-
DJNiteFlite said...
If the data is flawed, how is Rivals able to predict a full third of the best players in the country 4 years before it happens? Before their senior prom? Before they can grow facial hair? They're sophomores right now. Many of them have only started one year at the high school level. Yet and still the rankings find 30% of the best players before they even get to college.... but they're flawed. Yeah keep dreaming buddy.
Oh... and if they're flawed and Dantonio is so better than them at identifying talent, where are Dantonio's post-season honorees to back it up. When is the last time that Dantonio scouted a class and had 30% of his guys end up the best in the nation?
Dr Draymond
- 5 stars Rating: 96
863 votes total - Go Green
- (1542)
- 29 months
- Send Message
- Follow User
- Ignore User
- 5 stars
-
RPMadMSU
- 5 stars Rating: 93
9938 votes total - Eternal Bastion of Common Sense
- (16680)
- 29 months
- Send Message
- Follow User
- Ignore User
- 5 stars
-
fishrose said...
That article is riddled with false conclusions and poor methodology. If the recruiting services are ranking 1,800+ players as four stars, and those 1,800+ players only end up accounting for 34% of the 80-90 players named to All-American teams, that is NOT a good rate of success for the recruiting services. It's absolutely irrelevant that 1 in 53 four-stars were selected to All-American teams versus 1 in 173 three-stars over 2006-10. That is essentially guaranteed by the sample size of each group. The only thing that this data demonstrates is that recruiting services are incredibly inaccurate at deciding whether a player should be given three or four stars when the player isn't obviously great or obviously mediocre.
The author claims that rankings are accurate when viewed as an "investment" or a "projection of how likely a player is of becoming an elite contributor compared to rest of the field." However, that is not what the services are doing. They're selecting individual players and ranking them, specifically stating that individual player X is better than individual player Y. Similarly, coaches aren't recruiting every single four-star in each class, nor are they recruiting every single three-star in every class. They're recruiting individual players, and the author flat-out admits that these large-scale statistics are incredibly inaccurate on the level of individual assessment.
This article is all spin, which should really be expected since it was published by one of the largest and most profitable recruiting services in the business. If anything, this data suggests that far too many players are given four-star ratings for that "honor" to mean anything. If they really wanted to accurately predict which players will end up on All-American teams or as stars in the NFL, they would give at maximum 75-100 players per year a rating of four stars or better, and raise the threshold between three-stars and two-stars. Of course, there's no money in doing that, because every large fanbase wants to think it has 10-15 of the best high school players in the country. There just aren't that many "sure things" in the high school game, but those fans are willing to pay lots of money to be lied to.
-
DJNiteFlite said...
No I have never said anything of the sort. Urban Meyer and Nick Saban evaluate talent on their own just like most other great coaches. My only point is that it is not a mistake that the guys Urban and Saban think are good are always heavily featured on Rivals lists. It's because Rivals AND Meyer and Saban are good at evaluating talent. Like I've been saying... the rankings are predictors.
A faulty ranking system wouldn't so often agree with the best programs and coaches in the nation.
mentalstate
- 5 stars Rating: 86
8650 votes total - (7703)
- 29 months
- Send Message
- Follow User
- Ignore User
-
fishrose said...
It baffles me that people can't grasp the basic statistics behind what Rivals is trying to do. They're inflating the ratings of hundreds of players every year, and then hedging their bets statistically by inflating the rankings even further on the bottom end. It's absolute genius from a business standpoint; they know they don't have (and can't afford to pay) enough scouts and analysts to accurate rank all the high school football players in the country. They also know that they don't have to, because they can rig the numbers to look like they're making accurate predictions and fans will buy it (no pun intended) because they're being told what they want to hear.
RPMadMSU
- 5 stars Rating: 93
9938 votes total - Eternal Bastion of Common Sense
- (16680)
- 29 months
- Send Message
- Follow User
- Ignore User
- 5 stars
-
Emerald Green
- 5 stars Rating: 95
421 votes total - (370)
- 27 months
- Send Message
- Follow User
- Ignore User
- 5 stars
-
RPMadMSU
- 5 stars Rating: 93
9938 votes total - Eternal Bastion of Common Sense
- (16680)
- 29 months
- Send Message
- Follow User
- Ignore User
- 5 stars
-
fishrose said...
That article is riddled with false conclusions and poor methodology. If the recruiting services are ranking 1,800+ players as four stars, and those 1,800+ players only end up accounting for 34% of the 80-90 players named to All-American teams, that is NOT a good rate of success for the recruiting services. It's absolutely irrelevant that 1 in 53 four-stars were selected to All-American teams versus 1 in 173 three-stars over 2006-10. That is essentially guaranteed by the sample size of each group. The only thing that this data demonstrates is that recruiting services are incredibly inaccurate at deciding whether a player should be given three or four stars when the player isn't obviously great or obviously mediocre.
