After years of predicting Clemson football wins and losses by picking each individual game, KenPom's college basketball projections have prompted me to think more analytically and not consider any FBS game an absolute win or loss. With each game, there's a chance of losing (and winning, even in Tallahassee).
When last season's schedule was released, I gave it a look and on a game-by-game basis and thought we could win each one (remember we didn't know how good FSU was), but I never thought we would actually go undefeated. So, rather than taking random stabs at where we would stumble, @Robert_Reinhard and I created a simulator with which we can model the season using "Victory Likelihood" (a subjective prediction of our chances to win the game). I did this last season and the simulator correctly showed 10 regular season wins as the most likely outcome.
This is where it gets fun. You give me the victory likelihoods for each game in the comments section. I'll run your numbers through the simulator which plays each game 1,000 times and tell you in how many times (as a percentage) Clemson wins seven, eight, nine, 10, and 11 games. Reasonable inputs will lead to numbers below seven wins being about zero. To figure out the likelihood of a 12-0 season, you can just multiply all your percentages together.
I will try to get to as many comments as I can. The below table is my example. Be mindful that it is just an example, and don't read too much into the percentages. I've included it just to give some guidance.
|100%||South Carolina State|
|15%||at Florida State|
|85%||at Boston College|
|95%||at Wake Forest (THUR)|
|65%||at Georgia Tech|
Likelihood of 7 wins: 12%
Likelihood of 8 wins: 25%
Likelihood of 9 wins: 33%
Likelihood of 10 wins: 17%
Likelihood of 11 wins: 8%
Those percentages use the victory likelihood percentages from the table above. Leave your victory likelihood numbers in the comments and I"ll reply with the resulting outcome probabilities.