Should the final four only include 0 to 1 loss teams?

[cane_man]

-
I am not sure the authors exact thoughts, but think of it this way. CFB goes to RPI like basketball. TEam A is 12-0 and has a RPI of .50. Team B 10-2 and has an RPI of .66.. For the playoffs Team A is given a score of 6 (12*.5) and Team B is given a score of 6.6 (10*.66). Team B is given the nod over Team A.. make sense?

That sounds terrible

 

[Smart]

You only seem to care about notable victories. What about Oregon's loss? It was to an 11-1 team. Should Stanford's just be forgiven because it was to a 7-5 team? They got bonus points for having a great win, but don't lose anything for having a bad loss?

If you really believe Stanford and Oregon were even (or Stanford was better), then they played similar PAC schedules. The difference is Stanford beat UCLA twice and Oregon beat Arizona State. Was UCLA good? Sure, but not much better than AZ State. They played more or less the same conference schedule, and it was a wash record-wise.

I understand wanting to give Stanford credit for their OOC, but here's what you are missing: They lost to Notre Dame. And it's tough for me to fault Oregon for their weak schedule, when it was K-State who backed out at the last minute on them AND they kicked K-State's ass in a bowl.

All things considered, if you ignore MOV, Stanford was probably slightly better. You have to balance on one hand (Two wins over UCLA vs. one over AZ State) and (Loss to Notre Dame versus no elite OOC opponents).

While Stanford comes out slightly better there, it's not enough to offset Oregon playing better every week. It's one thing to say a win by 24 is better than a win by 17. That's debatable. But Stanford had seven wins by a score or less, and Oregon never won by less than 11. It's not debatable which team played better, and that overcomes whatever de minimis advantage Stanford's schedule had.

 

[Codaxx]

Having a strong SOS doesn't mean much when one of your losses is to a sub par / bad team.

-
have never understood that. TEam A plays the exact schedule of TEam B. Both go 12-1. Does it matter where that loss came from? The path is the same.

 

[Codaxx]

You only seem to care about notable victories. What about Oregon's loss? It was to an 11-1 team. Should Stanford's just be forgiven because it was to a 7-5 team? They got bonus points for having a great win, but don't lose anything for having a bad loss?

If you really believe Stanford and Oregon were even (or Stanford was better), then they played similar PAC schedules. The difference is Stanford beat UCLA twice and Oregon beat Arizona State. Was UCLA good? Sure, but not much better than AZ State. They played more or less the same conference schedule, and it was a wash record-wise.

I understand wanting to give Stanford credit for their OOC, but here's what you are missing: They lost to Notre Dame. And it's tough for me to fault Oregon for their weak schedule, when it was K-State who backed out at the last minute on them AND they kicked K-State's ass in a bowl.

All things considered, if you ignore MOV, Stanford was probably slightly better. You have to balance on one hand (Two wins over UCLA vs. one over AZ State) and (Loss to Notre Dame versus no elite OOC opponents).

While Stanford comes out slightly better there, it's not enough to offset Oregon playing better every week. It's one thing to say a win by 24 is better than a win by 17. That's debatable. But Stanford had seven wins by a score or less, and Oregon never won by less than 11. It's not debatable which team played better, and that overcomes whatever de minimis advantage Stanford's schedule had.

MOV is bad. I would be happy to use something like Pts per play - Defenses Pts yielded per play to get you MOV per play. That would adjust for pace. Hurry Spread teams tend to have higher MOVs than grind it out teams. The best loss or best win never concerned me. It is just a data point people will flip flop on based on which one suites their need/

 

[WNY_FOOTBALL_DUDE]

You only seem to care about notable victories. What about Oregon's loss? It was to an 11-1 team. Should Stanford's just be forgiven because it was to a 7-5 team? They got bonus points for having a great win, but don't lose anything for having a bad loss?

If you really believe Stanford and Oregon were even (or Stanford was better), then they played similar PAC schedules. The difference is Stanford beat UCLA twice and Oregon beat Arizona State. Was UCLA good? Sure, but not much better than AZ State. They played more or less the same conference schedule, and it was a wash record-wise.

I understand wanting to give Stanford credit for their OOC, but here's what you are missing: They lost to Notre Dame. And it's tough for me to fault Oregon for their weak schedule, when it was K-State who backed out at the last minute on them AND they kicked K-State's ass in a bowl.

All things considered, if you ignore MOV, Stanford was probably slightly better. You have to balance on one hand (Two wins over UCLA vs. one over AZ State) and (Loss to Notre Dame versus no elite OOC opponents).

