Do the Recruiting Stars Matter, 2023 Draft

[LouderVol]

I still haven't figured out where I messed up the numbers in the cumulative, and its got me reconsidering doing the rest of the draft. as I will just be making the numbers less and less accurate if I continue to work off of my own flawed numbers. No one probably actually cares, but apologies everyone. I may still try to go through and at least post all of the players and their rankings. we will see if I have the time.

 

[butchna]

Yes. There’s examples of games Saban has lost too. That doesn’t mean he’s bad at his job. 81% is really, really good.

“I can name guys they missed”-proves nothing. I can name guys Fulmer missed, Saban missed, etc. No model nor person can accurately predict the ability of a guy like Kelce to put on 60lbs and move to center

I DISREGARD contrived percentages…they’re really really boring.

Zackly! Fulmer was an elite recruiter and closer whose star went the other way when he got lazy, stopped utilizing his high school resources and went with the services. That’s why he netted highly ranked classes with names like Brooks and Donald…amongst others. Saban is 5 deep with 5 stars yet needs walk-ons like Levi Wallace and 3 stars like Josh Jacobs to help him win and was desperate for 3 star Tyler Steen to portal from VANDY to save his OL last season. Rankings matter.

Kelce and his brother BOTH are proof of how the nerd services can’t identify growth potential in athletes beyond that of the shiniest ones that get invited to their camps. Another example is 2 star Aaron Rodgers, putting up numbers in California. Doesn’t get invited to camps and doesn’t get offers from any four year institution. Went to Juco, shone at Berkeley and got drafted in the first round!. Percentages again…wonder how his story unfolded?

 

[Vol8188]

I DISREGARD contrived percentages…they’re really really boring.

Zackly! Fulmer was an elite recruiter and closer whose star went the other way when he got lazy, stopped utilizing his high school resources and went with the services. That’s why he netted highly ranked classes with names like Brooks and Donald…amongst others. Saban is 5 deep with 5 stars yet needs walk-ons like Levi Wallace and 3 stars like Josh Jacobs to help him win and was desperate for 3 star Tyler Steen to portal from VANDY to save his OL last season. Rankings matter.

Kelce and his brother BOTH are proof of how the nerd services can’t identify growth potential in athletes beyond that of the shiniest ones that get invited to their camps. Another example is 2 star Aaron Rodgers, putting up numbers in California. Doesn’t get invited to camps and doesn’t get offers from any four year institution. Went to Juco, shone at Berkeley and got drafted in the first round!. Percentages again…wonder how his story unfolded?

What does that prove in your mind? No one is proclaiming the services are perfect. But over 80% is really, really impressive

 

[butchna]

What does that prove in your mind? No one is proclaiming the services are perfect. But over 80% is really, really impressive

You’re closed off and drinkin that Kool-Aid. Embrace your percentages, they’ll never hurt you…like 8th grade did.

 

[Vol8188]

You’re closed off and drinkin that Kool-Aid. Embrace your percentages, they’ll never hurt you…like 8th grade did.

Yes, because % rather than random anecdotes is how we should approach the question. Troll on…

 

[butchna]

Yes, because % rather than random anecdotes is how we should approach the question. Troll on…

Getting drafted in a certain round vs legacy as a PLAYER. Your arguments and percentages fall on their face. I’m stating my takes and opinions. Characterizing that as TROLLING because it contradicts your own highlights your arrogance.

 

[Vol8188]

Getting drafted in a certain round vs legacy as a PLAYER. Your arguments and percentages fall on their face. I’m stating my takes and opinions. Characterizing that as TROLLING because it contradicts your own highlights your arrogance.

Legacy as a player isn’t a measurable attribute. But the biggest issue with that nonsense is that you’re attempting to proclaim the model sucks based off of something the model is not attempting to predict, how will the player will do in the nfl.

 

[butchna]

Legacy as a player isn’t a measurable attribute. But the biggest issue with that nonsense is that you’re attempting to proclaim the model sucks based off of something the model is not attempting to predict, how will the player will do in the nfl.

If we develop our 3-4 stars better than the service favorites 4’s and 5’s and they produce more in the NFL? It’ll mean more. And that’s not an “attempt”. If they’re not “attempting” to predict the best players you can bring into your problem then I have been wrong. I’ve always felt they’re a useful tool. That would prove they’re USELESS.

