Strength of Schedule -
Extra Point - 05-19-2014
This is strange. According to this site:
http://www.fbschedules.com/2014/04/2014-college-football-strength-of-schedule-ncaa-method/
the NCAA method for determining strength of schedule puts Stanford at number 42.
"The SEC leads with four teams in the Top 10 of the rankings Arkansas (1), Tennessee (3), Texas A&M (5), and Kentucky (6). The ACC is second with three teams Virginia (2), Syracuse (8), and Wake Forest (9)."
"The NCAA method is based solely on the opponents win/loss record from the previous season."
Add:
The part that I find strange is that anyone would think this is a valid metric. Is this for real? Could the NCAA actually consider this?
Re: Strenght of Schedule -
FarmBoy - 05-19-2014
That method of calculating SOS will always favor leagues that only play 8 games schedules. They trade a 50% chance of a loss against a league opponent for a 90% win against a creampuff.
Re: Strenght of Schedule -
martyup - 05-19-2014
Most notable is Alabama at 95. That is a deep hole to crawl out of to make the top 4.
Re: Strenght of Schedule -
yvonne - 05-20-2014
Is this the SOS that will be used to help determine the playoff participants?
Re: Strenght of Schedule -
stupac2 - 05-20-2014
(05-20-2014, 09:33 AM)Yvonne link Wrote:Is this the SOS that will be used to help determine the playoff participants?
Almost certainly not, since it's ludicrous on its face.
Re: Strenght of Schedule -
Papa John - 05-20-2014
I'm pretty sure that the NCAA will not use 2013 SOS to determine playoff participants for the 2014 season :P
But take a look at the upcoming schedule for Arkansas, #1 in projected SOS:
Auburn, 12-2
Nicholls State, 4-8
Texas Tech, 8-5
Northern Illinois, 12-2
Texas A&M, 9-4
Alabama, 11-2
Georgia, 8-5
UAB, 2-10
Mississippi State, 7-6
LSU, 10-3
Mississippi--I refuse to call it Ole M**s, 8-5
Missouri, 12-2
A few observations:
1. As FarmBoy ("as you wish") noted, the schools that play eight conference games are given an advantage, because they replace a tougher opponent with a creampuff that, in its league, is competitive and might have a decent overall record, such as...
2. Nicholls State, which is an FCS school. They have no business being included in the calculation. So Stanford gets penalized for playing 1-11 C.al, which would have no trouble beating Nicholls...nevermind.
3. Some teams that had great records last year, e.g. Northern Illinois, will not have as great a record this year.
4. Teams that win conference championship games will have an advantage over those that don't.
So, in my analysis, the conference that gets screwed the most by this formula is the Big 12. Like stupac2 said, this is a ludicrous way to determine SOS, and Texas, sitting at #26, and many others will scream bloody murder if the committee says that it will use it.
Re: Strenght of Schedule -
yvonne - 05-20-2014
(05-20-2014, 09:40 AM)stupac2 link Wrote:[quote author=Yvonne link=topic=10005.msg89776#msg89776 date=1400603580]Is this the SOS that will be used to help determine the playoff participants?
Almost certainly not, since it's ludicrous on its face.
[/quote]
OK, good. I was worried.
Re: Strenght of Schedule -
CTcard - 05-20-2014
(05-20-2014, 10:08 AM)Yvonne link Wrote:[quote author=stupac2 link=topic=10005.msg89779#msg89779 date=1400604036]
[quote author=Yvonne link=topic=10005.msg89776#msg89776 date=1400603580]Is this the SOS that will be used to help determine the playoff participants?
Almost certainly not, since it's ludicrous on its face.
[/quote]
OK, good. I was worried.
[/quote]
To chime in, and risk ruining the beauty of Yvonne's sarcasm, since when has using ridiculous formulas been a hurdle to selection committees?
Remember the NCAA invented the RPI, which is a version of this strength of schedule formulation, and now somehow thinks the RPI is a power rating.
To recap, the RPI is a simple formula: 25% your win-loss percentage, 50% your opponents' win-loss percentage, and 25% your opponents' opponents' win-loss percentage.
This is somewhat modified by adding a home/road adjustment. For baseball a home win counts as 0.7 wins, a home loss as 1.3 losses, a road win as 1.3 wins, and a road loss as 0.7 losses.
http://boydsworld.com/data/new_rpi.html
According to wikipedia, the basketball adjustment is 1.4 and 0.6.
http://en.wikipedia.org/wiki/Rating_Percentage_Index
Until 2011 there was also the "supersecret extra-component" of the RPI, which seemed to exist mainly so that nobody could call the NCAA on how they handled the RPI - since its secrecy meant nobody knew what a team's RPI was.
As far as I can tell, none of this is based upon any statistical, or other, analysis of anything.
I find it interesting that the NCAA got far more grief in the past using a rubric that included margin of victory than it does for using this simple, but obviously inadequate, measure to rate teams.
---------
And I will point out that the SEC-8 conference game schedule has a secondary benefit to this particular version of SOS. Since everyone else in your conference also gets an extra cream-puff - and a 90% chance of a win over an absolute 50% (averaging over the conference), that also gives all of your conference opponents a better win-loss percentage thus increasing your own SOS artificially.
Re: Strenght of Schedule -
stupac2 - 05-20-2014
I believe that the NCAA isn't making any statistics for the committee, and I know that ESPN is trying to make their new ones (which I think are based on Football Outsider's stuff) into the go-to stats for the committee.
Strength Strength Strength! -
martyup - 05-24-2014
For God's sake, this is The CARDboard. We should be strong spellers, or at least adequate editors.
Re: Strength Strength Strength! -
CompSci87 - 05-24-2014
(05-24-2014, 01:45 PM)martyup link Wrote:For God's sake, this is The CARDboard. We should be strong spellers, or at least adequate editors.
That was annoying me too. I've fixed the first post, and thus the thread title.
Re: Strength Strength Strength! -
martyup - 05-24-2014
(05-24-2014, 04:52 PM)CompSci87 link Wrote:[quote author=martyup link=topic=10005.msg90080#msg90080 date=1400964357]
For God's sake, this is The CARDboard. We should be strong spellers, or at least adequate editors.
That was annoying me too. I've fixed the first post, and thus the thread title.
[/quote]
You are a gentleman and a scholar.