Project Finished - Need some help from Football Experts!
Update: So I just finished the project I was working on, and as promised here is a copy. I pulled out some of the technical stuff that nobody else probably cares about. Thanks again for your help!
View attachment Elicitation of Expert Opinion on College Football Offenses - CF.pdf
I am doing a project for one of my grad school classes and I need CF's help. The class is Data Collection and Analysis, and the topic I've chosen is elicitation and use of expert opinion as it relates to FBS football. So what do I need from you?
Estimate of Yards per Rush
I'd like a single value of how many yards you think are gained per rush on average across all of FBS college football (not just the Cyclones), along with an estimate of the range of all values. For example, "I think the average yards per rush is 5 yards, and 80% of all rushes are between -1 and 8 yards." You can give me an average or a median as long as you let me know which it is. Or give me a median and an average if you're so inclined. Just let me know what it is.
Estimate of Yards per Pass
Same as above, but for pass attempts (note that I want per attempt, not per completion).
Estimate of Yards per Play
Same thing again, for all plays.
Estimate of Points Scored per Game Per Team
Same thing, for point scored.
Are you a former college player (any level) or a coach (Grades 9+)?
I will use this info to do some weighting of the data.
Please reply to this thread with your "expert opinions!"
I want your opinions, not actual data! I know I could just look up straight averages, but that doesn't give me any estimate of variabiility. And most importantly, it doesn't give me any estimate of the uncertainty in the opinions of the subject matter experts. I will just compile the data straight out of the thread.
So why do I need this data? I'm going to use the expert opinions gathere here to create baseline prior probability distributions for all FBS college football teams. Then I'll take real game data (certainly from ISU, but I could choose any team if you have other suggestions) and make Bayesian updates after each "game" through the entire season. Then I'll compare my Bayesian updates from each "game" to the real end-of-the-season distributions and see how long it takes for the updates to converge on the real answer. As a bonus, I'll do the same for any bowl pairings announced over the next week and see if they are useful for making bowl game predictions.
Let me know if you have any questions on what I'm asking for and I will explain further.Thanks in advance for your help, Fanatics!
Update: So I just finished the project I was working on, and as promised here is a copy. I pulled out some of the technical stuff that nobody else probably cares about. Thanks again for your help!
View attachment Elicitation of Expert Opinion on College Football Offenses - CF.pdf
I am doing a project for one of my grad school classes and I need CF's help. The class is Data Collection and Analysis, and the topic I've chosen is elicitation and use of expert opinion as it relates to FBS football. So what do I need from you?
Estimate of Yards per Rush
I'd like a single value of how many yards you think are gained per rush on average across all of FBS college football (not just the Cyclones), along with an estimate of the range of all values. For example, "I think the average yards per rush is 5 yards, and 80% of all rushes are between -1 and 8 yards." You can give me an average or a median as long as you let me know which it is. Or give me a median and an average if you're so inclined. Just let me know what it is.
Estimate of Yards per Pass
Same as above, but for pass attempts (note that I want per attempt, not per completion).
Estimate of Yards per Play
Same thing again, for all plays.
Estimate of Points Scored per Game Per Team
Same thing, for point scored.
Are you a former college player (any level) or a coach (Grades 9+)?
I will use this info to do some weighting of the data.
Please reply to this thread with your "expert opinions!"
I want your opinions, not actual data! I know I could just look up straight averages, but that doesn't give me any estimate of variabiility. And most importantly, it doesn't give me any estimate of the uncertainty in the opinions of the subject matter experts. I will just compile the data straight out of the thread.
So why do I need this data? I'm going to use the expert opinions gathere here to create baseline prior probability distributions for all FBS college football teams. Then I'll take real game data (certainly from ISU, but I could choose any team if you have other suggestions) and make Bayesian updates after each "game" through the entire season. Then I'll compare my Bayesian updates from each "game" to the real end-of-the-season distributions and see how long it takes for the updates to converge on the real answer. As a bonus, I'll do the same for any bowl pairings announced over the next week and see if they are useful for making bowl game predictions.
Let me know if you have any questions on what I'm asking for and I will explain further.Thanks in advance for your help, Fanatics!
Last edited: