Utah State WR

#151
#151
#152
#152
titansvolsfaninga said:
I like your analysis. Numbers don't lie, however there are other variables at play here. If all else is equal, which it is not, I think we have a better coaching staff too.
Click to expand...

UT has more talent. Neither team has experience as a strength but it would be accurate to say that they have more older players with experience.

If UT's staff is better, UT should win.
 
Last edited:
  • Like
Reactions: 1 person
#153
#153
Madmartigan said:
I've heard this number a bit, but it's not accurate. Here are players that logged starts last year that return.

Offense:

1 QB: Chuckie Keeton, 2. RB: Joe Hill, 3. Slot WR: JoJo Natson, 4. LT: Kevin Whimpey, 5. TE: Wyatt Houston 6. WR: Ronald Butler

Defense:

1. OLB: Kyler Fackrell 2. ILB: Nick Vigil (started at OLB last 1/2 of the season) 3. ILB Zack Vigil 4. NT Elvis Kamana-Matagi 5.DE: Jordan Nielsen 6. DE: Jordan Nielsen 7. SS Brian Suite
Click to expand...

If having started games is the measure of being a "starter" then UT returns several that aren't usually mentioned. These are guys who have started games but we would not consider "starters": Swafford, Croom, Smith, D Young, M Jackson, K Kerbyson (I think), M Lane, J Williams, T Saulsberry,... If I remember correctly all of these guys have been credited with a "start".
 
#154
#154
sjt18 said:
If having started games is the measure of being a "starter" then UT returns several that aren't usually mentioned. These are guys who have started games but we would not consider "starters": Swafford, Croom, Smith, D Young, M Jackson, K Kerbyson (I think), M Lane, J Williams, T Saulsberry,... If I remember correctly all of these guys have been credited with a "start".
Click to expand...

I wonder what metric Phil Steele uses to determine who he deems a starter: I"m guessing it's someone who started more than half the games. I'd love to see total snaps played as a metric deployed. Total field time would be fun to measure.
 
#155
#155
sjt18 said:
Sorry. I meant for '13 or any year in the last 3 or 4. You've done a good bit of the leg work. I would like to do some margin of games analysis and focus on conference games only.
Click to expand...

You can scroll through many posts on my blog, I think you will find the answer for 2013, but I think that is far back as I have discussed (maybe 2012).

Understand that I have to keep "my" evaluations separated from anything that I might have been paid to do. That is why I won't give away raw data; there might be blurred lines between my raw data and that with which I have been given to use by others. My evaluations tend to stop well short of anything more advanced. I love my job and never want to be questioned about my ability to keep proprietary data safe.

You can also check CFB Matrix - Setting College Football's Expectations. He has been doing similar work for much longer than I have. I believe you will find that he has tried to answer the same question you are (he has done some game by game calculations for CFB) and to my knowledge hasn't shown anything stable enough to be predictive of score.

Again I stress that these sorts of evaluations exist, but when they rise to the level of being predictive enough to beat the spread, you won't find them on message boards or pushed free to the public. To get to the level of being predictive of score, thus useful against the spread, requires (at least in my experience) far more data than you are wanting to evaluate and an incredible amount of time and effort ($$$) to parse it all.

That is just my two cents. Good luck on your search. I hope you find the answer and I can't wait to read about it when you do.
 
Last edited:
  • Like
Reactions: 1 person
#156
#156
titansvolsfaninga said:
I like your analysis. Numbers don't lie, however there are other variables at play here. If all else is equal, which it is not, I think we have a better coaching staff too.
Click to expand...

Well without considering anything else, you can arrive at a roughly 70/30 ratio of wins to losses using no other data.

You are right, coaching makes a difference. In fact, we have a coach with a history of over-performance. Jones averages 3 "extra" wins a year over his 7 year history in relation to talent.

