Thursday, December 06, 2007

Why the SEC’s Out-of-Conference Schedule is Irrelevant

And, what the hell is Jeff Sagarin doing anyway?

SEC critics and “haters” often point to conference members “out-of-conference” schedules as some sort of valid critique of the conference’s lack of worthiness.

I’m going to say this in real simple terms to those of you who think this way –

It doesn’t matter who you play out-of-conference if your in-conference schedule is harder than anyone else’s.

Yet once again in virtually any measure of schedule strength, SEC teams are near the top of the standings. First let’s look at Florida, who going into this season was held up to criticism for playing Western Kentucky, Troy and FAU.

Florida’s Final Strength of Schedule Rankings

NCAA – 8th
Coffey Ratings – 2nd
Colley Matrix – 2nd
Massey Ratings – 2nd
Sagarin Ratings – 9th

Even with WKU, Troy and FAU, Florida’s average SOS per the above was 4.6.

Not that Troy was even that bad at all. Per the four computer rankers above, Troy was –

Coffey Ratings – 37th
Colley Matrix – 40th
Massey Ratings – 56th
Sagarin Ratings – 57th

Which makes Troy a better team (higher ranked) than the following according to Coffey and Colley -

UCLA, California, Georgia Tech, Oklahoma St, Louisville, Maryland, Purdue, Colorado, Arizona, Indiana, Nebraska, Washington, Washington St, Kansas St, Iowa, Miami, NC State, Stanford, Northwestern, North Carolina and of course Notre Dame.

Maybe the Pac Ten, Big Ten, ACC, Big East and Big 12 should think about scheduling Troy to improve their strengths of schedule.

(Sagarin doesn’t have them as well, but more on Mr. Sagarin later.)

Next let’s take a look at all the SEC teams and where they were ranked in strength of schedule (along with some notables) –


4. Mississippi
6. South Carolina
8. Florida
9. Kentucky
10. Alabama
19. Tennessee
24. LSU
29. Auburn
43. Vanderbilt
45. Mississippi State
65. Arkansas

39. Oklahoma
47. West Virginia
48. Ohio State
80. USC

Per the NCAA, every SEC team but one had harder schedules than WVU and OSU, and all of them significantly harder than USC.

Coffey Ratings SOS

2. Florida
4. South Carolina
4. Mississippi
6. Tennessee
10. Kentucky
11. Georgia
14. LSU
15. Alabama
19. Auburn
28. Vanderbilt
30. Mississippi State
39. Arkansas

And the same notables –

41. West Virginia
43. USC
52. Oklahoma
65. Ohio State

Per Coffey, not a single one of the BCS teams above had a schedule comparable to any SEC team.

Massey Ratings SOS

2. Florida
3. South Carolina
8. Kentucky
12. Tennessee
13. Vanderbilt
14. LSU
22. Mississippi
27. Alabama
37. Auburn
47. Mississippi St
53. Arkansas

The notables –

45. USC
46. West Virginia
49. Oklahoma
55. Ohio State

Once again the Buckeyes, who are playing for the Mythical National Title, didn’t have a schedule as hard as any single SEC team.

Colley SOS

2. Florida
3. South Carolina
6. Mississippi
11. Tennessee
8. Kentucky14. Vanderbilt
17. LSU
20. Alabama
27. Georgia
29. Auburn
40. Mississippi St
60. Arkansas

The notables –

45. West Virginia
56. Oklahoma
59. Southern Cal
72. Ohio State

Once again the SEC rates pretty well.

(It is absurd that such untested teams are playing in BCS bowls, but that is a topic for another time.)

And finally –

Sagarin Ratings SOS

8. South Carolina
9. Florida
14. Tennessee
16. Vanderbilt
21. LSU
22. Mississippi
26. Kentucky
30. Georgia
38. Alabama
47. Mississippi State
48. Auburn
64. Arkansas

And the notables –

32. USC
42. West Virginia
59. Oklahoma
60. Ohio State

We’ve mentioned it before, but Sagarin – who won’t publicly reveal his ratings methods – seems to think much more highly of the Pac Ten than any other computer rating system. He puts USC with the 32nd best schedule, significantly ahead of any other rating system. Or, compare his SOS ratings for the Cal Bears –

Sagarin – 4th
Coffey – 18th
Massey – 19th
Colley – 24th
NCAA – 55th

Whatever Sagarin is doing in his mystery numbers has Pac Ten teams consistently rated with higher SOS’s than any other system. I think it is unacceptable the BCS allows a rating system to factor into its rankings that is not transparent but again, this is a topic for another time. (Why do I mention Sagarin at all then? Because if I didn’t, some commenter would. )

The salient point is thus – stated again –

You don’t need to play “tough” out-of-conference schedules when your conference schedule is demonstrably tougher than anyone else’s.

If the SEC played only SEC opponents during the regular season, they would still have a more challenging schedule than anyone else.Or, does anyone really believe Ohio State would be a one-loss team had they played even the softest of the SEC team’s schedules?

I’d like to believe this would put the SEC “out-of-conference” weakness nonsense to bed, but it won’t. Nonsense usually has legs.

As for the obvious garbage that is the BCS system that is likewise revealed by the above – that is a topic for another time.


Anonymous said...

If there was ever an argument for having a playoff, it should be to keep teams like ohio state from scheduling cupcakes and get rewarded when they only lose one game.

Mergz said...

Indeed tron. We have a situation where OSU will either be rewarded (by beating LSU) or, perhaps even worse, another situation like last year where they will be taken to the woodshed. Either way the system is a farce.

Also, in looking at Florida's SOS across the spectrum, how can Tebow not win the Heisman with the numbers he put up against the best teams in the nation?

Anonymous said...

