forums

Harrison-Hunte Update

hotshot said:
Paly said, "Many three-stars have significant pro potential." Also true many 4 and 5 stars turn out to have little or no pro potential, when all is said and done (we've had some of those, maybe more than our fair share).
Click to expand...

Most folks don't understand how ranking work. They think it goes like this.....

5* - Very good player, All-American
4* - very good player, but not quite All-American
3* - decent player role player
2* - probably won't start
1* - trash player

This is NOT how recruiting rankings work. Not at all. Lots of 2* and 3* players become All-Americans and All-Pro players. The rankings are not based on ability. They are however based on probability. It looks like this......

5* - Most likely players to become All-Americans
4* - 2nd most likely players to become All-Americans
3* - Only a small percentage this player becomes an All-American
2* - Relatively rare this type of players becomes an All-American
1* - Very small probability this players becomes an All-American

It's amazing to me how many people don't truly grasp how the rankings work in reality.

A 5* players is NOT better than a 3* player
A 5* player is more likely to become an All-American down the road than a 3* player
 
Advertisement
ShadowSpring said:
Most folks don't understand how ranking work. They think it goes like this.....

5* - Very good player, All-American
4* - very good player, but not quite All-American
3* - decent player role player
2* - probably won't start
1* - trash player

This is NOT how recruiting rankings work. Not at all. Lots of 2* and 3* players become All-Americans and All-Pro players. The rankings are not based on ability. They are however based on probability. It looks like this......

5* - Most likely players to become All-Americans
4* - 2nd most likely players to become All-Americans
3* - Only a small percentage this player becomes an All-American
2* - Relatively rare this type of players becomes an All-American
1* - Very small probability this players becomes an All-American

It's amazing to me how many people don't truly grasp how the rankings work in reality.

A 5* players is NOT better than a 3* player
A 5* player is more likely to become an All-American down the road than a 3* player
Click to expand...

Thank you. Yes this summarizes the point I was making.

And D$, imo you're still hung up by not distinguishing between star rankings as an indication of the probability that a given player becomes elite, vs star rankings as an absolute indication of the most elite 300 players. If your interpretation of the what the rankings are supposed to mean is correct, then there'd be no room for this statement when describing a 3 star recruit: "..many have significant pro potential." But that statement describing a 3 star recruit does exist. It's 247's very own own words. But how could that statement possibly exist based on your interpretation that the 300 4 star recruits are the very best 300 players and destined to completely fill all 256 draft picks? It couldn't exist. But it does.

In the final analysis, I don't think the guys at 247 (and Rivals) are so stupid that they don't even understand their own definition of what it means to be a 3 star recruit. I think they fully understand what they're implying when they award various numbers of stars. Your definition of things requires the 247 guys to be so stupid that they've failed to realize the inherent inconsistency in their narrative and mathematical criteria for a 4 star. I don't think that's the most likely explanation. rather, Boxing Robes definition (which many of us subscribe to) allows for both the narrative and mathematical criteria to co-exist (again, because it regards star rankings as an assessment of probabilities, not an absolute ranking). So Boxing Robes criteria is most probably correct imo.

And even in your use of the RJ McIntosh vs Joe Jackson doesn't make your point. You simply conclude that because Joe Jackson and McIntosh have had roughly similar careers, that their respective rankings ex post were incorrect. But all the ranking services said was that as a HS senior Joe Jackson was judged to have a higher probability of turning out to be an elite college player than RJ McIntosh. This doesn't seem egregiously wrong in hindsight, with McIntosh needing to prove he could make the conversion from basketball player to solely a football player; i.e. change his body, add weight and strength. Khris Bogle is another example. At 215 pounds he's not the 3rd best WDE in the country at this immediate moment. Instead, he's ranked #3 based on the expectation that he will gain weight, develop physically and catch up with some other WDE who are already at or close to ideal playing weight and strength for a college WDE. Once he does that, coupled with his elite athleticism, it's expected that he should perform like one of the very best WDE's in college football. But again, his ranking is based on the "probability" he'll become the 3rd best WDE, not based on any belief that he's already the 3rd best in the US.
 
