Updated Odds of Winning Out: 68.1%

[DiderotsGhost]

Here are our updated odds of winning out per ESPN's statistical projections.

vs South Carolina: 90.7%
vs North Texas: 99.6%
@ Missouri: 80.4%
vs Vanderbilt: 93.7%

Odds of Winning all 4 Games: 68.1%

Don't want to get too excited, because we all know that one of the three SEC games remaining will probably be closer than we expect. But certainly, if the early part of our schedule ended up being tougher than expected (with Oklahoma being better than last year and Florida being resurgent), then the latter half looks to be easier than expected preseason.

I also like that Mizzou is after that North Texas game. North Texas is atrociously bad this year. Sagarin has them rated below our I-AA opponent Western Carolina. We might be able to give some of our starters some rest vs North Texas, helping keep us in better shape for Mizzou and their tough D.

 

[daniel3417]

I don't want to overestimate Missou, either. They are very bad. We need to not let them even feel they belong on the same field on 11/21.

 

[LA Vol]

Oh, let's just go ahead and say 100%. Spurrier is gone, Mauk is gone, North Texas has no coach and only one win and we have an on-the-ropes Vandy at home to end the season.

 

[1orangews6]

I think we have a great shot of winning out but I like what I hear from the players in their interviews stating that they are playing one game at a time. If the staff can keep them grounded and focused on one game at a time then I think we will have a great season. I have always that that 8-4 is where we would end up. I think we could definitely have done better though.

 

[Biopsyman]

N Texas..LOL

 

[joevol33]

100% period.

 

[eskimovol]

I don't want to overestimate Missou, either. They are very bad. We need to not let them even feel they belong on the same field on 11/21.

My only concern about mizzou is if their d-line will make us play down to their level like last year. Not scared of the Mizzou offence at all.

 

[volstuff37641]

cain't figure,% against all 4 teams are above 80%,yet to win out only 68%.

 

[DiderotsGhost]

I don't want to overestimate Missou, either. They are very bad. We need to not let them even feel they belong on the same field on 11/21.

I think they are too well coached for that.

They are one of the more bi-polar teams I can remember. Reminds me of the 2005 Vols team where the defense was spectacular and the offense was pathetic. I'd rate Mizzou's defense an "A" and their offense would get something like a "D-" (and even that might be generous).

The main reason that game has potential to be problematic is turnovers. If we allow Mizzou's defense to get points on turnovers (like we did vs UK and UGA), and then we struggle to put up many points on them, it could be a scary game. I'd view taking care of the ball as priority #1 against them. We can probably hold their offense to virtually nothing, but if we let their defense score on us, we could be in trouble. They haven't given up more than 21 points all season.

On the other hand, they haven't faced a particularly prolific offense this season yet. They have Miss State next week, which should be interesting to see how well they can defend against a team that can score.

 

[sjt18]

More of the same non-sense. UT is better than each of the last 4 opponents just like they were better than the last 5 last week when some people here were "worried".

UT SHOULD win out. The schedule was heavily front loaded... almost like two different seasons in two different leagues talent wise.

I told folks here that the UK game shouldn't be close... it wasn't. None of these last games should be close with the possible exception of Mizzou who seems able to keep things from getting out of hand with their D. OTOH, they haven't really faced a good O all year. UGA was fresh off losing Chubb.

We'll know more about their D after the next two weekends when they play MSU and BYU. There's a good chance they're on a 5 game losing skid while UT will be coming off of a warm up game against one of the worst teams in CFB.

 

[VFL-82-JP]

cain't figure,% against all 4 teams are above 80%,yet to win out only 68%.

Chance of flipping "heads" once is 50%. Chance of doing it twice in a row is only 25%.

Same thing here. Just the individual event percentages are different.

 

[GBO97]

cain't figure,% against all 4 teams are above 80%,yet to win out only 68%.

The chance of winning all 4 is the 4 chances of winning each individual game all multiplied. Simple rithmetic bro

 

[VFL-82-JP]

More of the same non-sense.

SJT, just because you don't understand probability and statistics, that doesn't mean it isn't an accurate way of predicting future events.

Diderot's math is right. I posted exactly the same thing a few hours ago in another thread. It's factual, and accurate.

Go Vols!

 

[sjt18]

cain't figure,% against all 4 teams are above 80%,yet to win out only 68%.

I believe this is a misapplication of statistics.

If they were going to make this legit then they would have to go back to the original "probabilities" which they had absolutely no accurate means of determining and then predict the number of games UT would win. Most likely... it would have been 8 or 9.

I believe the best application of these probabilities which are suspect in and of themselves is to say the probability of winning out is equal to the lowest win probability... which would be 80.4%.

It is difficult though to determine how they derived at that number for MU when they've already been beaten by UK and VU... and UT's comparative strength to UK's is pretty fresh in our minds, right?

 

[DrDave]

cain't figure,% against all 4 teams are above 80%,yet to win out only 68%.

Prob win each of the games given as:
vs South Carolina: 90.7%
vs North Texas: 99.6%
@ Missouri: 80.4%
vs Vanderbilt: 93.7%

Prob win ALL 4 is .907 * .996 * .804 * .937 = .681 :hi:

 

[maotai]

cain't figure,% against all 4 teams are above 80%,yet to win out only 68%.