 

[Vol8188]

If we develop our 3-4 stars better than the service favorites 4’s and 5’s and they produce more in the NFL? It’ll mean more. And that’s not an “attempt”. If they’re not “attempting” to predict the best players you can bring into your problem then I have been wrong. I’ve always felt they’re a useful tool. That would prove they’re USELESS.

Obviously you’re attempting to find the best to bring into your program. And obviously the models do that. What’re you even disagreeing with the model on?

Is your entire argument that “it’s theoretically possibly, although highly unlikely, to find all the 3* guys who are underrated and win at a high level with them”?

 

[butchna]

Obviously you’re attempting to find the best to bring into your program. And obviously the models do that. What’re you even disagreeing with the model on?

Is your entire argument that “it’s theoretically possibly, although highly unlikely, to find all the 3* guys who are underrated and win at a high level with them”?

No…just don’t discount them because the nerds haven’t coronated them. Our coaches pursue them for a reason and not higher rated ones that get to go to camp!

 

[Vol8188]

No…just don’t discount them because the nerds haven’t coronated them. Our coaches pursue them for a reason and not higher rated ones that get to go to camp!

No one is saying that. You’re fighting a strawman. No one is saying “3* players will never amount to anything”. That’s not how this works. What is being said is that higher rated players are more likely to excel than lower rated players.

 

[StoVol]

If a 5* is more talented, how can you proclaim they “arbitrarily picked them”?

It is an important metric, sure. And they still hit over 30% on that specific metric. You don’t think that’s impressive?

We should also look at other metrics, like total drafted.

You claiming there is no arbitrary factors
What about the self fulfilling prophecy butchna suggested?
That the services follow who the coaches pursue?
Then it would be a function of a function
Your variables are starting to get the "sensitive dependence on initial conditions" Lorenz warned us about
I see you innacurately creating a strawman to easily burn down, we are trying to look for truth, not a caricature of it

 

[StoVol]

to me I think the issue is a difference between accuracy and precision. I think overall the recruiting analysts are accurate, these kids are talented; but not very precise, First Round vs 7th round.

the 5* hit rate is pretty dang close to the NFL draft overall success rate. I am not saying its perfect. Just pointing out that the recruiting services do get it right enough where you can see positive trends based on their ratings.

and with all of the variability I don't think looking at the first round only paints a good picture. Yes a 5* ranking means the recruiting guys think the kid is round 1 in 3 to 5 years; but I am not calling it much of a miss if most of the guys they call a 5* get drafted period. Obviously they found kids with enough talent/promise to be drafted. which is a high bar. and then you start playing the percentages of the draft rounds and their accuracy for when the 5* tend to go.

I didnt claim look at 1st round only, just that the services claimed they were 1st rounders in high school and are more wrong than right on that metric

You would need back propagation, complicated models better suited to ai and llm than relying on a dude named 88

Id like to see a function of 5 stars who went in 1st 3 rounds
Adjusted to the percent of overall players

Then the 4 stars, where they plot on a chart by round

Still it wouodnt tell me if they were a 4 star on merit or if on3 gave him 4 stars because saban offered one May afternoon

 

[Vol8188]

You claiming there is no arbitrary factors
What about the self fulfilling prophecy butchna suggested?
That the services follow who the coaches pursue?
Then it would be a function of a function
Your variables are starting to get the "sensitive dependence on initial conditions" Lorenz warned us about
I see you innacurately creating a strawman to easily burn down, we are trying to look for truth, not a caricature of it

If it is arbitrary as you’re claiming, why’s it so successful? You couldn’t randomly draw names and have the level of success the services had this year.

