(11-29-2018 12:16 PM)Frog in the Kitchen Sink Wrote:  I was playing around with the 538 numbers:

That's a neat site, but my gosh, hard to take seriously a model that says Notre Dame has a 79% chance of making the playoffs. Their actual chance is 99.9%.

Heck, the model says if all the favorites win - Alabama, Clemson, Ohio State, and Oklahoma - that ND's chances of making the playoffs fall to 64%. IMO that is nutso. They remain at 99.9%.

The only way they don't make the playoffs is if something like a giant meteor hits the USA and blows the country up, such that there are no playoffs.

(11-29-2018 12:16 PM)Frog in the Kitchen Sink Wrote:  I was playing around with the 538 numbers:

https://projects.fivethirtyeight.com/201...id=rrpromo

Basically if I'm on the committee, I want Alabama to win and one and only one of Ohio State, Oklahoma and Clemson to lose. Any of those combos results in 4 teams with >94% chance. Alabama losing and one and only 1 of those three losing isn't so bad for the committee- that results in 4 teams with >80% chance and only one team (alabama) with around a 20% chance, so really not too ambiguous.

All other combos results in probabilities of 2 or more teams are near each other for the last spot (i.e., chaos).

Yeah but it's numbers aren't necessarily gospel. Sorry but a scenario where
Georgia beats Alabama
Clemson wins
Oklahoma wins
Ohio St loses
isn't Clemson and Georgia >99%, Notre Dame 91%, and Oklahoma 84-Alabama 22%. Alabama WAY more likely to go to the playoff in that scenario than Oklahoma......

And it's numbers if everyone wins......
Oklahoma 68%
Ohio St 66%
Notre Dame 65%

sorry but don't buy that at all..

basically, they WAY overvalue the impact of the conference championship. It's no where near as absolute as they are painting it...

what is interesting is if we see Ohio St and Oklahoma lose with UCF winning...

they give it
35% Georgia
31% UCF
16% Texas
11% Washington!

(11-29-2018 12:24 PM)stever20 Wrote:  

(11-29-2018 12:16 PM)Frog in the Kitchen Sink Wrote:  I was playing around with the 538 numbers:

https://projects.fivethirtyeight.com/201...id=rrpromo

Basically if I'm on the committee, I want Alabama to win and one and only one of Ohio State, Oklahoma and Clemson to lose. Any of those combos results in 4 teams with >94% chance. Alabama losing and one and only 1 of those three losing isn't so bad for the committee- that results in 4 teams with >80% chance and only one team (alabama) with around a 20% chance, so really not too ambiguous.

All other combos results in probabilities of 2 or more teams are near each other for the last spot (i.e., chaos).

Yeah but it's numbers aren't necessarily gospel. Sorry but a scenario where
Georgia beats Alabama
Clemson wins
Oklahoma wins
Ohio St loses
isn't Clemson and Georgia >99%, Notre Dame 91%, and Oklahoma 84-Alabama 22%. Alabama WAY more likely to go to the playoff in that scenario than Oklahoma......

And it's numbers if everyone wins......
Oklahoma 68%
Ohio St 66%
Notre Dame 65%

sorry but don't buy that at all..

basically, they WAY overvalue the impact of the conference championship. It's no where near as absolute as they are painting it...

what is interesting is if we see Ohio St and Oklahoma lose with UCF winning...

they give it
35% Georgia
31% UCF
16% Texas
11% Washington!

I think they'd be the first to admit it isn't absolute, btw. The percentages are simply recognizing the level degree of uncertainty based on lots of variables. In fact, I think aren't you and quo are the ones arguing in absolute terms of what you think will happen.

But I agree they might be overestimating the value of conference championships in their formula.

No, I feel like the #1 team in the country and losing to the #4 team in the country is far more likely to still make the playoffs than some 22%. The ratings are for the WHOLE season and not just the last game.....

(11-29-2018 12:37 PM)stever20 Wrote:  No, I feel like the #1 team in the country and losing to the #4 team in the country is far more likely to still make the playoffs than some 22%. The ratings are for the WHOLE season and not just the last game.....

The words you use still imply some level of uncertainty so that’s good. I personally do not agree with “way” more likely, although if you said it was a 2:1 chance Alabama would make it in that scenario I wouldn’t disagree too much. Way more likely could be 80-20 and I’d disagree there. If I’m being honest I really really don’t know what they will do, so I’m more 50-50. I do agree that they seem to be underestimating Alabama’s chances with a loss.

(11-27-2018 05:07 PM)Wedge Wrote:  

(11-27-2018 04:44 PM)Gamecock Wrote:  If Texas, Northwestern, and Alabama won it would be extremely interesting to see who gets that 4th spot.

The committee would give it to Georgia, unless they are blown out by the Tide. (Maybe "swept away by the Tide" would be a more appropriate metaphor. Ha.)

This.

The only way UCF makes it is if Texas wins, Northwestern wins, Alabama wins by 40, and UCF wins by 40.

Even then, Michigan would get a serious look from the committee (which IMO is a mistake). It would help UCF's cause if Pitt beat Clemson, but only because it will make UCF's win over Pitt look better.

