Is There Any Doubt Okla St Will Be #2?

[Eagles_Legendz]

The entire schedule matters, not just the toughest four or five games. Just taking a simple average of the ratings of the opponents should give you a decent idea. A big reason is that the Big 12 plays nine conference games.

I don't know about a lot of the computers used, but I will vouch for Sagarin's model.

Lets do it this way: @Alabama > KState, Oregon (neutral) > Baylor, Arkansas at least = Oklahoma, Georgia (neutral) > (or equal if you really want to stretch it) @Texas. Yes?

OK, lets keep going: WVU at least = A&M, Auburn > Iowa State, Mississippi St=Tulsa?, Tennessee = Texas Tech, FL > Arizona, LA-Lafayette>Kentucky (?), Kansas > both WKU, NWST.

Support the system all you want. I have no problem with Sagarin. But I challenge anyone to defend ranking that schedule ahead of LSU's by any metric.

 

[IAppreciateIt]

Lets do it this way: @Alabama > KState, Oregon (neutral) > Baylor, Arkansas at least = Oklahoma, Georgia (neutral) > (or equal if you really want to stretch it) @Texas. Yes?

OK, lets keep going: WVU at least = A&M, Auburn > Iowa State, Mississippi St=Tulsa?, Tennessee = Texas Tech, FL > Arizona, LA-Lafayette>Kentucky (?), Kansas > both WKU, NWST.

Support the system all you want. I have no problem with Sagarin. But I challenge anyone to defend ranking that schedule ahead of LSU's by any metric.

Why would we do it that way? We have a model that has been fine-tuned over years of data designed solely to determine how good teams are based on predictive value. I'd take that over your (or my) personal opinions and inequality signs.

Tulsa, for example, is a quality team (ranked #31 in Sagarin), and Oklahoma State played them on the road. But they probably don't seem like a great matchup to the average fan.

Oklahoma State played seven top 30 Sagarin opponents, plus 31. They only played three outside the top 50, and none outside the top 100. While LSU played five and three respectively.

You're trying to tell me, objectively, playing teams that finished BY THE BCS OWN CALCULATIONS: @2, 5 (neutral), 6, 16 (neutral), @23, 25 is somehow easier than: 8, 12, 14, @24? How is that even possible?

Nobody is saying that. That part of LSU's schedule was tougher than Oklahoma State's schedule. But SOS is based on the entire schedule, not just an arbitrarily selected section of the schedule.

 

[Eagles_Legendz]

Why would we do it that way? We have a model that has been fine-tuned over years of data designed solely to determine how good teams are based on predictive value. I'd take that over your (or my) personal opinions and inequality signs.

Tulsa, for example, is a quality team (ranked #31 in Sagarin), and Oklahoma State played them on the road. But they probably don't seem like a great matchup to the average fan.

Oklahoma State played seven top 30 Sagarin opponents, plus 31. They only played three outside the top 50, and none outside the top 100. While LSU played five and three respectively.



Nobody is saying that. That part of LSU's schedule was tougher than Oklahoma State's schedule. But SOS is based on the entire schedule, not just an arbitrarily selected section of the schedule.

I'm not doing it arbitrarily. I'm going down the list based on strength of teams played. Order it by OKState's schedule and work your way down. You come out with LSU winning more of those matchups in terms of strength of schedule, unless you want to dispute any of what I said. I organized it by who the BCS ranked in the top 25 and compared it down the list from there.

If you want to say Tulsa is better than Mississippi State, Kansas is better than NW St etc, fine. You can't with a straight face say that outweighs LSU playing 3 teams ranked higher than OK State's best win. OK State played one BCS top 25 team away from home. LSU? 4.

My point is, I think, objectively, if you list it by Oklahoma State's most difficult teams and compare it to LSU's and go from most difficult to easiest, LSU ends up winning the comparison like 7-5 or 8-4, and, it is more front loaded, which should weigh more in a metric compared to the difference between Kansas and WKU.

BTW, speaking of drawing arbitrary lines by including tulsa at 31. LSU played, by Sagarin's methods, 32, 33, 34, and 35, which I'm sure you already knew.

Sagarin also has Iowa State rated higher than West Virginia. I'd like to see the line on that game on a neutral field.

 

[IAppreciateIt]

I'm not doing it arbitrarily. I'm going down the list based on strength of teams played. Order it by OKState's schedule and work your way down. You come out with LSU winning more of those matchups in terms of strength of schedule, unless you want to dispute any of what I said. I organized it by who the BCS ranked in the top 25 and compared it down the list from there.

If you want to say Tulsa is better than Mississippi State, Kansas is better than NW St etc, fine. You can't with a straight face say that outweighs LSU playing 3 teams ranked higher than OK State's best win. OK State played one BCS top 25 team away from home. LSU? 4.