If services were handing out ranking based on who went after them, Bama and Georgia wouldn’t sign 3* players annually. Do services evaluate guys after they get offers? Of course. If a player has an offer Bama and Georgia and 247 hasn’t evaluated, they’re not doing their job

 

[LouderVol]

You claiming there is no arbitrary factors
What about the self fulfilling prophecy butchna suggested?
That the services follow who the coaches pursue?
Then it would be a function of a function
Your variables are starting to get the "sensitive dependence on initial conditions" Lorenz warned us about
I see you innacurately creating a strawman to easily burn down, we are trying to look for truth, not a caricature of it

if the coaches can come around and figure out a player is better than they were the year before, Arion Carter, why don't the recruiting services get the same benefit of the doubt?

again, my whole point for this thread wasn't to say that the recruiting analysts are perfect. it wasn't even to say they were better at their jobs than the NFL or college scouts, or that just because a kid is a 5* means he was "guaranteed" to go in the first round. The point is they are right enough where you can make reasonable assumptions off of the data they generate as a predictive take on the whole. specific individuals bucking the trend doesn't mean the base assumption is wrong. No matter how that base assumption was made.

 

[LouderVol]

Cumulative through 3 rounds.

in the first 326 (blue chip) ranked players you get: 52 players
in the next 326 players you have only: 8 player.
in the third group of 326 players you get: 9 players
in the fourth group of 326 players you get: 7 players
in the fifth group you get: 6 players
in the sixth group you get: 2
in the seventh group you get: 0
in the eighth plus you get: 12

19 5 stars
35 4 stars
36 3 stars
11 2 star/unranked
(this is only 101 I am not sure where I lost someone).

I went back and I think its all coming from Round 3 screwing with me, but I am not sure where I am getting the numbers wrong. I will try to go back and look, and see if I can sort this out. hopefully I figure something out when I get a chance this afternoon/tonight.

@StoVol this was the cumulative for the first three rounds. note there is a mistake somewhere in my counts but I can't figure out where, but it should only be 1 player, so it shouldn't shift the narrative too much. I just didn't want to post additional numbers knowing it was flawed from the mid point on.

I am not sure if you are looking for the 5* as a percentage of all drafted, or a 5* as a percentage of a single class of 5*. but I have the total number of 5* drafted by round 3.

 

[StoVol]

If it is arbitrary as you’re claiming, why’s it so successful? You couldn’t randomly draw names and have the level of success the services had this year.

If services were handing out ranking based on who went after them, Bama and Georgia wouldn’t sign 3* players annually. Do services evaluate guys after they get offers? Of course. If a player has an offer Bama and Georgia and 247 hasn’t evaluated, they’re not doing their job

Is it successful? If they follow the coaches then its arbitrary
They do follow them some
Therefore it is partially arbitrary and one cannot know if rivals followed the coaches more or relied on their own expert evaluations

This is fish food for people who love recruiting and you eat the hell out of it with no understanding of the loop youre in

 

[LouderVol]

Is it successful? If they follow the coaches then its arbitrary
They do follow them some
Therefore it is partially arbitrary and one cannot know if rivals followed the coaches more or relied on their own expert evaluations

This is fish food for people who love recruiting and you eat the hell out of it with no understanding of the loop youre in

isn't that how all of talent evaluations go? They all follow each other. With something as arbitrary as trying to nationally rank high schoolers playing widely different positions you are going to use as many different sources as are available to you.

and they aren't as arbitrary as you make them sound. Or else Bama's and Georgia players would be 1-50. Tennessee's players would fall somewhere around the 500 mark. They apply a lot more variety and nuance than that. Nico rose on most sites last year after committing to us, and ignoring pretty much every school that would justify him having a higher ranking than when he committed to us.

 

[DaddyChad]

Interestingly this started out as a few local instate people ranking all the in state prospects. It evolved into a national service then quite a few national services. It is my belief that these national services start out ranking kids by the guys that evaluate in state talent, rankings are adjusted based on camp performances against other kids, ranking are further adjusted by what schools offer kids, further adjusted by what schools accept kids. Those schools that accept kids sometimes get a larger bump than other schools; that metric is moved dependent on the school's record of putting kids in the league.

So in summation stars do matter but how those stars are arrived at also matter bc you see offers matter more than just stars in and of themselves. Do you legitimately believe a four star that goes to Utah has the same percentile chance of going pro as a four star at Bama or Georgia?

 

[LouderVol]

Interestingly this started out as a few local instate people ranking all the in state prospects. It evolved into a national service then quite a few national services. It is my belief that these national services start out ranking kids by the guys that evaluate in state talent, rankings are adjusted based on camp performances against other kids, ranking are further adjusted by what schools offer kids, further adjusted by what schools accept kids. Those schools that accept kids sometimes get a larger bump than other schools; that metric is moved dependent on the school's record of putting kids in the league.

