A take on Tom's suggested ranking algorithm

Taking Tom’s suggestion, I thought it would be interesting to take the four conferences (Pac 12, SEC, Big 10, Big 12) and ND.

Methodology I:
A team’s W/L makes up 40% of the score
A team’s opponents W/L makes up 40% of the score (if an opponent is an FCS team, I subtracted 6 wins from that team’s W/L since a 10-1 FCS is LIKELY similar to a 4-7 FBS team in difficulty)
A team’s opponent’s opponents W/L makes up the remaining 20% (I did not include FCS opponent’s opponents in this calculation to reduce the amount of data)
Then, I adjusted the total by 0.05 for each FCS game scheduled and ranked the 49 teams.

1 Notre Dame (0.73)
2 Ohio St. (0.7)
3 Stanford (0.68)
4 Florida (0.65)
5 Texas (0.64)
6 UCLA (0.63)
7 Michigan (0.63)
8 Kansas St. (0.62)
9 Oregon (0.62)
10 Alabama (0.62)
11 Georgia (0.61)
12 S. Carolina (0.6)
13 Nebraska (0.6)
14 LSU (0.6)
15 Oregon St. (0.59)
16 Oklahoma (0.59)
17 USC (0.59)
18 Penn St. (0.58)
19 Oklahoma St. (0.56)
20 Texas A&M (0.55)
21 Arizona (0.54)
22 Michigan St. (0.54)
23 Mississippi St. (0.54)
24 Northwestern (0.53)
25 Washington (0.52)
26 Wisconsin (0.52)
27 Texas Tech (0.51)
28 TCU (0.51)
29 Vanderbilt (0.5)
30 Iowa St. (0.5)
31 Missouri (0.48)
32 Baylor (0.48)
33 West Virginia (0.47)
34 Mississippi (0.47)
35 Minnesota (0.46)
36 ASU (0.46)
37 Purdue (0.45)
38 Arkansas (0.44)
39 Tennessee (0.42)
40 Iowa (0.42)
41 Auburn (0.41)
42 Indiana (0.4)
43 Utah (0.4)
44 Cal (0.4)
45 Kentucky (0.38)
46 Illinois (0.35)
47 Kansas (0.34)
48 WSU (0.33)
49 Colorado (0.31)

Doing the same for each of the conferences (after normalizing the FCS opponents per team and FCS impact to 0.25):

1 Big 10 (0.381)
2 Pac 12 (0.355)
3 Big 12 (0.340)
4 SEC (0.305)

 

Methodology II:
The same as I above, except making all FCS teams worth 0-11.

The above order stays the same except the percentages and:
11 Nebraska
12 Georgia
13 S. Carolina

20 Arizona
21 Texas A&M

35 ASU
36 Minnesota
37 Arkansas
38 Purdue

No change in the order of the conferences

 

Methodology III:
The same as I above, except giving full credit FCS W/L and no penalty for FCS opponents.

1 Notre Dame (0.74)
2 Florida (0.73)
3 Ohio St. (0.71)
4 Kansas St. (0.69)
5 Texas A&M (0.69)
6 Oregon (0.69)
7 Georgia (0.68)
8 Stanford (0.68)
9 S. Carolina (0.68)
10 LSU (0.68)
11 Alabama (0.68)
12 Oklahoma (0.66)
13 Nebraska (0.66)
14 Oregon St. (0.65)
15 Texas (0.65)
16 UCLA (0.64)
17 Michigan (0.63)
18 Oklahoma St. (0.62)
19 Arizona (0.61)
20 Mississippi St. (0.61)
21 USC (0.59)
22 Northwestern (0.59)
23 Wisconsin (0.59)
24 Washington (0.58)
25 Penn St. (0.58)
26 Texas Tech (0.58)
27 Vanderbilt (0.57)
28 TCU (0.57)
29 Iowa St. (0.56)
30 Baylor (0.55)
31 Missouri (0.55)
32 West Virginia (0.54)
33 Mississippi (0.54)
34 Michigan St. (0.54)
35 Minnesota (0.53)
36 ASU (0.53)
37 Purdue (0.52)
38 Arkansas (0.52)
39 Iowa (0.49)
40 Auburn (0.48)
41 Tennessee (0.48)
42 Indiana (0.47)
43 Utah (0.47)
44 Cal (0.47)
45 Kentucky (0.45)
46 Illinois (0.42)
47 Kansas (0.41)
48 WSU (0.41)
49 Colorado (0.38)


