🏈 another point of view. what do the stats show?

[pop]

lets take a look at the trickle down system

Lets see how precise your season long predictions have. Lets call a blowout 17 pts or more, because that seems to be what you claim to be able to predict. Earlier in this thread you predicted close to such an event in the upcoming game (Utah <14>31, which is 17 or more point difference). In your predictions in 13 games you predicted 11 blowout victories for Alabama. In those 11 games where your system predicted a blowout of 17 pts or more, you were only correct 6/11 times, not any better than chance at predicting a blowout. These predictions also include glaring evidence of the bias inherent in your system, a predicted 21 point victory over the Florida gators. A florida team that against a much more difficult schedule amassed a statistical lead in every category over alabama, yet the trickle down analysis system somehow gave Bama the nod by 21 points. In fact that prediction of bama31-utah14 is giving Utah more credit than you gave florida. Hmm. What could account for that difference? Florida playing a great game? Maybe or is it more likely your system of analysis is inherently flawed and biased. The data should show the bias. Lets take a look. You averaged over-predicting the outcome in all 13 games by an average of 8.3pts per game this season alone. In the 7 games that were not blow outs (<17pts, Tulane, Georgia, Kentucky, miss, ten, lsu, fla) your system over predicted by 17.1points. What was the score you predicted between Utah-bama? a 17pt difference? Couldn’t be? No way. You system over predicts anywhere from 8pts against bad teams, and 17pts against teams that were able to compete with alabama. If you take out the inherent bias in your system for your prediction of the sugar bowl, the result would be a close game (8-17pts of pure bias). Kind of like what the actual data show, two statistically similar teams that will place a very close game.

 

[Guest]

I say just let the teams take care of it on the field. Statistics are useless in comparisons.

 

[ericd]

Stats are for chumps.

BTW, Ya'll have way too much time on your hands.

 

[TerryP]

Stats are for chumps.

BTW, Ya'll have way too much time on your hands.

Or perhaps, you have too many requests from the wife you never get finished...or boring hobbies...or...

:wink:

 

[ericd]

Stats are for chumps.

BTW, Ya'll have way too much time on your hands.

Or perhaps, you have too many requests from the wife you never get finished

:wink:

If you only knew.

Look all these stats don't mean jack. The only way you could really measure this match up is if both teams played the exact same teams, on the same fields, under the exact same conditions, with all the exact same players on every play.

The past doesn't matter. Only what's gone into preparation for this game and what's in these kids' heads that night. I know there's not much else to talk about, so folks are reaching, but, WOW! Either some of these guys posting have no life or have an incredible amount of time to sit in front of a computer. I wonder if they've ever even picked up a real football?

Anyway, I'm not trying to be mean spirited, just amazed.

 

[TerryP]

I understand what you are saying Eric, and I tend to agree up to a point.

Looking at stats isn't a be all end all to match-ups. But, in some situations it's a pretty good indicator of a lot of things.

Just as an example, when I looked at the Utah defense I noticed the number of tackles per unit and the disparity there was between the DE's and the DT's.

Tells me a couple of things...good pursuit, DE's spend some time in the backfield. When I look at those stats versus teams played you can start seeing areas where yards can be gained. In this case, you don't see their DT's making a lot of tackles which pretty well indicates that's a weak link in the running game.

Not a stretch, by any means, wouldn't you agree?

Now, after watching quite a few of their games I've seen where those stats match what they've fielded. But, those stats also led me to watch the DT's in particular while watching those games as well.

 

[Guest]

I say just let the teams take care of it on the field. Statistics are useless in comparisons.

You seemed to have condensed what I meant.

 

[pop]

\

Oh...as for comments that you will substituted 2nd and 3rd stringers on the OL who are bigger...I hope you do. There is a reason they are playing 2nd and 3rd string on a MWC team.

If Alabama plays its worst game of the year and Utah its best, it could be hairy.

If both teams have an average outing, Bama wins by 21.

If Bama plays well - it won't matter what you do, we win by 30+.

By halftime, there will be a lot of Utah fans doing the facepalm.

uh oh FACE PALM!

jk, it was just too easy. You guys played a great game. Your fan base is a class act, and Saban will take you guys far.

Go Utes!

 

[imported_Porter]

\

Oh...as for comments that you will substituted 2nd and 3rd stringers on the OL who are bigger...I hope you do. There is a reason they are playing 2nd and 3rd string on a MWC team.

If Alabama plays its worst game of the year and Utah its best, it could be hairy.

If both teams have an average outing, Bama wins by 21.

If Bama plays well - it won't matter what you do, we win by 30+.

By halftime, there will be a lot of Utah fans doing the facepalm.

uh oh FACE PALM!

jk, it was just too easy. You guys played a great game. Your fan base is a class act, and Saban will take you guys far.

Go Utes!

Damn.

 

[Guest]

heh. Nice one. Y'all earned it.

I admit the Florida prediction was based entirely on homerism, not matchups. The others were detailed breakdowns. When I did the actual breakdown, I came up with a 7 point Florida win being the most likely score (delta 14), but after picking 12 wins in a row, I was not going to jinx us.

Two misses were statistical anomalies in the early season. One was after a defensive injury and a second half letdown.

And I still think you miss the point, I don't predict the exact score. I predict a range based on statistical matchups - and then give the average as most likely prediction. Utah was hard as crap to break down because there were so many degrees separating our competition. No same-opponents and few similar.

My final average was 29-10 (fudged from 29-17) which was based on a range from 22-24 to 36-10. That was not accounting for the additional shuffling on the OL. That would have given another touchdown away and made the range 15-31 to 22-22 toss up with Utah favored...but you cannot predict injuries.

...and some of my commentary of Utah was for no other purpose than stirring you folks up. I made a comment to others about it in Private messages.

Utah won and the contributing factors have been beaten silly.

Hey, Bama site, Bama fan, what do you expect?

If you can find one instance of me posting on a Utah forum that we were going to win, you get a pink cupie doll.

You also don't understand what "Trickle down" refers to. It is based on the premise of talent saturation - that there are more high level players than scholarships available and the only accurate indicator of spreads is based on specific match-ups. The Utah's of the world can beat the Alabama's of the world on any given day if there are no specific matchup issues...and ours got taken away by a combination of injury and scheming by the Utah coaches. Good job there. I don't think Utah could hang with Florida or USC, because there are significantly more matchup issues. Too bad there is no way to find out.

If I could vote a #1, it would be Utah based on beating the number of top 25 teams you have - with no losses. No other team can claim that. If there was a playoff, we could find out who the best team is.

Good luck and congratulations on a great season.