So in summation stars do matter but how those stars are arrived at also matter bc you see offers matter more than just stars in and of themselves. Do you legitimately believe a four star that goes to Utah has the same percentile chance of going pro as a four star at Bama or Georgia?

Bama and Georgia pretty much only recruit blue chip players, and both had 10 players selected. 10/25=40% (single year results). Utah probably doesn't even get 10 blue chips a class, so yeah I would bet its a pretty similar ratio. Utah got 8 blue chip guys last year, they would only need 3 of those to make the NFL to have the same success rate as Bama/Georgia.

For the record Utah had 4 4* players in 2020. that's a pretty small sample size where you can't even cleanly get to 40% to accurately judge. (would 2 of 4, 50%, mean they were better developers than Bama/Georgia?)

 

[Vol8188]

Is it successful? If they follow the coaches then its arbitrary
They do follow them some
Therefore it is partially arbitrary and one cannot know if rivals followed the coaches more or relied on their own expert evaluations

This is fish food for people who love recruiting and you eat the hell out of it with no understanding of the loop youre in

You said one cannot know and then proclaimed I’m in a loop. If one truly cannot know, isn’t it just as likely that you are the one who is wrong and you’re stuck in a weird loop of attacking a great product?

 

[GUNTERSVOL]

2 5 stars
13 4 stars
15 3 stars
2 2 star/unranked

So the number of 4 stars did go up, but not nearly as much as I thought. 3 stars dominated the 2nd round with almost half of the total draft picks in this round.

using the same numbers from Round 1.
2/32 5 stars = 6%
13/300 4 stars =4%
15/700 3 stars = 2%
2/2109 (3141 - 1032) = 0.0009%

so your best bet of going highly in the draft is still to earn a higher rating. but now the numbers aren't nearly as dramatic as before. this trend says to me that the recruiting rankings are still more right than they aren't, but they aren't nearly as good at predicting this level of talent as they are Round 1 talent.

If there are more than enough 4 & 5 stars to totally populate the entire draft and ABOUT have the draft is comprised of the rest of the free world, that is a 50% failure rate. And that is if you only count 1 years worth of players when that is obviously not the case with 3 and 5 year guys available.

Should not be surprising since the MAJORITY of the high star guys earn them before their SR year and the later bloomers have 5 or so years to max out their abilities. It simply means you can't judge a team's class on signing day. Certainly cannot be used to establish ANY players ceiling which is really all that matters. Signing the right 3 stars can be better than signing a boat load of the wrong 4 stars.

 

[butchna]

No one is saying that. You’re fighting a strawman. No one is saying “3* players will never amount to anything”. That’s not how this works. What is being said is that higher rated players are more likely to excel than lower rated players.

Why are you putting quotation marks around the bolded…as if you’re quoting somebody? Lil TROLLISH on your part. And no one daring to counter you on this subject is denying we need the Nicos and Merklingers and their accompanying stars. Just that we’re not bound to lowered expectations because the star nerds haven’t signed off on the Squirrel Whites or Gage Ginthers. We were excited when Leacock gained his 4th star…but just for his deserved self-worth. We saw what player we were getting before the nerds copied Pope’s work.

 

[butchna]

You said one cannot know and then proclaimed I’m in a loop. If one truly cannot know, isn’t it just as likely that you are the one who is wrong and you’re stuck in a weird loop of attacking a great product?

“Great product”. The crux of our hopeless impasse. We’re just not into what gets you excited.

 

[LouderVol]

If there are more than enough 4 & 5 stars to totally populate the entire draft and ABOUT have the draft is comprised of the rest of the free world, that is a 50% failure rate. And that is if you only count 1 years worth of players when that is obviously not the case with 3 and 5 year guys available.

Should not be surprising since the MAJORITY of the high star guys earn them before their SR year and the later bloomers have 5 or so years to max out their abilities. It simply means you can't judge a team's class on signing day. Certainly cannot be used to establish ANY players ceiling which is really all that matters. Signing the right 3 stars can be better than signing a boat load of the wrong 4 stars.

but if you are doing a blind pick it would still be better ODDS to pick 25 random 4 stars than 25 random 3 stars.