A possible way to determine a "true" champion
by Jeremy Sharp on Dec 1, 2011 12:59 PM CST
Email
Print
or a participant in the BCS title game in the event of a tie.
This method tries to give more credit for the "good" wins, while at the same time penalizing a team for a "bad" loss.
The method is described in full detail after the jump.
This method gives positive points for a win and negative points for a loss. It attempts to factor in the quality of opponent (the better the opponent in a win gets more positive points, and the worse an opponent in a loss gets more negative points). You could even have margin of loss factored in (I think margin of loss (personal opinion) means more than margin of victory) if you want to.
Basically an overly simplistic way would be to take a ranking system (Sagarin for example) sorted in the usual fashion from from best to worst (regular sagarin) for the losses and the same list sorted worst to best (reverse sagarin) for the wins. Add up the rank of all the teams for wins and subtract the rank of all the teams for losses.
Here are some teams of interest that I have already done this for (Bama, OkieSt, KSU, and OU):
Bama wins (reverse sagarin):
Kent State +122, Penn St +222, N Texas +118, Ark +240, Florida +214, Vandy +208, Miss + 139, Tenn + 196, Miss St +216, Ga Sou + 174, Auburn +213
Bama loss (regular sagarin):
LSU -1
Bama’s "Jeremy’s end of regular season" ranking: 2061
OSU wins: (assuming a win over OU)
UL Laf +160, Ariz +184, Tulsa +220, aTm +235, KU +159, UT +233, Mizzou +229, BU +231, KSU +236, TTU +203, OU +243
OSU loss:
ISU -29
OSU’s "Jeremy’s end of regular season" ranking: 2304 (interestingly, OSU’s score NOT including OU at all is 2061, so losing to OU would drop the score to 2057)
KSU wins: (assuming a win over ISU)
EKU +92, Kent St +122, Miami +200, BU +231, MU +229, TTU +203, KU + 159, aTm +235, UT +233, ISU +218
KSU losses:
OU -4, OSU -3
KSU’s "Jeremy’s end of regular season" ranking: 1915
OU wins: (based on the above assumption that OSU beats OU)
Tulsa +220, FSU +221, MU +229, Ball St +145, UT +233, KU +159, KSU +236, aTm +235, ISU +218, OSU +244
OU losses:
TTU -44, BU -16, OSU -3
OU’s "Jeremy’s end of regular season" ranking: 1833
I developed this the other night, and I haven't made any changes to the mathematics of how this works, but I do now see one small problem with this method. It really only works to compare teams that have played an equal number of games. For instance, if UGA beats LSU in the SEX title game, then using this method to determine how LSU stands up to Bama and OSU is not really fair, because LSU would have one more game with which positive points were gained to help their cause (similarly how winning 20 games in a basketball season in 2011 is "easier" than winning 20 games in a season in 1960 because more total games are played these days).
That being said, if you are trying to determine whether Bama or OSU is the more deserving 1 loss team to play in the BCS title game, or who is the "true champion" in a 3-way tie for the Big 12 regular season among 7-2 OU, 7-2 OSU, and 7-2 KSU, I think this method does a pretty good job.
Maybe since this is a fanpost it will attract more comments than my original comment did in the 11-29 slate.
I know this isn't perfect (or somebody else would have already implemented something like this), and if this was done for all 246 FBS and FCS teams, there might be a glaring reason why this is NOT used. It is a rather slow process for me to do for a single team (takes me about 5 minutes per team to look up everything and do the math), or I would have done the remaining teams in the Big 12 to get a slightly larger sample.
Update Dec 2, 2011: As BlackCats pointed out in another thread where I posted something about this, if this method is going to be used to determine the "true" champion of a conference that has no championship game (Big 12), then you would not add in the points earned (or subtracted) from the non-conference games. You would only perform this exercise for the conference games. You would also only include the conference games to determine the team that should represent a division in the championship game (for example to settle the 3 way tie in the Big 12 South that occurred last season).
I welcome all comments and criticisms.
All comments, FanPosts and FanShots reflect only the view of the user creating them.
33 comments
|
0 recs |
Do you like this story?
Comments
I originally thought of this after reading
this article. To me, after looking at this guys blind stats chart, team B was (in my mind) clearly the better option.
There were some comments from others who thought this was still perhaps a flawed exercise, so I was wanting a numerical way to help substantiate what the blind chart was showing.