The author claims that rankings are accurate when viewed as an "investment" or a "projection of how likely a player is of becoming an elite contributor compared to rest of the field." However, that is not what the services are doing. They're selecting individual players and ranking them, specifically stating that individual player X is better than individual player Y. Similarly, coaches aren't recruiting every single four-star in each class, nor are they recruiting every single three-star in every class. They're recruiting individual players, and the author flat-out admits that these large-scale statistics are incredibly inaccurate on the level of individual assessment.
This article is all spin, which should really be expected since it was published by one of the largest and most profitable recruiting services in the business. If anything, this data suggests that far too many players are given four-star ratings for that "honor" to mean anything. If they really wanted to accurately predict which players will end up on All-American teams or as stars in the NFL, they would give at maximum 75-100 players per year a rating of four stars or better, and raise the threshold between three-stars and two-stars. Of course, there's no money in doing that, because every large fanbase wants to think it has 10-15 of the best high school players in the country. There just aren't that many "sure things" in the high school game, but those fans are willing to pay lots of money to be lied to.
I want to make it known that Steve Lorenz (Tremendous) is the reason my trial month became a permanent membership. Smart move 247.
-
mentalstate said...
Another post that proves you don't know too much: meyer has admitted a lot of their recruiting strategy is to see who Saban and Spurrier offer, and then offer them, as he is not a great HS talent evaluator. As for Saban, who is...Dantonio studied under him.
And your little side fight has allowed you to skate away from the post I did pointing out you linked an article from a year ago to back up your false claim of Dantonio not being good at developing talent....you talk about not people understanding statistics, then leave out 25% of the data you're arguing about.
This post was edited by RPMadMSU on 5/15/2012 at 1:32 PM
RPMadMSU
- 5 stars Rating: 93
9938 votes total - Eternal Bastion of Common Sense
- (16680)
- 29 months
- Send Message
- Follow User
- Ignore User
- 5 stars
-
mentalstate said...
Another post that proves you don't know too much: meyer has admitted a lot of their recruiting strategy is to see who Saban and Spurrier offer, and then offer them, as he is not a great HS talent evaluator. As for Saban, who is...Dantonio studied under him.
And your little side fight has allowed you to skate away from the post I did pointing out you linked an article from a year ago to back up your false claim of Dantonio not being good at developing talent....you talk about not people understanding statistics, then leave out 25% of the data you're arguing about.
This post was edited by DJNiteFlite on 5/15/2012 at 1:31 PM
I want to make it known that Steve Lorenz (Tremendous) is the reason my trial month became a permanent membership. Smart move 247.
-
Emerald Green said...
Wow, and here I thought that Mark Dantonio and his 40th ranked recruiting classes had exceeded expectations, only to come here and find out that MSU underachieves more than most of the programs in the Big Ten!!!!!!!!!!
Whatever shall we do!?!?!?!
I want to make it known that Steve Lorenz (Tremendous) is the reason my trial month became a permanent membership. Smart move 247.
-
RPMadMSU
- 5 stars Rating: 93
9938 votes total - Eternal Bastion of Common Sense
- (16680)
- 29 months
- Send Message
- Follow User
- Ignore User
- 5 stars
-
RPMadMSU
- 5 stars Rating: 93
9938 votes total - Eternal Bastion of Common Sense
- (16680)
- 29 months
- Send Message
- Follow User
- Ignore User
- 5 stars
-
DJNiteFlite said...
Nice.... you overlook the fact that the article you're citing was about developing nfl talent, not winning ball games. Oops.... good try though little bro.
I'm not the one spending time on another teams website trying to prove to them how poor their program is at developing talent.
For a team that is a distant 4th rival and that they don't even care about, MGOBLOG sure does spend a lot of time figuring out ways to marginalize Mark Dantonio.
As far as little bro, in light of current events...that is about as useful as an insult as calling that other team Ohio.
Emerald Green
- 5 stars Rating: 95
421 votes total - (370)
- 27 months
- Send Message
- Follow User
- Ignore User
- 5 stars
-
Emerald Green said...
I'm not the one spending time on another teams website trying to prove to them how poor their program is at developing talent.
For a team that is a distant 4th rival and that they don't even care about, MGOBLOG sure does spend a lot of time figuring out ways to marginalize Mark Dantonio.
As far as little bro, in light of current events...that is about as useful as an insult as calling that other team Ohio.
-
DJNiteFlite said...
In the second paragraph, you point out that the rankings aren't intended to do what the author finds them accurate in predicting. My question is so what? You don't dispute that they are relatively accurate when regarded that way, right? Finally, if you're arguing that Matt Hinton was biased, I'm done. There's really nothing more to be said.
This post was edited by fishrose on 5/15/2012 at 1:58 PM
tRCMB's resident Wayne State Warrior and Sam's Club Spartan fan.
-
Spartan8Ball
- 4 stars Rating: 62
5630 votes total - Baby Ruth?
