538's college football playoff projections

by Herbie Green 2016-11-13, 10:16

Pretty sweet.

http://projects.fivethirtyeight.com/2016-college-football-predictions/

by Herbie Green 2016-11-13, 11:13

Dwags. let me know if you need me to help explain this to you.

Louisville is in an interesting position at #4 with a 48% of making the playoff but it only goes up to 60% if they win out. This is where the model has to factor in the playoff committee criteria and human behavior. I don't think they have taken two teams from the same conference in the first two years? Louisville will have a very strong case this year

by **DWags** 2016-11-13, 12:49

Herbie Green wrote:Dwags. let me know if you need me to help explain this to you.

Louisville is in an interesting position at #4 with a 48% of making the playoff but it only goes up to 60% if they win out. This is where the model has to factor in the playoff committee criteria and human behavior. I don't think they have taken two teams from the same conference in the first two years? Louisville will have a very strong case this year

I get this. Could you explain 538's election forecast?

Em thanks.

by **NigelUno** 2016-11-13, 13:39

DWags wrote:
Herbie Green wrote:Dwags. let me know if you need me to help explain this to you.

Louisville is in an interesting position at #4 with a 48% of making the playoff but it only goes up to 60% if they win out. This is where the model has to factor in the playoff committee criteria and human behavior. I don't think they have taken two teams from the same conference in the first two years? Louisville will have a very strong case this year

I get this. Could you explain 538's election forecast?

Em thanks.

Herbie's not the brightest bulb.

If Clemson loses another regular season game, and Louisville wins out...Louisville will make the playoff.

If Clemson gets upset in the ACC Championship, and Louisville wins out...Louisville will make the playoff.

This brief (but enlightening for Herbie) foray into the "sports discussion" board is now concluded.

by **Blanch32** 2016-11-13, 22:24

DWags wrote:
Herbie Green wrote:Dwags. let me know if you need me to help explain this to you.

Louisville is in an interesting position at #4 with a 48% of making the playoff but it only goes up to 60% if they win out. This is where the model has to factor in the playoff committee criteria and human behavior. I don't think they have taken two teams from the same conference in the first two years? Louisville will have a very strong case this year

I get this. Could you explain 538's election forecast?

Em thanks.

lol. math clearly isn't your strong suite

by Herbie Green 2016-11-13, 23:17

NigelUno wrote:
DWags wrote:

I get this. Could you explain 538's election forecast?

Em thanks.

Herbie's not the brightest bulb.

If Clemson loses another regular season game, and Louisville wins out...Louisville will make the playoff.

If Clemson gets upset in the ACC Championship, and Louisville wins out...Louisville will make the playoff.

This brief (but enlightening for Herbie) foray into the "sports discussion" board is now concluded.

"If, If, if"

If UM, which currently has a 33% chance of making the playoffs wins out they have a 98% chance of making the playoffs. So they control their own destiny. Do you understand the difference?

I wouldn't have thought I had to explain that biut You just can't see the forest through the trees. You get caught
up on what is called "noise"

by **DWags** 2016-11-14, 06:01

Herbie Green wrote:
NigelUno wrote:

Herbie's not the brightest bulb.

If Clemson loses another regular season game, and Louisville wins out...Louisville will make the playoff.

If Clemson gets upset in the ACC Championship, and Louisville wins out...Louisville will make the playoff.

This brief (but enlightening for Herbie) foray into the "sports discussion" board is now concluded.

"If, If, if"

If UM, which currently has a 33% chance of making the playoffs wins out they have a 98% chance of making the playoffs. So they control their own destiny. Do you understand the difference?

I wouldn't have thought I had to explain that biut You just can't see the forest through the trees. You get caught
up on what is called "noise"

538's college football playoff projections 502811600

by Herbie Green 2016-11-14, 07:25

Nigel does accidentally bring up an interesting question though. I was thinking about a 2nd team from a conference making it but what about a lone representative from a conference that did not even make its conference championship game?
There is not a single example of this happening in the playoff's brief existence either, yet Nigel thinks if Clemson loses to South Carolina and/or the ACC title game that the committee will reward Louisville for sitting at home. This would be even more controversial. On paper Louisville would have a higher power ranking than the ACC champ but I have a hard time seeing the committee giving them a bid. They would like to lock out the ACC entirely at that point. Yet they will undoubtedly have a higher ranking than at least one other conference champ as well.
This is the kind of chaos I root for at this point.

by **NigelUno** 2016-11-14, 09:16

DWags wrote:
Herbie Green wrote:

"If, If, if"

If UM, which currently has a 33% chance of making the playoffs wins out they have a 98% chance of making the playoffs. So they control their own destiny. Do you understand the difference?