While Stanford comes out slightly better there, it's not enough to offset Oregon playing better every week. It's one thing to say a win by 24 is better than a win by 17. That's debatable. But Stanford had seven wins by a score or less, and Oregon never won by less than 11. It's not debatable which team played better, and that overcomes whatever de minimis advantage Stanford's schedule had.

Here's the difference between us: I don't care about MOV. I think it's a very deceptive and unrealistic stat. I care more about quality of schedule than I do about how well you beat an opponent. A win is a win.

UCLA beat Arizona State in the head-to-head match-up, and won 2 more games. How can you logically say there wasn't much of a difference?

Stanford gets the edge because they played a better schedule, won the conference, and beat Oregon in the head-to-head match-up. In a better system, both Oregon and Stanford would get the chance to win a National Championship.

How is losing to 7-5 Washington a bad loss, particularly when it happened early in the season and with a different QB? And it's not like Washington ripped apart Stanford, it was a close game.

 

[WNY_FOOTBALL_DUDE]

MOV is bad. I would be happy to use something like Pts per play - Defenses Pts yielded per play to get you MOV per play. That would adjust for pace. Hurry Spread teams tend to have higher MOVs than grind it out teams.

Right. It's a terrible stat, and was rightly removed from the computer formulas, and ironically it was Oregon, which lead the charge against MOV. There's just so many things wrong with it.

 

[The Authority]

-
have never understood that. TEam A plays the exact schedule of TEam B. Both go 12-1. Does it matter where that loss came from? The path is the same.

Hell yes it does.

If Team A loses to a 4 - 8 team and Team B loses to a 10 - 1 team obviously Team B gets the nod.

How else would you decide who was better?

 

[WNY_FOOTBALL_DUDE]

Hell yes it does.

If Team A loses to a 4 - 8 team and Team B loses to a 10 - 1 team obviously Team B gets the nod.

How else would you decide who was better?

Hold the phone on that one. What was the SOS numbers? Did either Team A or Team B win their conference? Did the two teams play each other?

 

[Codaxx]

Hell yes it does.

If Team A loses to a 4 - 8 team and Team B loses to a 10 - 1 team obviously Team B gets the nod.

How else would you decide who was better?

I still dont understand. That could mean the best win for Team A is 6-6 team, while Team B beat 10-1 team. It is a simple idea. 2 sprinters run 100 meters with 10 different sized hurdles. Both trip over a hurdle and run it in 13 seconds. Do we give the medal to the guy that tripped over the tallest hurdle? Is he superior, because he tripped on a 30 inch hurdle and the other guy tripped on a 28 inch hurdle? Or do you reward the runner that cleared the highest hurdle? I just dont get your pt.

 

[Codaxx]

Hold the phone on that one. What was the SOS numbers? Did either Team A or Team B win their conference? Did the two teams play each other?

-
Example was identical schedules. No head to head meeting. They played 12 games same exact way. Clearly hypothetical.

 

[Smart]

Using MOV as a number in a stats formula is stupid. Agreed. But when one team consistently crushes opponents and has their backups in the entire second half, while the other team is repeatedly in battles in the last five minutes it has to matter. Oregon and Stanford had the same PAC record. So there are two ways to separate them:

(1). Say that two wins over UCLA is so much better than one over Arizona State that Stanford gets the edge; or
(2.) Ask which team objectively played better throughout the season. This was definitely Oregon. There are some cases where MOV is close and it doesn't tell a story, but Oregon repeatedly crushed teams that Stanford struggled with. It wasn't a one-time pattern. It was more than half of their season.

I'm really struggling to see why Option #1 is a better choice than Option #2. It seems to be rewarding Stanford for something wholly out of their control. If there were some evidence that Oregon would have been greatly affected by the change, you might have something, but the evidence just doesn't exist. All signs point towards Oregon killing UCLA. What you are, in essence, saying is "Even though you were more easily and consistently more dominant all year, I am going to pick Stanford over you because they were given the chance to beat a team all signs point to you beating by 35 ad you weren't.

It seems to be punishing teams for something outside their control which is exactly what we don't want to do. Teams should be rewarded by who was the better team, and even with an OT loss at Autzen, it is abundantly clear that Oregon consistently outperformed Stanford by margins that are so wide it isn't because of style of play or pace.

 

[4down20]

I'll ask again.

If the polls are so bad at determining who should go to these games, please name the times/teams that have been screwed over in these things that the current expansion to 4 teams would not fix.

There is 2004 Auburn and 2011 Oklahoma St pretty much. Both of which the increase to 4 teams was designed to fix and would fix.

I keep saying people make these claims about the polls as if it's somehow truth, yet I have not seen anything to actually support the claim. If you want to say sometimes the polls have some #16 teams up higher than they should while another team is lower @ #19 and blah blah blah. Fine ok, but that has nothing to do with the playoffs and top teams where such rankings actually matter.