- (5701)
- 20 months
- Send Message
- Follow User
- Ignore User
- 4 stars
-
Johnny2x2x said...
Excellent post!
Furthermore, why have 5 stars, what an inaccurate way to create separation between players. How about 10 stars? This would require the services to actually have a clue about players rather than just look at their offers and make safe bets. Hmm, Rodney Jones is going to Florida to play DE, Florida has a great d Line coach and they have an extreme need at DE so he'll get a chance to play right away, Rodney Jones should get a 4th star and then I'll look smart when he makes all SEC his 3rd year. Was Rodney Jones any better than the 3 star he jumped who went to Louisville? Doubt it.
Michigan State does not and will not run the 3-4 defense.
SpartanRocky
- 5 stars Rating: 90
16743 votes total - Andrew Bell
- (22363)
- 29 months
- Send Message
- Follow User
- Ignore User
-
Spartan8Ball said...
SCum will be top 5 preseason. One reason for that is to build hype around the Bama game for ratings. College football and the conferences have sold their sole to television/rating numbers. Preseason polls are pure garbage. Polls shouldn't come out until week 7 or 8.
Michigan State does not and will not run the 3-4 defense.
SpartanRocky
- 5 stars Rating: 90
16743 votes total - Andrew Bell
- (22363)
- 29 months
- Send Message
- Follow User
- Ignore User
-
SpartanRocky said...
Actually, I'm a fan of the 247 rankings (100, 99, 98, etc.) because they provide a better degree of specificity than the general "5-star, 4-star, etc."
I look at it another way though; a certain # of kids are going be college stars. Those are your All-Conference and All-American performers. It doesn't matter what they were rated out of HS; if a kid has success, he's going to receive some sort of recognition.
I also believe that a higher % of 5-stars work out than do 4-stars, and a higher % of 4-stars work out than 3-stars do. That's usually where the analysis stops for the recruiting fanatics who believe stars are everything. [Note: I consider "working out" to mean that the guy plays college FB at a high level; he's a starter, usually for more than 1 year and receives some sort of all-conf. recognition. That doesn't have to be Greg Jones-recognition (multi-year AA); a guy like Trenton Robinson (2nd Team All B1G as a JR, 1st Team as a SR, 2.5 year starter, etc.) would qualify.]
Here's what those people don't stop to think about often, and why recruiting championships don't always translate into national titles: if the % of a group of kids that works out is anything less than 100%, you're going to have 5-stars/4-stars, etc. that do NOT work out. Let's say, just for argument's sake, that 4-star prospects work out 70% of the time, and that there are 300 4-star prospects in each class. That's 210 players who are going to be making an impact on some college team . . . and 90 that will not.
Think about that. 90 kids is enough to fill 4 22 or 23 man recruiting classes EXCLUSIVELY with kids that are busts. Keep that in mind.
Now, let's say there are 700 kids in a class who are called "3-star" prospects, and say that those guys work out at a 50% rate. That means that 350 of those players will not work out at a college level . . . but the other 350 will.
Here's where the star system fails at predicting college success: You cannot tell which 4-stars are part of the 210 "working out" group and which are in the 90-man "bust" group, or which group of 350 your class of 3-stars belongs to.
Obviously the successes and busts are spread-out amongst the classes and the country, but they aren't evenly distributed; one program's class of 10 4-stars and 10 3-stars gets 15 total kids to work out from that class, 8 3-stars and 7 3-stars, higher than the average, while another class of 10 4-stars and 10 3-stars will have 4 of those 4-stars and 4 of the 3-stars work out, lower than average.
This is exactly why a program like Virginia Tech can consistently win 10+ games/year and not have a plethora of 4-star recruits, and why teams like Texas, Penn State, Michigan, Notre Dame, FSU and Miami(Fl) have suffered downturns despite recruiting a consistently high # of 4-star prospects: Until the prospects take the field, there is no way to determine which guys are going to work out and which ones will not. A class comprised of 20 of the 350 "successful" 3-stars and 3 of the 210 successful 4-stars is going to do much better in CFB than one that only has 4 of the successful 4-stars . . . but 10 of the 4-star "busts", even though Class #1 will be ranked far behind Class #2.
In short, bringing in highly rated recruits hedges the bet in favor of your team, as the averages favor 4-stars over 3-stars, etc. It does no more than that, however, because the ability of the prospects is independent of the ratings assigned to them by recruiting services. That's why these matters are always best resolved on the field.
mentalstate
- 5 stars Rating: 86
8650 votes total - (7703)
- 29 months
- Send Message
- Follow User
- Ignore User
-
Spartan8Ball
- 4 stars Rating: 62
5630 votes total - Baby Ruth?
- (5701)
- 20 months
- Send Message
- Follow User
- Ignore User
- 4 stars
- Post a New Topic
- Back to Topics
- « Previous Topic
- Next Topic »
- Boards ▾
- Pages: 1 | 2 | 3 | 4 | 5
IF MSU has a different name what are they polled preseason?