I came up with the thought process and applied it to Bama and OSU (because through most of the conversations with my friends, those are the two teams still being the most heavily disputed), HOPING that OSU would come out on top, but not knowing for sure. I was EXTREMELY pleased when I saw the final results. It is not really even “close” when you use this method. This method gives OSU so much credit for the “better” opponents they player compared to Bama, that a 2 loss OSU team is only 4 points behind a 1 loss Bama.
First to identify a TEWWT (so what if it was unintentional)
I don't think the championship game is that much of a problem.
Because, just as it gives you extra bonus points if you win, it also give you extra negative points if you lose. This might actually help prevent 2001 and 2003 (and maybe 2011) scenarios from happening agaiin.
However, one way to work around the different sample sizes would be to use averages instead of totals, so:
OSU: 192
Alabama: 171.75
OU: 152.75
Hypothetical LSU score (because I’m not going to put as much effort into this as you have): 210.15
Would you like some Freys with that?
Yes, averages would be a way to work around the different numbers of games a team has played.
Obviously had not thought of that.
All of the scenarios I have played with so far I have had to assume that a particular team won or lost (depending on the scenario I am wanting to evaluate). Using averages would allow one to compare teams as they sit right now without having to assume an outcome or wait for the results of the game. There have been times where one team may have played 2 additional games over another team because of bye weeks and such.
I like it!
First to identify a TEWWT (so what if it was unintentional)
I like it
Plus it makes us the clear B12 Conference Champion if OU and we win Sat, as demonstrated HERE.
When life hands you lemons, make grape juice. Let them wonder how the F*ck you did it.
Fight till Hell freezes over and then fight them on the ice.
Just out of curiosity
Where would Boise or even Houston rank in this system? My instinct is that they would be a LOT lower and unfairly punished for playing the weak schedule they can’t necessarily control. I’m not saying that Houston deserves to be considered among the undefeateds and one loss teams because they really haven’t played anyone, but it just seems like this system would knock them completely out of the BCS and heavily favor the big conference teams.
You could maybe help control for that with margin of outcome.
The more points you win by, the more points you get.
Also, the more points you lose by, the more points you lose.
I don’t know exactly how to quantify that, but it could probably work.
I would say also that a 7 point or less loss or win wouldn’t add or subtract points from your total.
Would you like some Freys with that?
by ChrisP Wildcat on Dec 1, 2011 4:14 PM CST up reply actions
I also would like to think of a way to incorporate margin of loss/victory.
I’m like you Chris, I don’t know how to quantify that data. I remember reading somewhere that Sagarin has two different ratings – one that DOES incorporate a margin of victory component, and one that does not (this is what is used in the BCS computer rankings) – but I don’t know how he arrives at his numbers.
I would be interested to see what Bill C from RMN has to say about the method I have devised.
First to identify a TEWWT (so what if it was unintentional)
I hate margin of victory...why?
I hate running up the score on a team you clearly have the victory in your grasp. Also this method lends for more opportunity to have your starters injured in trying to rack up style points.
The difference between a successful person and others is not a lack of strength, not a lack of knowledge, but rather in a lack of will. - Vince Lombardi
I do agree in principle.
But I can’t see any other way to help Boise from being unfairly penalized for a soft conference schedule. They try to make up for it in the non-con, but they aren’t a big enough name to entice other big-names and they are too dangerous to schedule as a ‘cupcake’.
If you can think of another way to protect Boise, have at it.
Also, as I suggested capping the lower margins at 7 points (meaning if you won or loss by 7 points or less then it doesn’t affect the points you earn either way) you could also set a top-end margin.
Say, if you win by more than 21 points it all counts the same as we would count a 21 point victory.
Would you like some Freys with that?
by ChrisP Wildcat on Dec 1, 2011 11:02 PM CST up reply actions
And really, I say Boise in this, but this is about providing a fair chance for all non-major conference teams.
We don’t want to protect them too much and give them an easy path, for sure, but we need to keep them from being to heavily penalized if they actually try to beef up their schedule in the non-con. It isn’t their fault that they aren’t in the major-conferences.
Would you like some Freys with that?
by ChrisP Wildcat on Dec 1, 2011 11:04 PM CST up reply actions
after what we've seen the last two years
It isn’t their fault that they aren’t in the major-conferences.
Isn’t necessarily true anymore. Some of the teams (TCU, Boise, etc) were at least trying to do something about that by switching conferences.
First to identify a TEWWT (so what if it was unintentional)
by Jeremy Sharp on Dec 2, 2011 12:14 AM CST up reply actions
format fail
It isn’t their fault that they aren’t in the major-conferences.
is what I meant.
First to identify a TEWWT (so what if it was unintentional)
by Jeremy Sharp on Dec 2, 2011 12:14 AM CST up reply actions
And how did that work out for them?
