Prediction: Tennessee will go 8-4. See how the Vols might get there

Thumbnail image for derek-dooley-govolsxtra.JPG

KNOXVILLE, Tennessee — My official preseason prediction for Tennessee is due today. If you listened to my appearance on The SportsPage this morning, you already heard part of my explanation for why I picked the Vols to go 8-4 in 2012.

For math geeks, the rest of the story is here:

To start this scientific process, we begin with an utterly unscientific premise. We have to determine the likelihood of Tennessee winning each of its games. These numbers will change as the season goes on and we learn more about UT and its opponents. But for now, it’s merely an educated guess.

Tennessee’s 60 percent likelihood of beating N.C. State is actually the most objective number on the list. The 60 percent roughly corresponds to the moneyline for the game. Vegas is not infallible, of course, but at least the moneyline relies on the opinion of oddsmakers and thousands and bettors — not one dude with a blog.

Other games were much harder to put a percentage on. Obviously, Georgia State and Akron are gimmes. I upped Troy’s chances of an upset to 2 percent. But conference games were tough. If you think I’m way off on any of these numbers, let me know…

vs. NC State — 60 percent
vs. Ga. State — 99 percent
vs. Florida — 52 percent 
vs. Akron — 99 percent 
at UGA — 30
at Miss. State — 55
vs. Alabama — 20
at S. Carolina — 35
vs. Troy — 98
vs. Missouri — 60
at Vanderbilt — 65
vs. Kentucky – 70 percent

To get a baseline prediction, we can simply add up the expected win total. That is, .60 plus 0.99 plus 0.52, etc.

That gives Tennessee 7.43 wins. I think that’s in the ballpark, as 7.5 seems to be the consensus “over-under” win total for the Vols this season. Of course, you can’t win a half-game, so I rounded up to come up with my final 8-4 record.

The neat thing about this model is that you can adjust it as the season goes on. So if the Vols win the opener, the projection surges to 7.83 wins. If the Vols play well and you increase their expected win percentage in the rest of their games accordingly, the projection would likely be around 8.

On the other hand, if UT loses to N.C. State, the projection would quickly drop to 6.83 wins, and it could fall more dramatically if you decrease the Vols’ win likelihood in other games.

The other neat thing you can do with these numbers is calculate odds of certain scenarios. At the simplest level, if you have a 50 percent chance of winning Game A and a 50 percent chance of winning Game B, then you have a 25 percent chance of winning both (0.50 * 0.50 = 0.25).

With that in mind, let’s look at some different scenarios…

Chance of UT starting season 3-0: 30.8 percent
Chance of UT starting season 1-2 (losses to N.C. State and Florida): 19.2 percent
Chance of UT starting season 4-0: 30.6 percent
Chance of UT starting season 5-0: 9.1 percent
Chance of UT starting season 6-0: 5 percent
Chance of Tennessee going 12-0: 0.0944921 percent
Chance of UT closing season with 4 consecutive wins: 26.8 percent
Chance of UT going 0-3 against UGA, Alabama and South Carolina: 36.4 percent
Chance of UT starting the season 3-0 and ending the season 4-0: 8.3 percent

The caveat, of course, is that all these neat numbers are based on my win percentage guesses above. So if you disagree with my numbers, this won’t mean much to you. But if you think the game-by-game guesses are in the ballpark, this should provide a good guide for what to expect this year.

That doesn’t mean the Vols can’t overachieve or underachieve this season. But this provides a good baseline to draw that line between an underachieving season, an overachieving one and one that goes just about as expected.

4 thoughts on “Prediction: Tennessee will go 8-4. See how the Vols might get there

  1. Gerry Rose

    I like the article, but what I don’t understand is why it is hidden on the GVX site. Maybe I’m missing something, but I got to the article by clicking on your picture in the blogs section. That gave me the link to the article. Why isn’t the article posted as a blog entry on the main page? And why, if you click the blogs heading, do you still get sent to a page with Gribble’s and Ward’s blogs but not yours. I doubt anyone wants to go back and read the blogs from last year, but if those blogs are important enough to be there, I think yours should qualify for inclusion.

  2. Larry

    Interesting. So would you say 8 wins is a max and since you have 3 at 60% or less do you think 7 losses again are possible. I was saying before Rogers was dismissed 9-3 which if they win all games in which you have them at better than 50% would be 9 wins. I think 8 wins will guarantee Dooley another year 7 though will be very iffy and no more than 6 and he’s out.

  3. Evan Woodbery

    Larry, I don’t think I can calculate the odds of UT losing 7 games alone, but I can do the likelihood of UT losing 7 specific games.
    The 7 toughest, arguably, are NC State, UF, Miss. State, Mizzou, UGA, South Carolina and Bama.
    The odds of UT losing going 0-7 during that stretch are 1.9 percent.
    If there’s a math professor out there who can help me determine other probabilities, please let me know, but I think we would need to have some sort of engine that would “play” the games 50,000 times to determine the odds of UT finishing above or below a certain threshold.
    — Evan

Comments are closed.