One thing the 538 model doesn't seem to account for is MOV, which will surely matter.

E.g., if Alabama loses to Georgia, whether they still make the playoffs will depend greatly on how they lose. If they lose in seven OTs like LSU did to TAMU last week, they will almost surely still make the playoffs. If they get beat by 30 like Michigan did to Ohio State, they are far more likely to drop out.

Likewise, if it boils down to Ohio State vs Oklahoma for the last spot and both win, how each team wins will have a big impact. If Oklahoma barely scrapes by while Ohio State wins 50-0, Ohio State has a much better chance of getting the nod than if they both win by the same amount, etc.

(11-29-2018 01:02 PM)quo vadis Wrote:  One thing the 538 model doesn't seem to account for is MOV, which will surely matter.

E.g., if Alabama loses to Georgia, whether they still make the playoffs will depend greatly on how they lose. If they lose in seven OTs like LSU did to TAMU last week, they will almost surely still make the playoffs. If they get beat by 30 like Michigan did to Ohio State, they are far more likely to drop out.

Likewise, if it boils down to Ohio State vs Oklahoma for the last spot and both win, how each team wins will have a big impact. If Oklahoma barely scrapes by while Ohio State wins 50-0, Ohio State has a much better chance of getting the nod than if they both win by the same amount, etc.

It actually simulates the games (20,000 simiulations), so it is taking that MOV into account. The methodology is interesting. Some relevant passages:

Quote:The key characteristics of the model are that it’s iterative and probabilistic. It’s iterative in that it simulates the rest of the college season one game (and one week) at a time, instead of jumping directly from the current playoff committee standings to national championship chances. And it’s probabilistic in that it aims to account for the considerable uncertainty in the playoff picture, both in terms of how the games will turn out and in how the humans on the selection committee might react to them.

Games are simulated mostly using ESPN’s Football Power Index. We say “mostly” because we’ve also found that giving a little weight to the playoff committee’s weekly rankings of the top 25 teams helps add to the predictions’ accuracy. (We use the Associated Press Top 25 poll as a proxy for the committee’s rankings until the first set of rankings is released in the second half of the season.) Specifically, the model’s game-by-game forecasts are based on a combination of FPI ratings and committee (or AP) rankings — 75 percent on FPI and 25 percent on the rankings.1

Quote:Thus, teams from the Power Five conferences2 — especially the SEC — start out with a higher default rating.3 As a consequence of this, our system also gives teams from power conferences more advantages, because that’s how human voters tend to see them.

This conference-centric approach both yields more accurate predictions of game results and better mimics how committee and AP voters rank the teams. For better or worse, teams from non-power conferences (except Notre Dame, that special snowflake among independents) rarely got the benefit of the doubt under the old BCS system, and that’s been the case under the selection committee as well.

https://fivethirtyeight.com/methodology/...ions-work/

(11-27-2018 04:29 PM)stever20 Wrote:  So i was looking at the ESPN FPI numbers for this weekend
Alabama- 64% chance of winning
Clemson- 95.6% chance of winning
Oklahoma- 72.8% chance of winning
Ohio St- 83.9% chance of winning

so I come up with a few scenarios...
Oklahoma/Ohio St both win- 61.1% chance of happening. Would be controversial no matter how Alabama does.
Texas/Ohio St/Georgia all win- 8.2% chance of happening.
Oklahoma/Northwestern/Georgia all win- 4.2% chance of happening
Texas/Northwestern/Alabama all win- 2.8% chance of happening
All 4 of these would produce scenarios that would be contrversial for committee..... You add them up and we're at 76.3% chance of controversy. Might have something more with Clemson in there losing that would add to the controversy....

Chaos really occurs if two out of three of Clemson, Oklahoma and Ohio State lose. It is not really a significant problem if the Committee has too many schools to choose from. This has happened before. However, if two of these three schools lose, the committee will have no obviously good selection for the fourth playoff spot, with Notre Dame and the SEC champ each assured of a spot.

Using the numbers above, I calculate the following chances for two out of three of these schools to lose:

CU + OU + OSU win 58.39%
CU + OU 11.21%
CU + OSU 21.82%
OU + OSU 2.69%
Pitt + UT + NU win 0.19%
Pitt + UT 1.00%
Pitt + NU 0.52%
UT + NU 4.19%

Chance of two of three losing is 5.71% (1.00+.52+4.19). Chance of all three losing is .19%.

If two out of three lose, the committee could be saved by Georgia defeating Alabama. Alabama has been so strong this year, it would be a reasonable selection for the fourth playoff slot in any of these scenarios. Factoring in the potential for a Georgia win, the chaos arising from the scenario resulting from two out of three of Clemson, Oklahoma and Ohio State losing is reduced from 5.71% to a 3.65% chance that this happening would result in there not being four clear schools for the committee to choose.

Adding in the .19% chance that all three of Clemson, Oklahoma and Ohio State losing, and the potential for the lack of four clear teams for the committee to select is 3.84% (.19+3.65).