Illinois Football Recruiting Thread

Status
Not open for further replies.
#901      

altgeld88

altgeld88
Arlington, Virginia
altgeld88
aaeismacgychel said:
It won't be as impressive after I explain it, lol, but the major things I had to go on is that 24/7 claimed they used a gaussian distribution formula and that they ordered the recruits such that a school's top recruit was weighted 100% with each ensuing next top recruit getting weighted less. And it also greatly helped that I knew the answer to what the contribution of each athlete needed to be thanks to 24/7's class calculator.

A Gaussian function has the following form:
f(x)=A*e^{-.5*[(x-B)/c]^2}

Based on the fact that we know the top recruit for a given school is weighted 100%, that means the decay part of the function goes to e^0. So A is fairly easy to figure out since you know the contribution and rating of your top recruit. And when looking at a few schools to figure out how they got A from the recruit's rating, it became very apparent that A was just that recruit's rating minus 0.7 and multiplied by 100.

Similarly, knowing that there is an e^0 term for the first recruit, you know x-B equals 0 when x is 1, so B is equal to 1. That means A and B are solved for and all that's left is solving for C.

Since you have 11 recruits and know what their contributions are, you just have a system of equations and it's straightforward to finding c is about equal to 3.0. And after looking at an additional team with more recruits, this held true.

So as I said, not so impressive if you know what a Gaussian function looks like based on the info they provided. I personally did expect it to be more complicated as was being suggested, but the basic equation just worked, so no additional investigation was necessary.
Click to expand...
That's fabulous. I hereby cede the "Altgeld" handle to you, my friend. Mine will subsequently become "JanitorCloset88."
 
#905      

Indy Illini Fan

Indy Illini Fan
Indy Illini Fan
cookie86 said:
So I guess we're not going to see any commits today?
Click to expand...

We’ll have some announcements coming relatively soon … It is only Monday and I said 3 from this weekends visitors … We got 1 … I expect 2 more …

Patience Calm Down GIF by Brat TV
 
#910      

skyIdub

skyIdub
Winged Warrior
skyIdub
aaeismacgychel said:
It won't be as impressive after I explain it, lol, but the major things I had to go on is that 24/7 claimed they used a gaussian distribution formula and that they ordered the recruits such that a school's top recruit was weighted 100% with each ensuing next top recruit getting weighted less. And it also greatly helped that I knew the answer to what the contribution of each athlete needed to be thanks to 24/7's class calculator.

A Gaussian function has the following form:
f(x)=A*e^{-.5*[(x-B)/c]^2}

Based on the fact that we know the top recruit for a given school is weighted 100%, that means the decay part of the function goes to e^0. So A is fairly easy to figure out since you know the contribution and rating of your top recruit. And when looking at a few schools to figure out how they got A from the recruit's rating, it became very apparent that A was just that recruit's rating minus 0.7 and multiplied by 100.

Similarly, knowing that there is an e^0 term for the first recruit, you know x-B equals 0 when x is 1, so B is equal to 1. That means A and B are solved for and all that's left is solving for C.

Since you have 11 recruits and know what their contributions are, you just have a system of equations and it's straightforward to finding c is about equal to 3.0. And after looking at an additional team with more recruits, this held true.

So as I said, not so impressive if you know what a Gaussian function looks like based on the info they provided. I personally did expect it to be more complicated as was being suggested, but the basic equation just worked, so no additional investigation was necessary.
Click to expand...

cross-eyed.gif
 
#920      

Sutton_Place

S
Sutton_Place
aaeismacgychel said:
It won't be as impressive after I explain it, lol, but the major things I had to go on is that 24/7 claimed they used a gaussian distribution formula and that they ordered the recruits such that a school's top recruit was weighted 100% with each ensuing next top recruit getting weighted less. And it also greatly helped that I knew the answer to what the contribution of each athlete needed to be thanks to 24/7's class calculator.

A Gaussian function has the following form:
f(x)=A*e^{-.5*[(x-B)/c]^2}

Based on the fact that we know the top recruit for a given school is weighted 100%, that means the decay part of the function goes to e^0. So A is fairly easy to figure out since you know the contribution and rating of your top recruit. And when looking at a few schools to figure out how they got A from the recruit's rating, it became very apparent that A was just that recruit's rating minus 0.7 and multiplied by 100.

Similarly, knowing that there is an e^0 term for the first recruit, you know x-B equals 0 when x is 1, so B is equal to 1. That means A and B are solved for and all that's left is solving for C.

Since you have 11 recruits and know what their contributions are, you just have a system of equations and it's straightforward to finding c is about equal to 3.0. And after looking at an additional team with more recruits, this held true.