Advertisement
PalyCane said:
And D$, imo you're still hung up by not distinguishing between star rankings as an indication of the probability that a given player becomes elite, vs star rankings as an absolute indication of the most elite 300 players.
Click to expand...

But they rank the players for a reason. You are saying the rankings are based on "probability of success." The only way to hold them accountable for that probability is to measure the 300+ players who received 4* rankings against the players who actually turned out to be the Top 300 prospects.

In 2015, myself (and others) believed RJ McIntosh was more probable to have success than many four stars ranked ahead of him. It is not a hindsight argument. Rivals disagreed. Rivals was wrong.

I don't think Rivals is being stupid in setting out inconsistent criteria. I think they are covering their butts. That way, when they are wrong (like with McIntosh) they can still be right. If a four star player fails, he didn't meet his "potential." If a three-star player succeeds, he falls within the catch-all narrative definition. The only way Rivals can be wrong is if a three-star gets one of 66 All-American spots, or 32 of the first round spots. And even then, they can point to a growth spurt or something unforeseeable. That gives the rankings almost unlimited margin for error.

They have 300+ four star spots. Their own criteria states those spots are reserved for the Top 300 players. That's how they should be judged. They ranked RJ McIntosh below Jamie Gordinier, despite offers from Clemson and Florida and obvious athletic talent. They were dead wrong.
 
Last comment. We just see this one differently.

I'll say that I consider this to be an incorrect observation:

In 2015, myself (and others) believed RJ McIntosh was more probable to have success than many four stars ranked ahead of him. It is not a hindsight argument. Rivals disagreed. Rivals was wrong.
Click to expand...

That example in no way proves that Rivals was wrong. If I make a double-my-money wager and get get 2:1 odds, and you make a double-your-money wager at 1:1 odds, we can both win and double our money. But the fact that the outcome turned out to be exactly the same for both us doesn't mean that the probability of success was the same.

Separate "probability" from "outcome".
 
Kyle Wright and Forston were nowhere near "All American" quality and they not only were 5 stars, but the #1 ranked players among all the 5 stars at their positions nationwide.
 
Advertisement
PalyCane said:
Separate "probability" from "outcome".
Click to expand...

But what’s the objective measure for “probability?” We are accepting criteria at face value which make it almost impossible for Rivals or 247 to be wrong.

They bet against McIntosh. He didn’t get one of 300+ spots. Clemson and others disagreed with Rivals on the “probability of his success” and offered one of their limited scholarships. Many posters agreed with Clemson and disagreed with Rivals.

This disagreement existed in 2015 looking forward, not in hindsight. The services were wrong then and are wrong now.
 
DMoney said:
But what’s the objective measure for “probability?”
Click to expand...

The objective measure for probability is easy. It's whether performance in college (as measured by draft selection in this conversation) is distributed along the lines of the star rankings. That means that vast majority of 5 stars perform at an elite level (although by no means all 5 stars do so), that most 4 stars perform at an elite level although many do not), that many (although not a majority) of 3 stars perform at an elite level, that some 2 stars perform at an elite level (although the vast majority do not, etc. That's the expected distribution of outcomes using the definition of the various star rankings. And this distribution of outcomes is absolutely consistent with the a priori probabilities assigned by the ranking services.

The expected distribution of outcomes is not that every 5 star and every 4 star will turn out to be exactly the 325-350 best college players in the country, which is what you keep claiming.
 
ShadowSpring said:
Most folks don't understand how ranking work. They think it goes like this.....

5* - Very good player, All-American
4* - very good player, but not quite All-American
3* - decent player role player
2* - probably won't start
1* - trash player

This is NOT how recruiting rankings work. Not at all. Lots of 2* and 3* players become All-Americans and All-Pro players. The rankings are not based on ability. They are however based on probability. It looks like this......

5* - Most likely players to become All-Americans
4* - 2nd most likely players to become All-Americans
3* - Only a small percentage this player becomes an All-American
2* - Relatively rare this type of players becomes an All-American
1* - Very small probability this players becomes an All-American

It's amazing to me how many people don't truly grasp how the rankings work in reality.