Boise still isn’t in a major conference.
TCU is, but only because the Big 12 was getting desperate.
So, it isn’t their fault they aren’t in a major conference. They are obviously putting the effort and resources necessary into trying to be a big time program, but still they haven’t received an invitation.
I’m not talking about Wyoming here, I’m talking about the programs that have tried and failed. Houston and SMU would fall into this category too, probably.
Would you like some Freys with that?
by ChrisP Wildcat on Dec 2, 2011 7:38 AM CST up reply actions
When you insist on playing on blue turf you probably are not going to be taken seriously by a new conference.
The difference between a successful person and others is not a lack of strength, not a lack of knowledge, but rather in a lack of will. - Vince Lombardi
I seriously don't have a clue why that is such a big deal to people.
I have no problem with it and kind of like it, but, whatever.
Would you like some Freys with that?
by ChrisP Wildcat on Dec 2, 2011 10:33 AM CST up reply actions
First and foremost...
from a functional standpoint when they wear blue uniforms you can’t fricken see the players clearly due to the field. Second, no serious FBS program would ever do this. If they want to paint the endzones blue with orange lettering than I could get behind it. It is a big joke around all of college football. Do you really want people to discuss your “Smurf Turf” or the fact that you just beat Georgia? Don’t give others any reason to dilute what you are doing on the field by trying these kind of stunts.
The difference between a successful person and others is not a lack of strength, not a lack of knowledge, but rather in a lack of will. - Vince Lombardi
You call it a stunt, but like I said I think it is cool.
Obviously I am in the minority on this one, but the prevailing argument seems to be that it is not acceptable, because it is a “stunt” to get attention. The reasoning behind this conclusion is because the traditional “big-boy” programs wouldn’t do that, so therefore it must be some childish attempt at garnering attention.
I reject that line of thinking. There is no rule that says your turf has to be green. I think it is great myself, and wish more people would do it. Who cares what others think? Having green turf won’t make them a better football team, just like having blue turf doesn’t make them a better football team. It shouldn’t be a criterion in judging whether they are a serious football program, since it has absolutely ZERO impact on the quality of your football program. That would be like saying teams that don’t play in a horseshoe style stadium aren’t taking the game seriously.
As far as having a hard time making out the players, that is a legitimate complaint, but that doesn’t bother me too much. That is just personal preference though, so I don’t have a problem with that viewpoint.
Would you like some Freys with that?
by ChrisP Wildcat on Dec 2, 2011 1:12 PM CST up reply actions
Fair enough.
Would you like some Freys with that?
by ChrisP Wildcat on Dec 2, 2011 1:58 PM CST up reply actions
Is it a stunt?
If they were the original and have not gotten rid of it? I think the copy cats in FCS football are more like the ones pulling the stunt.
Bobby Hill: What's a meat examination team?
Hank Hill: It's like a debate team, only instead of doing something useless you get to grade the cut and quality of meat!
and after looking at the numbers I posted below
Boise is actually better off than you might think.
Houston is the team that needs help. They don’t have any losses (therefore are not getting ANY negative points), yet a 2 loss OU already has a better overall ranking with one game yet to play! I mean, WOW!
First to identify a TEWWT (so what if it was unintentional)
by Jeremy Sharp on Dec 2, 2011 12:17 AM CST up reply actions
Should have kept reading before I replied above.
And really, with the schedule that Houston has played this year (a relatively easy non-con) they don’t deserve any special treatment.
Would you like some Freys with that?
by ChrisP Wildcat on Dec 2, 2011 7:40 AM CST up reply actions
Right.
I was actually surprised at how Boise’s final ranking turned out. Houston’s is just pathetic though. And I don’t think any amount of “margin of victory slanting” would help them.
I was actually a bit surprised at Stanford’s number as well (and VT actually). There has been a decent amount of talk on here about how weak the PAC (and ACC) is this year, but Stanford is legitimately in the conversation with Bama and OSU when talking about whop deserves to be in the MNC game with LSU (at least right now anyway). Remember, Bama and OSU are tied with a score of 2061 right now, but that is with Bama’s full 12 game schedule complete and OSU still having to play OU. If OSU beats OU, then there should be no question, as they will jump up to 2304.
The other thing is that while this exercise can be done throughout the season (maybe starting the same time as the BCS rankings), they can change a decent amount as teams are juggled around in the sagarin rankings. But since the bowls make their decision Saturday night/Sunday, I don’t know if the new sagarin rankings would be out yet, so using the rankings I have seem reasonable to me.
