UT Georgia computer model prediction

#51
#51
That's funny, because I think if UGA can't stop our offense, they will get beat by at least 2 scores. They had a rough time with Florida last weekend and Florida doesn't have nearly the sort of playmakers on offense that we do. Despite UGA's numbers, they don't have the offense to hang with us. Stetson Bennett is not Bryce Young.
 
#52
#52
GBOplayer said:
ESPN Computer's Prediction For Georgia vs. Tennessee

A computer simulation using advanced metrics predicts that Georgia will win easily, 75% chance that Georgia will win....I guess that's why we are almost 2 touchdown underdogs
Click to expand...
A computer simulation didnt figure / rank Tennessee to be un the top 25...go figure. The scoreboard will tell who won Saturday. Got my money on Vols...for more reasons than im a fan
 
  • Like
Reactions: lifeisdeep
#53
#53
VolFan4Life87 said:
In the article: “For Tennessee, a loss would probably spell the end of their national title run.”

Why? If UGA does win and UT wins the remaining games on the schedule, who the F is the better one loss team than UT?
Click to expand...

The problem is we would have to depend on at least 1 of Clemson/TCU losing. And even then we'd also have to depend on UGA beating Bama in the SECCG (assuming Bama wins out). The committee won't leave out 1 loss UGA team if they lose in the SECCG to Bama.

Even then, a 1 loss USC, Clemson, or TCU that wins their conference would likely be selected before us. The committee values conference champions. Like it or not, the loser of this weekend is likely out of the playoffs unless some other teams fall out of the picture.
 
#57
#57
Thus the difference between Virtual and Reality. Pmuters don't play the game. TN by 24. Kirby goes Ballistic.
 
  • Like
Reactions: VolRage
#58
#58
313402808_10160108418561826_8837429183694168395_n.jpg
 
  • Like
Reactions: captennmike, 90 Vol, Beagle Bill and 4 others
#61
#61
The toughest SEC opponent they’ve faced has been Florida and they allowed 371 yards and 271 passing. I really think we can score on their defense!
 
#64
#64
#65
#65
GBOplayer said:
ESPN Computer's Prediction For Georgia vs. Tennessee

A computer simulation using advanced metrics predicts that Georgia will win easily, 75% chance that Georgia will win....I guess that's why we are almost 2 touchdown underdogs
Click to expand...
The major problem with computer models in CFB is that they either do or don't use transitive principles. Doing so might make the computer more impressed with UGA's win over Kent St (3-5) than UT's win over Akron (1-6). At a minimum they'd get more "points". But those are two very similar teams whose game was decided 33-27 this past weekend.

You do that over 8 games and it compounds itself. Realizing that's true... the programmers have to introduce "corrections" which may help in some ways but introduce additional bias in other ways.

I don't think this game is very predictable but if I gambled I would have jumped all over that early spread. One of the reasons it is so unpredictable is that no one has actually stopped or even slowed UT's offense. We don't know where the upper limit is. The same is true of UGA's D to an extent. They've allowed no more than 22 points this season.

Flip the ball and the computers will be a lot less impressed with UT's D vs UGA's O... but the O most similar to UGA's that UT has faced is UK. UT's improving D throttled them.

Something will break. It could be UGA's OL vs UT's D front like Mizzou did to them. It could be UGA's D vs UT's O just on shear talent. But there isn't an obvious answer.
 
  • Like
Reactions: Arrested Devolopment
#66
#66
If UF, LSU, UK, nor Bama could figure out how to stop our offense, much less slow it down, then I don't thing UGa will all of a sudden find the answer. This is not last years team. Even though we were up 10-7 at the end of the 1st qtr on UGa. The pressure is on them not us.
 
#67
#67
DiderotsGhost said:
This isn't true at all. The computer models had us as a big favorite over Kentucky. I only looked at a couple, but I know Sagarin had us favored by over 12 points and hasn't been "high" on Kentucky all season; in fact, the computers have never liked Kentucky as much as the media this year.
Click to expand...
Looks like Sagarin has it somewhere between 5 and 8 points.
 
  • Like
Reactions: DiderotsGhost
#69
#69
Hard to get meaningful data for college football, I would think, so it’s amazing that these models are right as often as they are (I’ve read that it’s about 75%). ESPN FPI has us ranked behind Ohio State, Alabama, and Michigan as well as UGA. I have no doubt we would beat Michigan, and I would feel very good about an OSU matchup as well. Alabama we’ve already beaten. Good thing the game is played on grass and not silicon.
 
#71
#71
DiderotsGhost said:
This is hilarious when people try to discredit the most accurate models. These types of models are used by Vegas. And while they aren't perfect, they are far more accurate than the alternative of humans randomly making bad guesses. The good computer models get things right more often than not and they are certainly more reliable than random internet message board posters.

That said, Tennessee's rapid improvement this season is difficult to factor into those models. The team that barely got by Pitt in overtime would've gotten clobbered by Georgia. The team that beat Bama, obliterated LSU, and destroyed Kentucky has a shot. It's rare for a team to improve this much this quickly and the models have difficulty factoring that in.

Georgia should be considered the favorite. It's in Athens and they've destroyed everyone they've played this season with the sole exception of Mizzou. But we have a better chance than the models suggest if we play like we have over the past few weeks.
Click to expand...

Agree with everything except you left off Kent State. Georgia didn't destroy them either and really struggled.
 
#72
#72
DiderotsGhost said:
This isn't true at all. The computer models had us as a big favorite over Kentucky. I only looked at a couple, but I know Sagarin had us favored by over 12 points and hasn't been "high" on Kentucky all season; in fact, the computers have never liked Kentucky as much as the media this year.
Click to expand...

Sagarin's models are utter garbage until November (at least strength of schedule rankings). He even admitted years ago that his formula doesn't work until teams have played their full slate.
 
#73
#73
BondJamesBond said:
I wonder what the computer said about the likelihood of undefeated Tennessee playing against undefeated Georgia. I bet it was less than 25%.

This team is special in part because they have somehow defied the odds in many competitions. Claiming to predict a team who you have been unable to predict for 8 weeks... let's just say the history of predictions for this team doesn't give me confidence that the modelers have a good handle on it. This offense is a unicorn and we don't have a long history of predicting unicorn behavior.
Click to expand...
Bet it was less than 5%.
 
  • Like
Reactions: BondJamesBond and mpfvol
#74
#74
GAVol said:
I think 7-8 is probably where it lands. Everybody pretty much knew that 11-12 was a little much. Defending national champs vs. 15 years of dumpster fire is a hard bias to reconcile in a betting line.
Click to expand...
Plus they get the 2-3 points for home field.
 
#75
#75
Just remember this same computer has Texas at 6. So this computer doesn’t really matter at all
 
New posts
You must log in or register to reply here.

VN Store



Back
Top