Those percentages are based off the expected Vegas lines for those games. So if Vegas has made you a 10 pt fav then historically that team has won the game say 70% of the time.
Based off those percentages on the link, we have a 1.4% chance of going undefeated.
If we beat FSU (and the lines stay the same for the remaining games) our chances of winning the remaining games jumps to 7.8%
Based off those numbers , the Vegas over/under for wins will probably be 9 games which is actually pretty good since you will not see many teams with a much better O/U number.
Very good post. You have the correct perspective. The vast majority of posters on forums like this have little clue. They'll see 3 point projected favoritism in every game and act like the oddsmakers are expecting us to go undefeated.
A 10 point favorite in college football is actually closer to 77% than 70% win likelihood. Here is a link that breaks it down well:
https://www.boydsbets.com/college-football-spread-to-moneyline-conversion/
In the NFL it is higher than 77%. College favorites are not as reliable to win the game as pro favorites. Considerably greater dispersion from spread to actual result. That's why teasers are laughable poor value in college football. We used to restrain snickers behind the sportsbook counter whenever somebody was making college football teaser bets.
I normally cite money lines around here as how the season win over/unders are derived. Naturally the conventional wisdom types don't like that. They envision a big secretive brain trust in which every conceivable player and matchup and variable is deciphered. Hilarious. I've seen posts about the good old days on some other forum, a decade or more ago. Good thing I wasn't there at the time or I would have effortlessly deflected all the garbage at that point. Basically that type of thinking had a free undeserved reign of ignorance.
Since most people who don't bet sports are unfamiliar with money lines (-260...+220) and how they vary from sport to sport, the projected win percentage is probably a better reference point on forums like this. When a sportsbook is assigning the season win over/unders they'll translate the spread into a percentage for each game and then feed those percentages into a formula to spit out the projected wins and therefore the over under. Let's say the win expectancy is 60% each time among the first 5 games. That ESPN link and the way it was authored would assert, "Projected Record: 5-0." Meanwhile, the betting over/under among those 5 games would be 3-2.