My point is, I think, objectively, if you list it by Oklahoma State's most difficult teams and compare it to LSU's and go from most difficult to easiest, LSU ends up winning the comparison like 7-5 or 8-4, and, it is more front loaded, which should weigh more in a metric compared to the difference between Kansas and WKU.

BTW, speaking of drawing arbitrary lines by including tulsa at 31. LSU played, by Sagarin's methods, 32, 33, 34, and 35, which I'm sure you already knew.

Sagarin also has Iowa State rated higher than West Virginia. I'd like to see the line on that game on a neutral field.

You are right about the top 30. Sagarin already has it listed on his ratings, so I used that.

The point is it's the entire schedule that matters. People always say "top to bottom" and then immediately ignore it and only look at the top, but in this case it really is top to bottom.

Iowa State is a 2.5 favorite over Rutgers. A pretty good predictor of Vegas lines is taking the average of Sagarin's overall rating and his predictor rating. Perhaps leaning a bit towards the latter.

 

[Eagles_Legendz]

I understand what you're saying but I'm honestly not attempting to discount the games after the first few. I simply do not think Iowa State beats WVU/Auburn on a neutral field more often than not. Perhaps I'm wrong. I think looking at the totality of the schedule, LSU hard a more difficult road than OSU, not only at the top, but also between teams 4-7 that they played in comparison.

Similarly, Sagarin has LSU and Alabama close in terms of SOS, which I also disagree with.

 

[IAppreciateIt]

I understand what you're saying but I'm honestly not attempting to discount the games after the first few. I simply do not think Iowa State beats WVU/Auburn on a neutral field more often than not. Perhaps I'm wrong. I think looking at the totality of the schedule, LSU hard a more difficult road than OSU, not only at the top, but also between teams 4-7 that they played in comparison.

Similarly, Sagarin has LSU and Alabama close in terms of SOS, which I also disagree with.

I'm discounting where games were played, which the MSM idiots do and is should never be done, so I will shame myself in advance, but it was easier this way -- OK State played 6 road games, LSU played 5 and 2 neutral, so we'll just call it even.

Sagarin rankings of Oklahoma State's opponents:
4
10
12
13
15
16
29
31
44
62
87
89

LSU's:
2
6
7
18
32
33
34
35
51
84
101
107
176

Even if you get rid of all the garbage teams (which I think you shouldn't -- playing a bad team like Arizona is still far tougher in terms of preparation and physical fatigue than playing a team like NW State), Oklahoma State still plays a tougher schedule if you use a simple mean. Even if you do it the "matchup" way (which I think is pretty rudimentary), Oklahoma State wins after the toughest three opponents.

The Big 12 teams are almost uniformly underrated because of their poor records -- but that's because they beat each other up, playing nine conference games with only one bad team.

Auburn is only a 1 point favorite over UVA. They stink. ISU would be at least a field goal favorite over them.

 

[Eagles_Legendz]

I'm discounting where games were played, which the MSM idiots do and is should never be done, so I will shame myself in advance, but it was easier this way -- OK State played 6 road games, LSU played 5 and 2 neutral, so we'll just call it even.

Sagarin rankings of Oklahoma State's opponents:
4
10
12
13
15
16
29
31
44
62
87
89

LSU's:
2
6
7
18
32
33
34
35
51
84
101
107
176

Even if you get rid of all the garbage teams (which I think you shouldn't -- playing a bad team like Arizona is still far tougher in terms of preparation and physical fatigue than playing a team like NW State), Oklahoma State still plays a tougher schedule if you use a simple mean. Even if you do it the "matchup" way (which I think is pretty rudimentary), Oklahoma State wins after the toughest three opponents.

The Big 12 teams are almost uniformly underrated because of their poor records -- but that's because they beat each other up, playing nine conference games with only one bad team.

This goes out the window when you do it based on BCS rankings or put a team's name next to the rankings. Like I said, using that blind method, it insinuates Iowa State is better than West Virginia, which I don't think is true. I know arguing about what 'experts' think etc is largely irrelevant, but WVU beats Iowa State 7/10 times to me. If you disagree with that, fine, but I don't think it is an unreasonable stance.

I agree with the who wins in this matchup being a rudimentary way of doing things, but not when it almost uniformly goes LSU's way. You say it doesn't based on Sagarin's rankings. Okay, then we won't agree.

The SEC had one awful team as well this year by that standard (Ole Miss). Kentucky/Tennessee are probably ~ Texas Tech.

