Ill take the over on 3 wins for any amount of money this writer wants to wager.
That's not exactly what those ESPN gomers are saying.
The FPI (a computer program, not a human...though designed by humans) actually projects we will win somewhere between 4 and 6 games (technically, between 3.8 and 6.2).
How can it do that if it has us the underdog in all but 3? Math.
Take an imaginary team. Call them the Vandybuilt Toiletdoors. Say they have a 49% chance of winning 9 of their 10 games, and a 0% chance of winning the last game, against the Tennessee Volunteers. So you'd say they're picked to lose all 10 games and will go 0-10, right? Nope.
Because those first nine games are almost 50/50 splits (slightly below, but only slightly), in reality the FPI would be saying that the Toileteers will split the 9. So they'd give them a 4.5 win projection for the season. Then they hedge even that, and predict something between 3 and 6 wins.
And that's how it goes for our "projected season" in FPI, too. We have some games that the FPI thinks are basically a toss-up, like vs USCe (46.8% is pretty close to 50/50) and Kentucky (48.1%). So if we split those, both technically having "picked us to lose", we're no longer 3-7 but 4-6.
And on it goes, with "partial win" projections for every other game that at first you'd say they have us losing.
The math used by these prognosticators (prob & stats) doesn't deal in certainties. It deals in probabilities. To read anything they say as a "certain prediction" is to misread it.
Nonetheless, they're a bunch of gomers and they're wrong. We're winning 7 this year.
Go Vols!
Tennessee 2020 FPI - Volunteers - ESPN
EDIT and p.s. Thinking through why the FPI might project us as underdogs against teams we beat last year and against whom we have a significant talent advantage. I mean, USCe and Kentucky in particular. But also why they have us such significant underdogs against a team we beat just a couple of years ago (Auburn) when we had a brand new coach who now has 3 years experience, and a long track record beating Auburn's coach. Those three projections in particular didn't make sense to me, so I got to wondering why.
The FPI is just a program. It's only as good as (a) its algorithms and (b) the data fed to it for the upcoming season. The algorithms aren't bad, we already know that. FPI has done about as well as Vegas over the past few years projecting winners. So the algorithm, which ESPN's statisticians are constantly fine-tuning, is probably okay. That means the data must be bad.
So what bad data could the statisticians be throwing into the program about Tennessee? I think it's this: I think they're believing every word Jeremy Pruitt says.
All through July and August, and now into September, Jeremy has been pretty open about our challenges. He's not quite Lou Holtzing it (remember how Holtz always bad-mouthed his own team and talked up the opponent to be god-like?), Jeremy's not doing that, but he is spending significant time talking about the negatives. The DLine that just won't meet expectations. The QBs who need to become more consistent. The covid contact tracing that's keeping everyone out of practice. And on and on.
I think ESPN are believing him. I mean, I think they're inserting some "confidence factor" kind of adjustment in favor of Tennessee's opponents because of what they're hearing CJP tell the world.
But CJP just talks that way. He tells it as he sees it, sure, but he tends to harp on what needs to be improved rather than what's going well. Saban does the same thing. But the ESPN goombas have gotten used to Saban, and tune it out. They're not yet tuning it out for Pruitt.
That's my orange-colored theory, anyway. I think they're underestimating the Vols, because they're taking our coach too completely at face value.
We'll see starting this weekend.
Go Vols!