If there is no guaranteed to win one game that means the chances of winning multiple games decreases as the number games increases.

So if you have a 90% chance of winning one game but a 95% chance to win the second game you only have a 95% chance of a 90% chance to win both games. So your chances of winning both games are 0.90*0.95=0.855=85.5%

So for the espn percentages listed:
0.907*0.996*0.804*0.937=0.680553489456 rounded to the tenth of a percent = 0.681 or 68.1%.

 

[sjt18]

SJT, just because you don't understand probability and statistics, that doesn't mean it isn't an accurate way of predicting future events.

Diderot's math is right. I posted exactly the same thing a few hours ago in another thread. It's factual, and accurate.

Go Vols!

I understand just fine. I understand the technique they're trying to apply. There's nothing wrong with the calculations themselves. They simply do not apply. Football isn't a random set of events. Football is subject to willful decisions and controlled variables. If the games' outcomes were determined by meeting at mid-field with 100 cards marked for each team proportionate to those probabilities... then it would be a good use of that method.

Again, the only way this would apply is if you could accurately compute probabilities at the beginning of the season and produce a projected win total.... that total would have likely been 8 or 9. Even then... it would only apply generally across all teams and would be virtually useless to predict a particular team.

Also, what are the numbers behind the numbers? Where does the 80% for MU even come from? We ARE talking about a team that was beaten by VU and UK. Their best win is 3-5 USCe?


Let me explain why this is a poor use of that method. Daj2576 has a matrix that is "generally" predictive of outcomes using Rivals' rankings. So the team with more talent wins on avg 70% of the time and loses 30% of the time. But let's apply this method to Bama's chances of winning 10 games which they have in every season since 2007. Their chance of winning 10 games would be .7 e10... or just under 3%.


Bottom line... it simply doesn't work in the way it is being applied.

 

[Other_Guy]

Only UT worries me in these contests. Even.before the UK game we all agreed none of these teams belong on the same field right now.
That.being said, UT still starts slow and gives away freebies. Mizzu getting a couple early TOs and 3 n outs, if they capitalize UT will be.forced to throw it more and could be in trouble.
Still, comparing teams UT should beat SCjr 38-23, Mizzu 20-10 and Vandy 27-13 at the.very least.

 

[maotai]

I understand just fine. I understand the technique they're trying to apply. There's nothing wrong with the calculations themselves. They simply do not apply. Football isn't a random set of events. Football is subject to willful decisions and controlled variables. If the games' outcomes were determined by meeting at mid-field with 100 cards marked for each team proportionate to those probabilities... then it would be a good use of that method.

Again, the only way this would apply is if you could accurately compute probabilities at the beginning of the season and produce a projected win total.... that total would have likely been 8 or 9. Even then... it would only apply generally across all teams and would be virtually useless to predict a particular team.

Also, what are the numbers behind the numbers? Where does the 80% for MU even come from? We ARE talking about a team that was beaten by VU and UK. Their best win is 3-5 USCe?


Let me explain why this is a poor use of that method. Daj2576 has a matrix that is "generally" predictive of outcomes using Rivals' rankings. So the team with more talent wins on avg 70% of the time and loses 30% of the time. But let's apply this method to Bama's chances of winning 10 games which they have in every season since 2007. Their chance of winning 10 games would be .7 e10... or just under 3%.


Bottom line... it simply doesn't work in the way it is being applied.

Sigh...

 

[DiderotsGhost]

I believe this is a misapplication of statistics.

...

It is difficult though to determine how they derived at that number for MU when they've already been beaten by UK and VU... and UT's comparative strength to UK's is pretty fresh in our minds, right?

Sjt, not to rag on you, but I don't think you understand how probability models work. I play poker tournaments occasionally and this stuff is pretty common in poker.

Let's say you go all-in on 5 different hands and get called, each time with an 80% chance of winning. What that means is that the moment you lose, you're out of the tournament. Here are you odds of advancing to the 5th hand and winning.

Win 1st hand: 80%
Win 1st and 2nd hand: 64% (.8 X .8 = .64)
Win 1st, 2nd, and 3rd hand: 51.2% (.8 X .8 X .8 = .512)
Win 1st, 2nd, 3rd, and 4th hand: 41.0%
Win all 5 hands: 32.8%

Even though you are favored to win every hand, the odds of actually winning all 5 hands is relatively low at 32.8%.

This makes sense in college football, as well. After all, Alabama was probably favored to beat every team it played, and yet it lost 1 game. If you play enough games with a 90% probability of winning, you'll lose one eventually. On average, it should happen around the 7th game you play. A team that has a 90% chance of winning all 12 games it plays should, on average, probably finish the season either 11-1 or 10-2. Though, sometimes it will get lucky and go 12-0 or get unlucky and go 9-3 or 8-4.

I understand just fine. I understand the technique they're trying to apply. There's nothing wrong with the calculations themselves. They simply do not apply. Football isn't a random set of events. Football is subject to willful decisions and controlled variables. If the games' outcomes were determined by meeting at mid-field with 100 cards marked for each team proportionate to those probabilities... then it would be a good use of that method..

