College Football Nerds model has Tenn 44 LSU 26

Well, there's that, guess I don't have to watch.

quote:

---

How often is their model right or close?

---

Idk. 18 points just seems ridiculous to me.

We don't have a qb, so it sounds about right.

My experience has been that it is hit or miss. It's largely based off of per play and per attempt statistics, which would probably favor Tennessee. That being said, the accuracy is partially dependent on common opponents and the size of the sample. We have no common opponents with Tennessee yet, so it's pretty much based on stats alone at this point. Generally, their score predictions are in the right neighborhood, although not necessarily strongly predictive of the outcomes.

quote:

---

per play and per attempt statistics,

---

What a dumb way to come up with an outcome. Totally worthless.

quote:

---

We don't have a qb, so it sounds about right.

---

There’ll be a ton of you internet-dipshit coaches rooting against LSU this weekend so you can be “right.”

I am over simplifying it a bit, but my recollection is that those statistics have a sizeable impact. Of course, they also don't share their model's inner workings, so it's a bit of an inference from things they say. But it's sure not bulletproof, and they readily disagree with the model at times.

quote:

---

We don't have a qb, so it sounds about right.

---

This would allow them to get 44 points?

The line has moved to +2.5 already. Seems like such a trap game for people taking Tenn. But then listening to them and their model, idk what to think