation between the outliers (both positive and negative examples) and the other variables (like Hoke as your coach)?

And as it applies to us: how do we rise up a tier? How do we continue the upward trend in recruiting?

I'm not sure if your use of the phrase "plucking the recruiting numbers" was meant to be dismissive or not. In any case, the correlation between class rankings and winning is significant. This is not an opinion. It is a statistical conclusion of significance based on analysis of 12 years of data.

So knowing that, you have a threshold or baseline from which you can look at the importance of other factors you mentioned, which are much more difficult to quantify. The expectation is that higher class rankings will result in higher winning percentages over time. In the short-term, it's reasonable to assume that coaching, resources, and money will impact those outcomes, either positively and negatively. It would not be statistically unusual to find some coaches who consistently outperform class rankings (btw, those are the guys you want to look for). But again, the occurrence of those outcomes (either positive or negative) do NOT disprove the correlation between rankings and winning. What it probably does "prove" is that those coaches who consistently outperform their team talent rankings are highly valuable (where value = winning).

What can SU do to try to win more? Without question, the option with the highest probability of success is to recruit and sign higher rated players. Short of that, the next best option, IMO anyway, is to make sure you have a HC who is consistently outperforming the rankings of the talent he/the school is getting.

 

[tep624]

I'm not sure if your use of the phrase "plucking the recruiting numbers" was meant to be dismissive or not. In any case, the correlation between class rankings and winning is significant. This is not an opinion. It is a statistical conclusion of significance based on analysis of 12 years of data.

So knowing that, you have a threshold or baseline from which you can look at the importance of other factors you mentioned, which are much more difficult to quantify. The expectation is that higher class rankings will result in higher winning percentages over time. In the short-term, it's reasonable to assume that coaching, resources, and money will impact those outcomes, either positively and negatively. It would not be statistically unusual to find some coaches who consistently outperform class rankings (btw, those are the guys you want to look for). But again, the occurrence of those outcomes (either positive or negative) do NOT disprove the correlation between rankings and winning. What it probably does "prove" is that those coaches who consistently outperform their team talent rankings are highly valuable (where value = winning).

What can SU do to try to win more? Without question, the option with the highest probability of success is to recruit and sign higher rated players. Short of that, the next best option, IMO anyway, is to make sure you have a HC who is consistently outperforming the rankings of the talent he/the school is getting.

But couldn't this analysis simply be proving that successful teams get their classes ranked higher?

The measure of a recruiting class's impact is typically 2-4 years after they are recruited, so that recruiting class literally has 0 impact on the current state of the team and its success, or lack there of.The analysis you reference compares recruiting classes from 2002-2012 and team performance during the same period. I would argue that at least a third of that data set is skewed because 2002-2003 is generally going to be impacted on the field by recruits from the 1999-2001 classes and the 2011-2012 classes likely have little, if any impact on the 2011 and 2012 seasons. And it's even worse for schools that generally don't recruit plug and play blue chip players.

What they should have done is compare the 2002-2012 classes to the 2004-2014 seasons, at a minimum, for the power programs. And the scale probably slides even more for schools that recruit the 2 star types that need more time to develop.

 

[TheCusian]

I'm not sure if your use of the phrase "plucking the recruiting numbers" was meant to be dismissive or not. In any case, the correlation between class rankings and winning is significant. This is not an opinion. It is a statistical conclusion of significance based on analysis of 12 years of data. So knowing that, you have a threshold or baseline from which you can look at the importance of other factors you mentioned, which are much more difficult to quantify. The expectation is that higher class rankings will result in higher winning percentages over time. In the short-term, it's reasonable to assume that coaching, resources, and money will impact those outcomes, either positively and negatively. It would not be statistically unusual to find some coaches who consistently outperform class rankings (btw, those are the guys you want to look for). But again, the occurrence of those outcomes (either positive or negative) do NOT disprove the correlation between rankings and winning. What it probably does "prove" is that those coaches who consistently outperform their team talent rankings are highly valuable (where value = winning). What can SU do to try to win more? Without question, the option with the highest probability of success is to recruit and sign higher rated players. Short of that, the next best option, IMO anyway, is to make sure you have a HC who is consistently outperforming the rankings of the talent he/the school is getting.

No - I've been consistent in this. I think those numbers are solid. There is a correlation.

I think your point on it being a baseline was kind of what I was getting at. You said it better .

And obviously the other variables all add to and re-enforce the recruiting numbers.

Some of Coyles comment were very encouraging re:recruiting...