It's not a set of "random events", but it is a series of events that can be projected on past data and outcomes. How do we know this? The profitability of Las Vegas bookmakers is pretty good evidence of it.

Obviously, the data is not perfect and football is more dynamic than say picking a card out of a stack. But the same logic applies. A team with a 90% chance of winning every game does not have a 90% chance of going 12-0. That is simply bad math. It, in fact, has a 28.2% chance of going 12-0.

Could you argue that 90% was an inaccurate projection? Sure.

Can you argue that the laws of mathematics are suspended and magically conform to a rule you completely made up based on ignorance of math? You can, but you'll be wrong.

 

[Cartervol]

USCjr and Mizzou are all that's really left. Of the two Missouri is the one that you have to think will be tougher. SC has played a little better since Spurrier left, but they just don't have the horses.. Could be closer than expected but i kind of doubt it.

 

[VFL-82-JP]

Yeah, SJT, I like you man, and you are hugely smart in most areas we discuss on these boards, but you're showing in your "I understand" responses that you don't really understand

It's all very common sensical, once you take the time to learn how all the logic pieces fit together. And a LOT of different fields use the stuff, extensively. Not just Vegas, not just the markets, not just sports and games of chance, not just the military, not just medicine, not just...well, the list is all but endless. Every field with a scientific component uses prob & stats to help predict future events.

And it fits football just as well as many of those other "fuzzy" fields if applied properly. I see no reason to believe the stats guys at ESPN aren't applying it well.

Go Vols!

 

[volunteerfan56]

I believe this is a misapplication of statistics.

no

 

[volunteerfan56]

I understand just fine. I understand the technique they're trying to apply. There's nothing wrong with the calculations themselves. They simply do not apply. Football isn't a random set of events. Football is subject to willful decisions and controlled variables. If the games' outcomes were determined by meeting at mid-field with 100 cards marked for each team proportionate to those probabilities... then it would be a good use of that method.

Again, the only way this would apply is if you could accurately compute probabilities at the beginning of the season and produce a projected win total.... that total would have likely been 8 or 9. Even then... it would only apply generally across all teams and would be virtually useless to predict a particular team.

Also, what are the numbers behind the numbers? Where does the 80% for MU even come from? We ARE talking about a team that was beaten by VU and UK. Their best win is 3-5 USCe?


Let me explain why this is a poor use of that method. Daj2576 has a matrix that is "generally" predictive of outcomes using Rivals' rankings. So the team with more talent wins on avg 70% of the time and loses 30% of the time. But let's apply this method to Bama's chances of winning 10 games which they have in every season since 2007. Their chance of winning 10 games would be .7 e10... or just under 3%.


Bottom line... it simply doesn't work in the way it is being applied.

you obviously do not understand

 

[sjt18]

Sjt, not to rag on you, but I don't think you understand how probability models work. I play poker tournaments occasionally and this stuff is pretty common in poker.

I understand just fine. Poker is a game built on a finite number of possible outcomes wholly dependent on statistically random events (cards dealt). Football isn't that way.

The improvement made in practice this week isn't random nor is it equal across teams. Some teams lack the capacity to improve as much as others. There are differences in effort, motivation, etc.

Once you introduce purposeful, controlled actions... the model no longer applies at the team level.

This makes sense in college football, as well. After all, Alabama was probably favored to beat every team it played, and yet it lost 1 game. If you play enough games with a 90% probability of winning, you'll lose one eventually. On average, it should happen around the 7th game you play.

No. That is even MORE illegitimate. It should happen most likely against the opponent they have the least likelihood of beating regardless of where they appear on the schedule.

If they had even a 98% chance of winning 11 games.... then by the model you applied there is only a 56% chance they would do so. If we use the 90% you suggest above then it falls to 31%.

The model simply does not apply at this level. Too many variables are NOT random.

A team that has a 90% chance of winning all 12 games it plays should, on average, probably finish the season either 11-1 or 10-2. Though, sometimes it will get lucky and go 12-0 or get unlucky and go 9-3 or 8-4.

The method you are using does not have the ability to predict that... and "luck" has little to do with it.

It's not a set of "random events", but it is a series of events that can be projected on past data and outcomes. How do we know this? The profitability of Las Vegas bookmakers is pretty good evidence of it.

But they won't bet just UT's odds... if they bet it at all then they'll bet A LOT of teams to improve the level of "randomness" and to provide some control. If Jones and UT beat these "odds" then someone else doesn't. Ultimately, the house wins.

Obviously, the data is not perfect and football is more dynamic than say picking a card out of a stack. But the same logic applies.

We're chasing our tails and will not agree... but the same logic does NOT apply. No matter how practiced or prepared or talented or improving the "pickers" of the cards are... their probabilities of victory do not change. That is not true of football teams and coaches.

A team with a 90% chance of winning every game does not have a 90% chance of going 12-0. That is simply bad math. It, in fact, has a 28.2% chance of going 12-0.

No. But you will continue to believe otherwise so we might as well drop it.