A 5* players is NOT better than a 3* player
A 5* player is more likely to become an All-American down the road than a 3* player
Click to expand...
You have a point to make but undermine it by mischaracterizing what the ranking services actually do. They are not running probabilistic models. You are trying to present a rationale for what they should be doing, and it’s logical. But they just don’t do it that way.

They are basically collecting information on various measurables, on which programs offer and recruit a kid, on how kids do on camp circuits, and how they do in competition (some qualitative). They then come up with rough position group rankings, and then rerate over the year as they see which kids are being recruited by which programs. At the end of the day, they overrate kids who are being chased by multiple major programs relative to kids who commit early and don’t therefore engage in competitive recruiting. They also intentionally overrate kids whose programs are major revenue drivers for cfb fan sites. That’s a systematic bias.

The point here is that it’s reasonable to observe that expecting their top 250 kids from Hs to be the same top 250 kids four years later is unrealistic and fails to understand uncertainty and probability. But they also run a flawed, skewed ranking system, and it’s equally reasonable to observe that they underrate some kids for identifiable reasons.

The proof would be whether someone can realiably identify a meaningful pool of underrated kids in advance, say around NSD of their senior year (and then be right four years later by comparison to how the services rated them). If not, then they may be ranking better than anyone else can, and errors are random. But if so, then they are probably not as statisitically driven as you seem to think.
 
Advertisement
Sigma_Cane_7 said:
Bruh. Curious, are you posting from a PC. Your font is always official. Makes me feel font inferior lol.
Click to expand...
So, you got Font Envy.

Reminds me of Woody Allen in a movie; he admitted to having ***** Envy, which, up until then, was only known as a condition Freud attributed to women.
 
DMoney said:
But they rank the players for a reason. You are saying the rankings are based on "probability of success." The only way to hold them accountable for that probability is to measure the 300+ players who received 4* rankings against the players who actually turned out to be the Top 300 prospects.

In 2015, myself (and others) believed RJ McIntosh was more probable to have success than many four stars ranked ahead of him. It is not a hindsight argument. Rivals disagreed. Rivals was wrong.

I don't think Rivals is being stupid in setting out inconsistent criteria. I think they are covering their butts. That way, when they are wrong (like with McIntosh) they can still be right. If a four star player fails, he didn't meet his "potential." If a three-star player succeeds, he falls within the catch-all narrative definition. The only way Rivals can be wrong is if a three-star gets one of 66 All-American spots, or 32 of the first round spots. And even then, they can point to a growth spurt or something unforeseeable. That gives the rankings almost unlimited margin for error.

They have 300+ four star spots. Their own criteria states those spots are reserved for the Top 300 players. That's how they should be judged. They ranked RJ McIntosh below Jamie Gordinier, despite offers from Clemson and Florida and obvious athletic talent. They were dead wrong.
Click to expand...
You are missing an important mathematical point.

Just take two pools of 100 kids. Say one groups has a 50% probability of being drafted in the nfl four years later, and the other pool of kids have a 5% probability of being drafted at the same time.

It’s a mathematical fact that some of the 5% kids will be drafted amd some of the 50% kids will not be. That doesn’t mean the pools were defined wrong. You have got to account for the possibility that kid A was in fact a better prospect (measured by probability) than kid B coming out of Hs, but still kid B can turn out better four years later. It does not mean they were wrong in the rating coming out of HS.

There are lots of flaws in the rating services to point out, but the math we should all agree on.
 
I am just laughing, after all of this, its over a player that was a second team All-ACC player and 5th Round pick.

Seems like a poor player to plant your flag in to make your point.
 
Advertisement
BoxingRobes said:
I am just laughing, after all of this, its over a player that was a second team All-ACC player and 5th Round pick.

Seems like a poor player to plant your flag in to make your point.
Click to expand...
You seriously don’t think the thyroid issue that was flagged at the combine didn’t affect where he was picked?.. come on now
 
Last edited:
BoxingRobes said:
I am just laughing, after all of this, its over a player that was a second team All-ACC player and 5th Round pick.

Seems like a poor player to plant your flag in to make your point.
Click to expand...

If we had a steady stream of starting DTs of McIntosh's ability playing for us every year, would that be anything short of a success?
 
Boxing Robes has to be the king of negativity. Dude’s got a serious case of PTCD.
 
Advertisement
Back
Top