Gonna be interesting!
If Dr. Bob would only read this post and then take the "it" factor into consideration on his plays....he'd be the best....
For example, for the last 3 games A&M played at Texas Tech, Dr. Bob had A&M as a 3*, 2*, 3*. Knowing the long hatred and more recent bloodshed between these two schools, my attorney friend emailed him and guaranteed him that he would lose these games due to the "it" factor. Sure enough, A&M hasn't covered in Lubbock in forever, much less won.
This "it" factor is bigger in college football than in any other sport I believe. That is why we love it. :toast:
Numbers can only take you so far, the emotional factor & the trend as you stated about A&M are over looked by guys that strictly deal with a statistical program. If you combine the statistical program with everything else you can get your hands on, then I believe you have something that will probably beat the Vegas number 58% or better consistently.
Ok I am going to put you on to a site (colleyratings.com). It is a computer based rating system that uses simple statistical principles, uses only Div. 1 wins and loses, does not take into account margins of victory, but does consider strength of schedule. It uses the Laplace statistic and the interative technique. If your boys are not aware of it by now they better check it out. We are all interested in seeing what MIT can come up with so this is my contribution to the cause. Hope it helps.
yea my friend would kill me if he knew i was listing his group's project picks or whatever the fuck it is....He said he will give me the picks that are over 55% on the program and tell me the percentage for each game..maybe someone can equate that to units...he wants me to buy him some cool hard drive if I end up making money for him squealing to me the programs picks...who knows if it will work...maybe they will suck, either way I'm excited and its better than me picking a coin
yea should be interesting...thanks for the links but i doubt they are interested....they dont care what i think or what info i can contribute...if the school...which essentially is funding the program-experiment etc found out my friend was giving me the picks...they would be in deep shit...
The bottom line...
programs can simulate trend data, but that data has to be given a value. How do you determine what variable gets what value? It has to be a specific #. If Team A is simulated to outrush Team B by 200 yds, does that = a 17 pt win? You have to know more about college football then you do about math IMO, but what makes it difficult is there are too many variables that you can't quantify, or put a # on. Plus the ones that you can tend to change for all sorts of reasons under all sorts of conditions.
programs can simulate trend data, but that data has to be given a value. How do you determine what variable gets what value? It has to be a specific #. If Team A is simulated to outrush Team B by 200 yds, does that = a 17 pt win? You have to know more about college football then you do about math IMO, but what makes it difficult is there are too many variables that you can't quantify, or put a # on. Plus the ones that you can tend to change for all sorts of reasons under all sorts of conditions.
I think they are aware of that, one of the professors who proposed doing the program was a big gambler at some point and probably got alot of info from insiders etc especially if the school gave the green light to do this...whatever the case at least I'll have ten picks a week from god knows where schools to put some money on!
Maybe you can run this idea by your boys... This is something I have always been curious about:
How can you ultimately quantify the strength of an offensive line and defensive line?
Many people may respond rushing yards, yards per carry, def. yards allowed, etc. But all of these aspects can be dependent upon the opponent, the weather, one long breakaway run to taint the stats, one or two mental mistakes by a Free Safety, or other various factors. Screw all skilled positioned player stats. If you know the beef, you know the winners. Look at a team like 2008 Georgia Bulldogs: all skill, little line (overrated) vs. a team like 2008 Alabama Crimson Tide: all line, little skill (underrated).
"Bold" statement of the day: If you can say that "Team A has a better offensive line than Team B, and Team A has a better defensive line than Team B," I guarantee you'll find amazing ATS numbers (plays and fades). I would be very interested in a Top 10 offensive lines and Top 10 defensive lines in the country, especially a Top 10 under-rated O-lines and D-lines.
The question is: how can you accurately do this?
It would be extremely valuable these first 4-6 weeks of the season.
How can you ultimately quantify the strength of an offensive line and defensive line?
Many people may respond rushing yards, yards per carry, def. yards allowed, etc. But all of these aspects can be dependent upon the opponent, the weather, one long breakaway run to taint the stats, one or two mental mistakes by a Free Safety, or other various factors. Screw all skilled positioned player stats. If you know the beef, you know the winners. Look at a team like 2008 Georgia Bulldogs: all skill, little line (overrated) vs. a team like 2008 Alabama Crimson Tide: all line, little skill (underrated).
"Bold" statement of the day: If you can say that "Team A has a better offensive line than Team B, and Team A has a better defensive line than Team B," I guarantee you'll find amazing ATS numbers (plays and fades). I would be very interested in a Top 10 offensive lines and Top 10 defensive lines in the country, especially a Top 10 under-rated O-lines and D-lines.
The question is: how can you accurately do this?
It would be extremely valuable these first 4-6 weeks of the season.