As expected, this also reorders the "quality" of the conferences:
1 SEC (0.596)
2 Big 12 (0.582)
3 Big 10 (0.561)
4 Pac 12 (0.558)

 

Based on the above comparisons, the comparitively high number of FCS scheduling in the SEC and Big 12 has a direct impact on the success of those conferences as compared to the Pac 12 and Big 10. Although discussed in my previous Pac 12 subject, this yet again demonstrates that the perceived strength of the SEC (and Big 12 to a lesser degree) are a direct consequence of the number of lesser opponents those conferences schedule. When the SEC as a whole is stronger, the lesser OOC scheduling doesn't hurt. However, in years where the SEC is down, such as this year where there are fewer tough teams, the lesser opponents still bloat the overall W/L to give the impression that the conference is still just as strong when it is quite apparent if you look at the data that it is not.

 

13's my lucky number! And that's a higher ranking than the AP gives us!!

 

Close to what I expected with the first methodology. If it is easy in your calculations to make a slight change to the percentages, I would like to see the following:

50% + subtract .01 from final percentage for every FCS team scheduled
35%
15%

Reason: I'm still seeing Tennessee (39), Auburn (41), and Kentucky (45). This is too high for the records of those schools.

 

Easy enough to do. Although, modifying an algorithm to find an expected result doesn't make for good scientific process.

I'll email you the results rather than gobbing up the forum.

 

I understand. However, given the 1st and 2nd level having the same % did not follow what I had in mind. Diminishing % for lower levels seems more appropriate.

 

No worries, Tom. My thoughts with making the first two levels equal is that it's just as important to know against whom you competed as it is to know those results. An 8-3 Sun Belt team is NOT the same as an 8-3 Big 10 team.

I'm not sure I have your correct email or phone numbers. Can you message me with them?

 

I got them and did not see much difference at all.

Fornyone else, my preferred e-mail is correct if you just click on the e-mail button at the bottom of any of my posts.

 

This is great stuff, Kes.....and Tom for your suggested formula.

 

I like it... but I'd bet there are probably NOT 19 teams that would feel good about taking on Texas A&M right now.

(Our once in a lifetime occurrence of two FCS teams in a year is skewing us a bit low me thinks...)

 

oh gob it all up. I'm totally enjoying this.

 

Scott, I realize that your present schedule doe not reflect your normal schedule. I am all for ending the FCS scheduling and a system like this <del>could</del> WOULD promote just that!

 

I know Tom, I'm just poking Kesley.

I worked with him for too long to question his numbers.
He's an engineer... to the BONE.


As his final lists shows, if A&M had played soggy FBS teams instead of FCS we'd be right up there.

 

The only thing that gives me pause about those numbers is that WSU isn't last. :evil:

Scott, once you get into the spirit of SEC scheduling, you'll probably have THREE tough FCS teams every year! Only THEN will it be a true murderer's row of games! :twisted:

 

As am I... an Electronics Design Engineer, but an Engineer none the less. 8)

 

:wink: guess Dave wont be commenting on this one... Good job Kes and Tom...

 

Florida fares well in either methodology above.

If Texas A&M played Texas every year in a home and home then I'll bet they have the same budget constraints as the rest of the SEC and wouldn't be adding a home and home with a Michigan or ND, etc.

Now since they don't play an annual OOC game vs. Texas who knows what they might do scheduling wise.

Except for the boredom of some of the "filler type" games......like JSU last week.....I agree with UF's schedule as long as it helps the Gators rather than hurts them in the rankings and BCS playoff selection criteria.

UF may not win this Sat. vs. FSU but at least I think they are rested, healthy and in position to give it their best shot.