First to identify a TEWWT (so what if it was unintentional)
Here are some more teams
Boise:
UGA +230, Toledo +197, Tulsa +220, Nevada +180, Fresno St +147, CSU +110, AF +177, UNLV +100, TCU -21, SDSU +194, Wyoming +178, New Mex (not played yet) = 1712 or 155.6 avg
Stanford:
SJSU +152, Duke +138, Ariz +184, UCLA +195, Colo +151, Wazzou +168, Wash +207, USC +238, Ore St +166, Ore -5, Cal +217, ND +223 = 2034 or 169.5 avg
VT:
App St +140, ECU +150, Ark St +193, Marsh +161, Clem -28, Miami Fl +200, Wake +176, BC +157, Duke +138, GT +204, UNC +199, UVA +190 , Clem (not played yet) = 1880 or 156.7 avg
Houston:
UCLA +195, N Tex +118, LA Tech +201, GA St +40, UTEP +149, ECU +150, Marsh +161, Rice +153, UAB +93, Tulane +74, SMU +179, Tulsa +220, So Miss (not played yet) = 1733 or 144.4 avg
Bama:
Kent State +122, Penn St +222, N Texas +118, Ark +240, Florida +214, Vandy +208, Miss +139, Tenn +196, LSU -1, Miss St +216, Ga Sou + 174, Auburn +213 = 2061 or 171.8 avg
OU:
Tulsa +220, FSU +221, MU +229, Ball St +145, UT +233, KU +159, TTU -44, KSU +236, aTm +235, BU -16, ISU +218, OSU (not played yet) = 1836 or 166.9 avg
OSU:
UL Laf +160, Ariz +184, Tulsa +220, aTm +235, KU +159, UT +233, Mizzou +229, BU +231, KSU +236, TTU +203, ISU -29, OU (not played yet) = 2061 or 187.4 avg
KSU:
EKU +92, Kent St +122, Miami Fl +200, BU +231, MU +229, TTU +203, KU + 159, OU -4, OSU -3, aTm +235, UT +233, ISU (not played yet) = 1697 or 154.3 avg
This continues to prove the point of most on BOTC that our body of work is more impressive than Boise, Houston, or VT…The reason I say this is because our numbers are reasonably close (average) to these others and we HAVE ONE MORE LOSS (or two in the case of Houston).
First to identify a TEWWT (so what if it was unintentional)
by Jeremy Sharp on Dec 2, 2011 12:12 AM CST up reply actions
How did we fall below OU?
Yesterday we were ahead right?
The difference between a successful person and others is not a lack of strength, not a lack of knowledge, but rather in a lack of will. - Vince Lombardi
These scores are not including the final game.
So, one less loss for OU and one less win for KSU.
Would you like some Freys with that?
by ChrisP Wildcat on Dec 2, 2011 10:34 AM CST up reply actions
Chris nailed it. The numbers you saw yesterday were for the scenario of KSU beating ISU and OSU beating OU.
That’s why I like the idea of using the average instead of raw score. It gives you a better comparison of two teams when the schedules aren’t complete.
And remember, a win against ISU is going to boost KSU’s total score more than the loss to OSU will hurt OU’s score. That’s the whole point of this. OU will be losing to the #3 sagarin team, so they will only lose 3 points. ISU is the #29 sagarin team so beating them will boost our score by 218 points.
And also, as Black pointed out, when comparing OU, OSU, and KSU with each other for conference bragging rights, you don’t include the non-con games, and that benefits KSU and hurts OU (their non-con opponents are actually pretty decent).
First to identify a TEWWT (so what if it was unintentional)
by Jeremy Sharp on Dec 2, 2011 11:08 AM CST up reply actions
Ok gents. Got it.
The difference between a successful person and others is not a lack of strength, not a lack of knowledge, but rather in a lack of will. - Vince Lombardi
will be interesting to see what happens
after this weekend’s games are completed and the new sagarin rankings come out. Shouldn’t change a lot, but in the 3 way tie for Big 12 champion, it could potentially flip flop OU and OSU for 2nd and 3rd place since they were only one point apart to begin with.
First to identify a TEWWT (so what if it was unintentional)
I just wanna know
Why is there MATH in my FOOTBAW!!!
Bobby Hill: What's a meat examination team?
Hank Hill: It's like a debate team, only instead of doing something useless you get to grade the cut and quality of meat!
Haven't you heard? Everything is about math.
When life hands you lemons, make grape juice. Let them wonder how the F*ck you did it.
Fight till Hell freezes over and then fight them on the ice.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
