sfink16
All American
Posts: 3,571
Joined: Jul 2013
Reputation: 73
I Root For: Temple
Location: Dubois, Pa
|
RE: Rank'em...
(10-01-2017 11:04 AM)slhNavy91 Wrote: (10-01-2017 10:44 AM)sfink16 Wrote: (10-01-2017 10:27 AM)slhNavy91 Wrote: (10-01-2017 10:15 AM)sfink16 Wrote: (10-01-2017 09:24 AM)NestaKnight1 Wrote: You mean those same "National" polls that had FSU ranked 2nd? Ok. I'm not sour against USF, the fact is they have played the weakest schedule in the conference. Houston, SMU, Navy, Memphis and probably others would be undefeated if they had played USF's powderpuff schedule. USF's schedule is hurting their perception, it's that simple, no rationalization needed. That being said:
1) UCF
2) USF
3) Navy
4) Memphis
5) SMU
6) Houston
7) Tulane
8) Temple
9) Cincy
10) ECU
11) Tulsa
12) UCONN
The FSU correction has been made and is past and still may have a decent season by the end considering the superior talent they recruit. Apparently you like living in the past. You are also blinded by the facts.
Lets look at Houston for example. They have played one team with a winning record in Texas Tech. Had they beat Texas Tech, they wouldn't have a winning record. Houston and USF have one common opponent in Temple. Temple was in the Houston game until the end but was completely overpowered and blown out by USF. Like Houston, USF played a similar (slightly less it can be argued) team to Texas Tech in Illinois. Had Illinois won, they'd be 3 and 1, like Texas Tech is now, instead of 2 and 2.
SMU is a good team but has played one good team in TCU and lost badly, giving up 51 points. That 51 points is more then half the total points USF has giving up. How can a team with that kind of defense be mentioned in the same breath as USF when they failed so miserably against TCU? Would USF beat TCU? Who knows, but they certainly couldn't do any worse.
Navy has played a worse schedule then USF, not one power team on their schedule, 3 lower AAC teams and a G4 team.
Memphis best defensive effort came against La Monroe, giving up 29 points, while giving up 31 to an FCS foe, and just got blown out by UCF. Hard to mention them in the same sentence as USF either.
The only AAC team, up to this point worthy of mentioning with USF is UCF. All the others do not have great resumes.
Check yourself.
Massey Composite has Navy's opponents 96,73,80,95
USFs opponents 122, 85, 84, 112 for FBS, and let's give you Sagarin's137 for your FCS. Your P5 is below 2 of our 3 AAC wins you deride, just as are your conference wins.
Average opponent ranking Navy 86, USF 108
Any rating system is a work in progress at this time of year making it crazy to use as gospel. Some teams have played as little as 3 games while other have played 5. Assuming both those teams are undefeated, the team that played 3 games played only 60% of the team that played 5 games. The team that played 4 games only played 80% of the games. This type of disparity makes the system useless at this time of year.
Further, the team that played more distributed more losses to the Massey system then the team that played less. Once you get to 11 or 12 games, then the Massey system may make more sense. Then if a team played 1 or 2 less games, the percentage difference is much smaller then 60% or 80%.
In week 5, Massey is apples and oranges statistically.
You don't math much, do you?
I was using the COMPOSITE rankings which Massey compiles on his site. An appropriate response to that is that I'm flawed to use the composite on Sunday morning when it is only compiling 14 different rankings; it will be better value later in the week.
As far as 4-5 games being a smaller sample than 12 games, well duh. But the results to date are the sample we have.
Given that, I challenge you to find ANY SoS calculation that has USFs SoS better than Navy's.
You will have better luck finding a data driven ranking that has Illinois higher than Tulane and Cincy,but its a mixed bag.
Some of those ratings systems use data from the previous year to determine early year ratings, since several of the rating systems are looking for trends as predictive models. Using previous year data is one way with helping determine trends. How else would they be able to predict games with such a small sample of data? The smaller the sample, the less accurate they are.
So yes, they are not terribly accurate this early in the year. That said, while USF hasn't gone through murders row in schedule, neither has Navy, using the data you provided. It's hard to imagine why USF would beat the very same teams on Navy's schedule.
|
|