I think if you list the teams based on Sagarin rankings, the teams LSU played win that matchup. I think LSU's schedule is more difficult objectively by any metric, including the totality of their schedule. I understand why you don't, and that Sagarin doesn't (which was my issue originally), so we'll agree to disagree.

 

[88Deac]

Spaulding: I want, I want, LSU-Oklahoma State. I want Alabama-Stanford. I want Boise State-Oregon….

Judge Smails (interrupting): YOU will get Alabama-LSU II and Michigan-Va. Tech, and LIKE it!

 

[IAppreciateIt]

This goes out the window when you do it based on BCS rankings or put a team's name next to the rankings. Like I said, using that blind method, it insinuates Iowa State is better than West Virginia, which I don't think is true. I know arguing about what 'experts' think etc is largely irrelevant, but WVU beats Iowa State 7/10 times to me. If you disagree with that, fine, but I don't think it is an unreasonable stance.

I agree with the who wins in this matchup being a rudimentary way of doing things, but not when it almost uniformly goes LSU's way. You say it doesn't based on Sagarin's rankings. Okay, then we won't agree.

The SEC had one awful team as well this year by that standard (Ole Miss). Kentucky/Tennessee are probably ~ Texas Tech.

I think if you list the teams based on Sagarin rankings, the teams LSU played win that matchup. I think LSU's schedule is more difficult objectively by any metric, including the totality of their schedule. I understand why you don't, and that Sagarin doesn't (which was my issue originally), so we'll agree to disagree.

Well there are two different arguments here. I was responding to the idea that Sagarin's SOS calculations were illogical. Obviously, it's based on his own rankings.

If you disagree with the rankings of his teams, then obviously you won't agree with his SOS rankings since they come from the same source.

Even if you say that WVU is better than Iowa State, they probably wouldn't win 7 out of 10 matchups -- that amounts to about seven points, and WVU would not be a seven point favorite over Iowa State. So I would say that's an unreasonable stance.

 

[Haros]

There are a few other problems, a number of them mandated by the BCS THEMSELVES, in order to rig the entire process in favor of major conference teams over smaller schools.

1) Beyond actually having automatic bids (which, other than a few exceptions, I'm okay with), there is a different threshold that teams must meet in order to be selected for a BCS bowl. How the heck can a conference champion from a BCS league be an automatic bid at #40 (Uconn last year, WVU is #20 this year), but a conference champ from the MWC can be #4 and potentially not get a bid if the CUSA champ is #3?

On to the computer rankings themselves:

1) They are not just some guy throwing darts at a dart board. Please. They are computer rankings, not subjective rankings. That's the whole point. Even if you don't like the methodology going into them, every team is judged by the same set of formulas within the computer, and they are not tweaked midseason. They are allowed to make changes in the offseason to refine their methods.

2) In order to measure team strength over a 12 game schedule, you need to know more than whether a team was winning or losing, you need to know how much they are winning by. There aren't enough data points for the rankings to be entirely meaningful otherwise. Of course, that hasn't stopped the BCS from disallowing the use of margin of victory in the computer rankings. What results is, even from the computer formula creators point of view, a less than ideal set of rankings. There is a reason Sagarin posts an overall set of rankings, which are his 'ideal' set, along with the two components of it. Unfortunately, that ideal set is not the one allowed to used by the BCS, and you instead get the 'ELO Chess' half of his rankings, which do not factor in the important margin of victory factor which is a huge part of determining just how good each team is. Extrapolate that explanation to the other 5 computers as well. Blame the BCS for that.

3) As IAI pointed out, almost all of the SOS rankings use a simple or weighted average of opponent strength. Almost every major conference opponent will be in the top 80. Two-thirds in the top 40. There is much less difference between them than there is between team 101 and 176. Yet, the chance of a top 5 team beating #20 on a neutral field (about 70%), #40 (85%), or #80 (95%) varies more than the winning percentage against #100 (98%) and #180 (99.9%). A simple log formula that made the bottom teams have less impact would easily solve this, but that's frowned upon. I'll let you guess who encourages this practice (with the idea that non-BCS conference teams that invariably have a few more awful teams on their schedule than a BCS conference team would).

4) 9 conference games (and just 3 non-conf.) vs. 8 and 4. It makes a difference if you schedule a second creampuff with that extra conference game to earn money on an extra home game. In the current environment, you have to account for that tradeoff.

5) The reason computers are used at all is illustrated by the back-and-forth over Auburn/Iowa St/West Virginia. The natural inclination is to belive the bigger name will always be the better team, but that will not necessarily be true in a given season. It is a fair assumption that Iowa State is better than Auburn this year, and not substantially inferior to WVU, even if that is counter to conventional wisdom due to past seasons. If you still can't get your mind quite there, see Wake Forest 2006 for an illuminative example.