So lets see the proof that the polls don't work.

 

[Codaxx]

I still dont understand. That could mean the best win for Team A is 6-6 team, while Team B beat 10-1 team. It is a simple idea. 2 sprinters run 100 meters with 10 different sized hurdles. Both trip over a hurdle and run it in 13 seconds. Do we give the medal to the guy that tripped over the tallest hurdle? Is he superior, because he tripped on a 30 inch hurdle and the other guy tripped on a 28 inch hurdle? Or do you reward the runner that cleared the highest hurdle? I just dont get your pt.

-
I wanted to add best win actually makes more sense. In this case you know one runner is capable of clearing a 30 inch hurdle. We do not know the other runner can, because his highest cleared was 28 inches. We also know that the runner that tripped at 28 inches is capable of clearing that hurdle. because he cleared 30 inches. Given that how would you logically award the race to the runner that tripped on the highest hurdle?

 

[WNY_FOOTBALL_DUDE]

-
Example was identical schedules. No head to head meeting. They played 12 games same exact way. Clearly hypothetical.

In that case, I would pick the better loss.

 

[Smart]

And why does winning the conference matter? It's just another surrogate for head-to-head. It's counting the same factor twice.

Teams should be ranked on merit. I'm tired of head to head mattering, but common opponents being irrelevant. It takes an entire season of work and throws it out the window for one night. And that's not how statistics or common sense works. In statistics, it's very clear that more data points are better. If one team beats another by 52, that's one thing. It shows one team was conclusively better. But when a team wins in overtime and played worse the entire rest of the season? You seem to be saying "One play in this game tells me more about merit than hundreds of plays throughout the season." And it's bullshit.

But there's something to be said for close wins. Some teams (including possibly Stanford) have a knack for winning close games. If you win more games and come up big in clutch situations, there's something to that. But here's the thing: STANFORD AND OREGON HAD THE SAME RECORD IN THE PAC-12. So that argument is gone. Now all you have is two teams with the same record, and you saying "one win (AS I KEEP EMPHASIZING, IN OVERTIME) is more important than eight other games, which clearly shows one team is better. And the two have the same record."

It's the single most illogical aspect of college football to me. I just don't get how anyone could support it, and it seems like everyone does. Nobody would even consider Bama over Auburn last year, for example, even though Bama was clearly the better team the entire regular season and lost on a fucking kickoff return in Auburn. The message seems to be "let's reward the worst team because they happened to draw you at home this year, even though thy have the same record." People purposely chose the less deserving team!

 

[Smart]

I'll ask again.

If the polls are so bad at determining who should go to these games, please name the times/teams that have been screwed over in these things that the current expansion to 4 teams would not fix.

There is 2004 Auburn and 2011 Oklahoma St pretty much. Both of which the increase to 4 teams was designed to fix and would fix.

2013 Alabama. People can roast me all they want for it, but there is no objective way to reach the conclusion that Auburn played better than Bama before the bowls last year.

 

[Codaxx]

In that case, I would pick the better loss.

-
I made the case against this using runners, mostly to avoid inherent biases based on CFB seasons. I would like to hear the logic one uses to derive the better team being the one that had the best loss.

 

[4down20]

2013 Alabama. People can roast me all they want for it, but there is no objective way to reach the conclusion that Auburn played better than Bama before the bowls last year.

We had our chance and blew it.

 

[occupant]

Winning one game head-to-head (in overtime, no less) doesn't undo seven games of clearly being outplayed. And of course, one of those games was a loss to a team Oregon beat by 31. Or to reframe this:

A beat B by 3 (in overtime)
B beat C by 31
C beat A by 4

The logical conclusion is that B was the best based solely on these games. And there can be no doubt that B was better outside these games, as they literally did better against every opponent. To argue that B's much better resume (again, including a game where they won and A lost) is offset by an overtime loss is beyond ludicrous.

It all depends on what conference the refs were from.

 

[Smart]

We had our chance and blew it.

You only "blew it" because of a system that that rewards the status quo ranking system over actually picking the best team.

How can you be eliminated by "blowing it" when you lose by 6 in Auburn but Auburn didn't blow it by losing by 14 to LSU? You didn't "blow it" against LSU...you beat them by 21.

You only had one close game all regular season (@ TAMU) and guess what....you were still closer than Auburn. They had seven wins by 8 or less. You had 1.

But somehow, because the schedule makers put you in Auburn and you lost in the closest game possible, you were less deserving even though you had the same number of losses and played better. Uh, what? If you take off your "that's the way CFB is" glasses and look at it, there is no doubt who deserved to play in the NCG. And it wasn't Auburn. I just hope given a similar situation, the committee will see the same thing.