(04-27-2015 11:18 AM)DrGonzo Wrote: [ -> ]http://cfbmatrix.com/returning-starts-2014-all-fbs/
The Belt
12: Appalachian State
16: Georgia State
23: NMSU
28: stAte
91: Georgia Southern
91: Troy
94: Idaho
101: ULM
113: ULL
117: Texas State
124: South Alabama
Still want to know what this list looked like in 2014 and compare to final win loss record.
App's is probably distorted by injuries and flipping different people on and off the OL.
(04-27-2015 11:30 AM)EigenEagle Wrote: [ -> ]App's is probably distorted by injuries and flipping different people on and off the OL.
We would have fallen significantly in this ranking if we'd kept the same starting roster through the entire season that we had in game one. I wonder if we would overtake Temple if we'd done the opposite.
It obviously bodes a whole lot better for teams who played well last year.
App's starting lineup didn't change much the last six games, and only included two seniors (CB and OT).
Our Dline starts of 8 is inaccurate.
We had one senior on the two-deep at DE and he started 6-8 games. That number should be 24 at a minimum.
A-State lost 2 guys that matter. Q Lee and Artez Brown.
We're loaded - deeper and more talented at all positions except those two.
(04-27-2015 12:07 PM)Saint3333 Wrote: [ -> ]App's starting lineup didn't change much the last six games, and only included two seniors (CB and OT).
Our Dline starts of 8 is inaccurate.
We had one senior on the two-deep at DE and he started 6-8 games. That number should be 24 at a minimum.
Deuce was the senior, correct?
That's is encouraging. Thanks.
(04-27-2015 11:27 AM)panama Wrote: [ -> ] (04-27-2015 11:18 AM)DrGonzo Wrote: [ -> ]http://cfbmatrix.com/returning-starts-2014-all-fbs/
The Belt
12: Appalachian State
16: Georgia State
23: NMSU
28: stAte
91: Georgia Southern
91: Troy
94: Idaho
101: ULM
113: ULL
117: Texas State
124: South Alabama
Still want to know what this list looked like in 2014 and compare to final win loss record.
This or over significant time period as a large sample size would show if there is any correlation to returning starters and winning %. I am sure there probably is but not as strong as most think. Losing two or three studs over five or more starters could be worse depending on the talent level returning. Sometimes it is addition by subtraction. I would rather have a returning stud QB rather than losing a couple of average linemen.
Looks like the OL and WR position is what dropped us so low.
Let's just say the "expert" statisticians that put out preseason predictions use this metric. They certainly get rankings wrong, but if they didn't see a correlation they wouldn't continually use this metric.
I think you have to take the prior year standings and W/L record into account as a starting point and then move that team up or down based on attrition and recruiting for this year's prediction. This is just one of a dozen data points.