various reasons. There is a reason the schools have been reluctant to incorporate Week 0 into the regular season calendar and still require schools (FBS and FCS) to get a waiver before they can agree to play a Week 0 game. You may not think it is a big deal to practice in hotter conditions, but the schools may see it as a liability issue. Since the only college football games I watch on TV involve VT and sometimes JMU, I would have no interest in a Clemson-Penn State game regardless whether it is a bowl game or Week 0/1 regular season game. The whole discussion boils down to a simple economic one for me. If a location thought enough people would travel there in late August and companies were willing to sponsor the game, then we would be seeing more of these games being plan than we currently do now. The other piece of it is the network piece. Right now the bowl games are spread out with some played during the work week (early afternoon, mid-afternoon, and night) to minimize the number of games in direct competition with each other. For example, the Famous Idaho Potato Bowl is scheduled to play this year on December 22 (Monday) at 2pm. Since you said these games are essentially for TV ratings, how good of a ratings do you think the game would get if it were to play next year on August 24 (Monday) at 2pm? I count a total of 14 bowl games currently scheduled to play on a week day and kickoff before 5 pm. They should all get pretty good ratings because of the number of people off work or school during the holidays. Do you truly believe they would get the same ratings if they played on the same day of week and at the same time in August? If you believe those games would be limited to the evenings and/or the weekend, then their ratings should go down because they will be in direct competition with at least 1 other bowl game. If multiple bowl games are played at the same time, then they will all see a decline in ratings which doesn't make good business sense given that ESPN owns all except a handful of bowl games now.