Texas + Darrell Scott = 52.7% probability of love
by Ultra Horn on Jan 28, 2008 10:22 PM CST
Email
PrintAfter a search in google news, I stumbled upon an article about a professor from Mercer University, Georgia who created a college recruiting simulator based on several criteria, such as official school visits, BCS school, bowl games, final AP ranking, home town and home state proximity, conference and national titles and other factors.
The predictor calculated a 52.7% of Texas landing Scott vs. a 40.3% probability for Colorado and a 7% chance for LSU.
Every other athlete Texas is trying to recruit is in accordance with the predictions of the software. I won't summarize any longer, just hit ctrl+f to search for all the Texas players on file and be proud that we were the school with a 50+ million dollar budget that was mentioned to highlight the importance of college football :)
Predictor Site:
http://ssbea.mercer.edu/recruiting.htm
All comments, FanPosts, and FanShots are the views of the reader-authors who create them.
11 comments
|
0 recs |
Do you like this story?
Comments
Cool
Someone else posted about this some time ago, but I hadn't seen the link to the actual site until now. I'm hoping they'll keep doing this every year. It will be interesting to see how the Shepard and Gilbert situation is predicted.
Don't worry...
The last time it was posted, they hadn't posted the entire data on the website, it was just an article on SI.com showing the prediction for Pryor, Scott and Julio Jones I think.
The one thing scares me, is that Miami FL, Bama and Colorado are 3 teams that stick out as getting recruits they "shouldn't" be getting.
Also, the model has no control for family relationships, which changes things. Either that or they decided that the variable was not significant in their model (I actually took a class in discrete choice, so im semi familiar with the method they used)
by fathead on Jan 29, 2008 2:54 AM CST up reply actions
hmmmm
But does the model take into account the thermal radiation coefficient? I think not. {throws model in the trash}
I guess it's interesting as a predictive instrument, but I find it amusing that it takes credit for early commitments such as Humphrey and Aaron Williams for us. Now, when it comes to crunch time and there are only a couple dozen players left with limited scholarships to hand out, that's one thing. But a 78% hit rate? Arguable.
Still fun, though. If it's right about Scott, great. If it's wrong, it sucks.
Actually
They're not "taking credit" for it. What they're saying is that they applied their model to each player and that's the result they got. Considering the paper is getting published, I doubt that they tweaked the model for each player, in which case it would make their 78% hit ratio look like shit.
You have to remember, these aren't journalists.
by fathead on Jan 29, 2008 3:12 PM CST up reply actions
hmm indeed
bigfatdrunk,
It really doesn't rule out the possibility of Scott going to Colorado. As fathead mentioned, as far as we know, the model doesn't take into account family relationships, or the thermal radiation coefficient for that matter =]
I believe he would regret going to Colorado and not winning any major league, bowl game thus he will choose Texas. I think he is smart enough to figure out that not many decent, good and star football recruits have family in the school. He would be their one and only national caliber signing in the next 4 years if no dramatic changes take place.
And so we wait for an answer :]
true
Mostly, I'm getting a kick out of an analysis system trying to guess the whims and desires of 17 and 18 year old kids. I haven't read the whitepaper behind the system, but unless it involves sex, alcohol, and/or drugs, I would worry about the true accuracy of their model.
Hmmmm, seems to me those three evils are considerations when you are 37, too.
Well
Considering these same models are used to predicts the decisions of all demographics about all types of products and services, I don't see how this method is any less appropriate. It's the same method used to figure out if you're gonna buy coors light or bud light, whether you're gonna use public transportation or not, etc, etc.
Trust me, the people that do this deal with analyzing things a lot more fickle than 17 and 18 year olds.
by fathead on Jan 29, 2008 9:25 PM CST up reply actions
Colorado Recruiting
He would be their one and only national caliber signing in the next 4 years if no dramatic changes take place.
Colorado got an O lineman in the rivals top 100 last year and a LB and an O lineman in the rivals top 100 this year. What is your def of national caliber?
Colo
A player they would build their offense around, I guess I was thinking QB, RB. I don't know much about Colorado and underestimated their recruiting capabilities - it appears they've recruited a lot more talent from away states than Texas has this year. Then again the state of Colorado's football player pool is much smaller than the one of Texas.
Disregard what I posted about national caliber :} It appears Colorado has a pretty good class coming in from I've looked at in rivals.
In conclusion Texas still has more to offer than CU.
Bad nightmares
Reading that article gave me nightmares about my Econometrics class while I was at Texas.
by The Tres Leches on Jan 29, 2008 9:41 AM CST reply actions
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
