re: Graphic: FPI Probability of wins by week

Those percentages are stupid.

35% chance of beating Florida, A&M, Miami
22% chance of beating Bama and Georgia
17% chance of beating auburn

LOL where did you get this from?

I really dont see this being that far off. We will be major underdogs in like 3 games this yr and underdogs in a few more.

U can't look at last years schedule and not know this years is much, much tougher.

quote:

---

put together by a guy on reddit, who did one for every FBS team

---

Sounds like a real expert and somebody who analyzed tons of empirical data beforehand.


Kind alike this.....

quote:

---

I'd say 50% chance to beat Miami, 98% chance to win W2 and 25% to win @ Auburn so whatever that would work out to.

---

I mean Miami is the real head scratcher but it's the first game of the year. There have been wide ranging predictions for that game.

But 25% to beat Auburn. That's not really much better than 17%. Both percentages represent extremely low probability events.

Using S&P+ data, instead of FPI, which at this point with zero games played, neither is any better than fancy statistical guesswork, here is the year end chart for every team in the SEC.



Look at the wins mode value, which is the most likely number of wins for each team.

SEC West
11 Bama
9 Auburn
8 State
7 LSU, A&M, Ole Miss
6, Arkansas

SEC East
10 Georgia
7 Florida, Mizzou, S. Carolina
6 Kentucky
5 Tennesse, Vandy

The statistics are valid, but are the input probabilities? At this point who the heck knows?
Check back after game 6 and they'll be different and closer to reality.

quote:

---

I'd say 50% chance to beat Miami, 98% chance to win W2 and 25% to win @ Auburn so whatever that would work out to. I could see 10% chance to start 3-0.

---

Hmmm