Objectively speaking...
Wouldn't scheduling relatively easy OOC games actually help the conference strength of schedule? By inflating the records of the teams in the conference, since you play mostly in conference teams, this should increase your strength of schedule...

Perhaps just nonconference SoS would be more relevant?

Anonymous said...

Urban ain't taken these guys "out-of-conference", forget west coast road trips, he says "I try to have a curfew when we go on the road..but I'm not going to be out chasing them around." Meyer's was trying to explain about two more Gators arrested Wednesday night for verbally abusing and throwing sandwiches and cups at restaurant workers. Can't blame him for not wanting to take the players anywhere OOC but to maybe Thug U Miami. Feel sorry for Meyers he says "complete ignorance...I have to deal with it a lot."

Anonymous said...

Mergz, doesn't this post/rant contradict your power rankings just a couple of posts below on your blogsite?

Mergz said...

Anon - Well, maybe my Power Ratings suck ;)

To answer your question - not really. My power ratings look total offense and total defense modified by SOS (the NCAA's version) and overall record. In OSU's case, they have the following -

- 397.1 ypg offense, which puts them 58th in the country, and

-225.25 ypg defense, which puts them 1st in the country, modified by

-an NCAA SOS that puts them 48th.

So they come out pretty well with that number 1 D, and when the Offense is added to the Defense and the Overall Record (11-1), they come out on top.

The real question is - do I need to change my Power Rankings? While I modify total O and total D by SOS, maybe I need to do the same for W-L record? Am I overweighting the W-L record as it is?

This is the first year of the Power Ratings, and they are subject to change. What I want to see is how teams from different conferences compete in the Bowl Games and see how it compares to my rankings.

And I don't feel at all bad about my Power Ratings not be a final product. The BCS seems to change its formula every year.

J Tin said...

@ Tron/Mergz, re: "Playoff should be instituted to keep OSU from scheduling cupcakes"

--So if I'm getting this right, from Tron and Mergz, the point is: because the SEC is so badass, other conferences better schedule tough OOC games, while the SEC can play weak teams? That's nonsense.

Take a look at my blog...the SEC's OOC schedule is not all that impressive. They dominate non-BCS opponents, but only went .500 in inter-conference matchups with other BCS conference teams.

Anonymous said...

Of course, you can also point out that the argument that OOC strength of schedule in the SEC is pretty damn good. Just look at some of the teams the SEC played OOC this year.

Miss St--WVU
UGA--Oklahoma State and Tech
Auburn--Kansas State and USF

How can anyone legitimately argue that every one of those teams outside of Arkansas put what looked at the beginning of the season as at least one toss up game?

As a point of reference, through about 9 weeks of the season, Central Michigan was the best team that anyone in the Big10 other than Illinois or Michigan had played OOC. Of course, most of those teams rely on ND to fill their decent OOC slot--but aren't they really just another team that hasn't won a bowl game since 1995?

Every year the SEC gets slammed for not playing anyone OOC because writers are either too ignorant or lazy to realize that SEC teams tend to play big OOC games either at the middle or end of the season. See, UGA-Tech, UF-FSU, and USC-Clemson

Anonymous said...

J tin,
That's not what I'm saying. OSU and Michigan were supposedly the two best teams in the country last year but were exposed in their bowl games. If michigan made it to the title game last year, they would have never been exposed as the paper tigers they were. But a playoff system would have rectified that problem.
It has nothing to do with which conference is the best. What I meant was that teams shouldn't be rewarded for weak scheduling. But as we saw last year, it's very hard to get a true picture of a team's strength from the regular season.

Unknown said...

I already have college basketball odds on the gators winning the Championship this year... GO GATORS!!!!!

Anonymous said...


Actually, I think the Saurian Sagacity Power Ratings are pretty good.

I do have questions regarding the various SoS calculators.

If the SoS calculators are accurate, then one would posit that where teams with identical or close won/loss records play, the one with a statistically significant difference in SoS ought to win.

It would seem that the genesis for this type of rating would need to emanate from interconference games played during the season and the bowl season, since that is where the bulk of the interconference games between reasonably strong opponents, at least in theory, takes place.

So, aren't the most accurate SoS calculations only available after all football has been played?

The end sum of this being, shouldn't the interconference performances be the ultimate test of the accuracy of the SoS stat?

Henry Louis Gomez said...

The problem is that the current system discourages teams in the historical power conferences from scheduling meaningful interconference games. Michigan turned Hawaii down and instead scheduled Appy State. Looking back on it, a loss to Hawaii would have been much more respectable and would have given Hawaii a bigger beef. But Michigan wanted a patsy because if you go undefeated as a Big 10 team (or PAC 10 or SEC or Big 12 or ACC) you are going to have a much better chance at the mythical national title. In short, the part of the schedule you control should be as weak as possible. For the mid majors, it's the opposite. They want to schedule good teams to elevate themselves but finding opponents is tough as referenced above.

Anonymous said...

My opinion as to why the SEC is always rated so highly by computers:
They almost never lose to scrubs.

Look at this year. Untill Alabama dropped the ball against Monroe, the SEC hadn't lost a game outside of a power conference. Compare to the big te-leven, where they lost a full 5 games. (Plus another to a 1 win duke team).
The Pac10 lost 7 times, including twice to Notre Dame.
This means, entering conference play, everyone is already beafed up to 3 or 4 wins. And the losses are always "respectable".

Just a thought.

Anonymous said...

Just an observation about using Notre Dame as a metric for conference strength...
Against ACC :: 1-2
Against Big-10 :: 0-4
Against Academies :: 0-2
Against PAC-10 :: 2-1

If they had scheduled more PAC-10 teams maybe they would have had a winning record!