Jones also has a history of increasing the recruiting averages that he inherited. These two facts combined tend to put to rest the hypothesis that Jones only rode Kelly's coat-tails.
 
Last edited:
  • Like
Reactions: 2 people
#157
#157
daj2576 said:
You can scroll through many posts on my blog, I think you will find the answer for 2013, but I think that is far back as I have discussed (maybe 2012).

Understand that I have to keep "my" evaluations separated from anything that I might have been paid to do. That is why I won't give away raw data; there might be blurred lines between my raw data and that with which I have been given to use by others. My evaluations tend to stop well short of anything more advanced. I love my job and never want to be questioned about my ability to keep proprietary data safe.

You can also check CFB Matrix - Setting College Football's Expectations. He has been doing similar work for much longer than I have. I believe you will find that he has tried to answer the same question you are (he has done some game by game calculations for CFB) and to my knowledge hasn't shown anything stable enough to be predictive of score.

Again I stress that these sorts of evaluations exist, but when they rise to the level of being predictive enough to beat the spread, you won't find them on message boards or pushed free to the public. To get to the level of being predictive of score, thus useful against the spread, requires (at least in my experience) far more data than you are wanting to evaluate and an incredible amount of time and effort ($$$) to parse it all.

That is just my two cents. Good luck on your search. I hope you find the answer and I can't wait to read about it when you do.
Click to expand...
I hope you have not misunderstood the whole time.

I am NOT saying that you can say that team A will beat team B by X number of points.

I am saying that based on talent, it is likely that UT beats USU by more than 17 pts. I think there is better than a 50% chance that's the case. There's a chance but not a likelihood that USU wins or that they keep it closer than 17.

Very, very seldom do I say something "will" happen and those occasions usually involve me having a great deal of personal and direct control.
 
#159
#159
sjt18 said:
I hope you have not misunderstood the whole time.

I am NOT saying that you can say that team A will beat team B by X number of points.

I am saying that based on talent, it is likely that UT beats USU by more than 17 pts. I think there is better than a 50% chance that's the case. There's a chance but not a likelihood that USU wins or that they keep it closer than 17.

Very, very seldom do I say something "will" happen and those occasions usually involve me having a great deal of personal and direct control.
Click to expand...

We seem to be talking past each other.

That is fine.

Everything is available for you to understand my position, and vice versa. Let's not corrupt a whole message board arguing about your 17 point spread.
 
  • Like
Reactions: 1 person
#160
#160
vols4evah said:
thats a nice website.

is it yours ? edit: ok NM I guess thats your Blog.

Very well done, I enjoy reading it and have it book marked now.
Click to expand...

Thank you. It is "mine" but it wouldn't exist in the form it is now if I didn't have some amazing contributors who volunteer their time and viewpoint.

We sometimes disagree on our view, and that tends to confuse many readers who believe that a website should have a uniform editorial flavor. I tend to believe that our disagreements often lead to a better view of the reality affecting our Vols if one can get past the rare contrast of opinion.
 
Last edited:
  • Like
Reactions: 1 person
#161
#161
daj2576 said:
Thank you. It is "mine" but it wouldn't exist in the form it is now if I didn't have some amazing contributors who volunteer their time and viewpoint.

We sometimes disagree on our view, and that tends to confuse many readers who believe that a website should have a uniform editorial flavor. I tend to believe that our disagreements often lead to a better view of the reality affecting our Vols if one can get past the rare contrast of opinion.
Click to expand...

Have you used your model to predict the NC winner and each league winner? That could be a fun exercise.
 
#162
#162
Madmartigan said:
Have you used your model to predict the NC winner and each league winner? That could be a fun exercise.
Click to expand...

I don't want to step that far out of my comfort zone yet. I tend to believe that those milestones, at least the championship game, rely too heavily on voter subjectivity. I haven't spent the time to try to model that although I have read some data that suggests it is predictable behavior.
 
New posts
You must log in or register to reply here.

VN Store



Back
Top