I wouldn't have thought I had to explain that biut You just can't see the forest through the trees. You get caught
up on what is called "noise"

This board is "noise".

Herbie's bulb gets dimmer by the day.

by Herbie Green 2016-11-14, 09:54

Here is what Silver does in his model. Note this is a footnote of a change to the first year's model that had TCU with a 91% chance of making the playoff:

"Second, we assign a small bonusx
Because we don’t have a firm idea of how much the committee rewards conference champions, we treat the magnitude of the conference championship bonus as being uncertain, and it varies from simulation to simulation. In some simulations, winning the conference championship is associated with a fairly large bonus; in others, it’s associated with no bonus at all. (The bonus can never be negative, however.) On average, however, it’s fairly small, and it acts as the equivalent of a tiebreaker in otherwise close cases"

by Herbie Green 2016-11-14, 10:43

So we don't know what this factor is but I am going to guess from that wording that it is too small. IMO that factor is somewhere between two full games and an outright bare minimum for consideration.
It might not be smart of me to go against Silver but he has been wrong before and in this case he has no data to back up his assumption.

by **DWags** 2016-11-14, 12:15

As a layman, I can pretty much tell you that if Michigan wins out, they're in. If Wisconsin wins out, they're pretty close to being in, I'd say better chance being in the playoffs than out. OSU is the most interesting case we have in my opinion. They win out, they have the hyde of the #4 team on their wall, they went down south and beat Oklahoma, they have the #6 teams hyde on their wall in a night game at Wisconsin, they lost to a top 10 team in a night game at that teams house, on a blocked field goal. But they will not be in the championship game for their own conference.

That's a damn good resume. How do you keep them out? If Wisconsin wins out and beats PSU in the championship game, how do you put them in over OSU? They lost at home to OSU.

I just love the chaos. Next two weeks will no doubt straighten some of this stuff out. But all in all, it will be fun.

by **Robert J Sakimano** 2016-11-14, 12:20

I'm going to feel so much better about myself if scUM doesn't make it.

by **Death Roe** 2016-11-14, 12:46

Yeah, 538 is really accurate. They got the election right.... 538's college football playoff projections 502811600

by **Blanch32** 2016-11-15, 07:54

Death Roe wrote:Yeah, 538 is really accurate. They got the election right....

only 2% off. so they sunshine and you simpleton

by Herbie Green 2016-11-15, 08:16

Yeah, I was just about to addtess this topic. 538 had Trump with a 29% chance of winning, more than about every other forecast. They were widely criticized for giving Trump such a high chance yet still went on to warn that Trump could actually win with just a small sample error.

Here is Nate Silver's article that explains it very well.

http://fivethirtyeight.com/features/why-fivethirtyeight-gave-trump-a-better-chance-than-almost-anyone-else/

by Herbie Green 2016-11-15, 08:33

Back to the playoff model, the Big Ten might end up the best test of the conference champion bonus and two teams from a conference question.
538 has Ohio State with a 95% chance of making the playoffs if they win out and Penn State with an 18% chance if they win out. Yet if both of these things happen then Penn State is winning the big ten championship while OSU is sitting at home.
Once again I think Silver is underestimating the conference title factor. Maybe that is my own bias of how I think it "should" work. I like things being settled on the field. How can you say "well Penn State won the Big Ten Championship and beat Ohio State but Ohio State is really the better team so they are going to the playoffs as the Big Ten representative".
I suppose with my line of thinking you might as well just pick four out of five conference champions and end this whole charade.

by **Blanch32** 2016-11-15, 10:02

Herbie Green wrote:Back to the playoff model, the Big Ten might end up the best test of the conference champion bonus and two teams from a conference question.
538 has Ohio State with a 95% chance of making the playoffs if they win out and Penn State with an 18% chance if they win out. Yet if both of these things happen then Penn State is winning the big ten championship while OSU is sitting at home.
Once again I think Silver is underestimating the conference title factor. Maybe that is my own bias of how I think it "should" work. I like things being settled on the field. How can you say "well Penn State won the Big Ten Championship and beat Ohio State but Ohio State is really the better team so they are going to the playoffs as the Big Ten representative".
I suppose with my line of thinking you might as well just pick four out of five conference champions and end this whole charade.

this. either they both make it or both are left out.

Louisville is in the same scenario fyi

by Herbie Green 2016-11-15, 21:38

Is Herbstreet just talking out of his ass? I haven't watched this show until now. He seems to be more on Silver's side but I don't know if he is a paid shill

by **Blanch32** 2016-11-15, 21:41

uofm #3 lol

by Sponsored content

538's college football playoff projections

538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections

Re: 538's college football playoff projections