So as I said, not so impressive if you know what a Gaussian function looks like based on the info they provided. I personally did expect it to be more complicated as was being suggested, but the basic equation just worked, so no additional investigation was necessary.
Click to expand...
Sorry to disappoint but the assumption of Gaussian Distribution is not empirically supported and a long stretch. For small sample like this, you need students t distribution that allows thicker tails or excess kurtosis. If you throw couple of 4 or 5 stars players in the mix, that will result in skewed distribution or non-zero skewness, so may be a chi-squared distribution may end up being useful. Just must 2 cents. (Dan please remove this message if it's inappropriate)
 
#921      

TentakilRex

TentakilRex
Land O Insects between Quincy-Macomb-Jacksonville
TentakilRex
aaeismacgychel said:
It won't be as impressive after I explain it, lol, but the major things I had to go on is that 24/7 claimed they used a gaussian distribution formula and that they ordered the recruits such that a school's top recruit was weighted 100% with each ensuing next top recruit getting weighted less. And it also greatly helped that I knew the answer to what the contribution of each athlete needed to be thanks to 24/7's class calculator.

A Gaussian function has the following form:
f(x)=A*e^{-.5*[(x-B)/c]^2}

Based on the fact that we know the top recruit for a given school is weighted 100%, that means the decay part of the function goes to e^0. So A is fairly easy to figure out since you know the contribution and rating of your top recruit. And when looking at a few schools to figure out how they got A from the recruit's rating, it became very apparent that A was just that recruit's rating minus 0.7 and multiplied by 100.

Similarly, knowing that there is an e^0 term for the first recruit, you know x-B equals 0 when x is 1, so B is equal to 1. That means A and B are solved for and all that's left is solving for C.

Since you have 11 recruits and know what their contributions are, you just have a system of equations and it's straightforward to finding c is about equal to 3.0. And after looking at an additional team with more recruits, this held true.

So as I said, not so impressive if you know what a Gaussian function looks like based on the info they provided. I personally did expect it to be more complicated as was being suggested, but the basic equation just worked, so no additional investigation was necessary.
Click to expand...
1. Logarithms are neat, wish high schools spent more time on them (practical use: you can calculate who long it will take for your money to double in an account with any given interest).
2. This is some message board genius stuff, but in this case genius is not being used sarcastically. Well done 👏👏👏.
 
#922      

DeonThomas

DeonThomas
South Carolina
DeonThomas
Sutton_Place said:
Sorry to disappoint but the assumption of Gaussian Distribution is not empirically supported and a long stretch. For small sample like this, you need students t distribution that allows thicker tails or excess kurtosis. If you throw couple of 4 or 5 stars players in the mix, that will result in skewed distribution or non-zero skewness, so may be a chi-squared distribution may end up being useful. Just must 2 cents. (Dan please remove this message if it's inappropriate)
Click to expand...
Not inappropriate......fascinating, actually. But perhaps a new thread..........

nerdy the passing zone GIF by America's Got Talent
 
#923      

TentakilRex

TentakilRex
Land O Insects between Quincy-Macomb-Jacksonville
TentakilRex
Sutton_Place said:
Sorry to disappoint but the assumption of Gaussian Distribution is not empirically supported and a long stretch. For small sample like this, you need students t distribution that allows thicker tails or excess kurtosis. If you throw couple of 4 or 5 stars players in the mix, that will result in skewed distribution or non-zero skewness, so may be a chi-squared distribution may end up being useful. Just must 2 cents. (Dan please remove this message if it's inappropriate)
Click to expand...
Engineer fight! jk

(Apparently we can and should create some IllinoisLoyalty Analytics section and posts ASAP.
 
#924      

logjam

logjam
logjam
Sutton_Place said:
Sorry to disappoint but the assumption of Gaussian Distribution is not empirically supported and a long stretch. For small sample like this, you need students t distribution that allows thicker tails or excess kurtosis. If you throw couple of 4 or 5 stars players in the mix, that will result in skewed distribution or non-zero skewness, so may be a chi-squared distribution may end up being useful. Just must 2 cents. (Dan please remove this message if it's inappropriate)
Click to expand...

TentakilRex said:
1. Logarithms are neat, wish high schools spent more time on them (practical use: you can calculate who long it will take for your money to double in an account with any given interest).
2. This is some message board genius stuff, but in this case genius is not being used sarcastically. Well done 👏👏👏.
Click to expand...
Star Trek Wow GIF
 
#925      

EyeoftheIllini

EyeoftheIllini
EyeoftheIllini
TampaToo said:
There was this Witherspoon kid that had a weak offer list who committed to Illinois and he turned out ok. Ya just never know. 🤷🏼‍♂️
Click to expand...
Very true and I trust the staff. But I do get a kick out of posts that invoke the the names of Illini greats as proof or rationale for something. (Not meant to be slam at you or the post). In the bball forum I chuckle when people say that DGL can contribute as a freshman cuz Dee and Deron did it. I have no idea if he can or not — have my opinion but not really important. But because 2 Illini greats did it really isn’t any kind of argument for DGL being able to do it.
 
Status
Not open for further replies.