I’ve done this for several seasons now. It’s a simple, but usually reliable process…I look at each game, independently, put my most objective % that we win the game, based on opponent, venue, where it falls in the schedule (i.e., who we played the week before and play the week after), etc. Then, I sum up the percentages, and that’s my predicted number of wins. In probability and statistics, it’s know as the “expected value”.
Normally, I’ve been on the number, or one win/loss away. Last year, I had perhaps my worst ever (can’t remember for sure how far off I was in 2006, but I believe it was worse)…I only predicted 6.15 wins (regular season) and we had 8…so I was almost 2 off.
Here we go:
@ Miss State…60%
Add it up, and it comes to 7.55 wins, which means it will come down to the 4 virtual toss-up games (A&M, @ MSU, @ BYU and Bama) to see if we win 2 or 3 of those 4 games to determine if we are 7-5 or 8-4.
Note that I never include Bowl game results (regular season only), because we have no idea who we will play in a Bowl, and the match-up matters.
By the way, I plugged in the most recent ESPN FPI percentages that I could find (they used to show those numbers on the ESPN schedule, but I don’t see them any more), and those figures project 6.5 wins.
Now…to find out what REALLY happens instead of reading everyone’s prediction/projection.