538 Idea for football scheduling

UpTheMiddlex3Punt

All-Conference
May 28, 2007
18,207
4,353
113
I know all the SPS Liberals** are big into fivethirtyeight, so they've probably already read this article. For the rest of us, here it is: http://fivethirtyeight.com/features/make-college-football-great-again/

Essentially, teams in a conference like the SEC (Big Ten is used as an example in the article) will play a nine game conference schedule (non-conference and bye weeks would be scattered in the season somehow). Five of the games will be predetermined each year, some based on traditional rivalries, and others based on random draws or cycled over the years. The first four games each year are played against predetermined opponents and then the teams are paired up based on record where you play the team not already on your schedule that has the closest record to you. If your schedule was front-loaded with cupcakes, you'll quickly be beaten back into submission against actually good teams. If you had a tough break to start the season, you'll get a chance to beat up on a lesser team to climb back up. This repeats every week there is not a predetermined matchup, so by the end of the season you'll have a situation where the three or four best teams in the conference will have all played each other. There is no conference championship game, but you could declare a conference champion based on record and tie-breaks. You would also have one or more games at the end of the season that would be de facto championship games.

I know there's been talk about getting rid of the championship games because they're either a big mismatch (see almost every SECCG since 2009) or irrelevant (ahem, Penn State this year) to the national championship picture. You would also avoid situations where a team might make it to the championship because of a weaker cross-division schedule.

The biggest disadvantage of this type of scheduling is logistics. While you would know your home game dates in advance, going to away games would be nearly impossible to schedule, not to mention teams would have to make travel plans on a week-by-week basis and lining up charter planes a week out would be pricey.
 

DawgatAuburn

All-Conference
Apr 25, 2006
11,192
2,326
113
I only read your recap, but that seems like an insanely terrible idea. As a fan, you have to be able to plan where you are going more than a week out.
 

EurekaDog

Redshirt
Nov 10, 2010
598
0
0
Travel Plans...

For the TBD games, I would expect the SEC Office to reserve the hotel rooms and charter planes and/or buses months prior to the season. The SEC office and the schools would work out an agreement on the finances.

They should not tamper with the rivalry games.

I would expect that ticket prices could spike or bottom out depending upon the situation in addition to trying to make sure your group can sit together at the stadium.

As others have noted, the fans would be the most negatively affected in regard to travel plans.
 

EngDawg

Redshirt
Mar 29, 2016
2,037
4
38
Has the election caused Nate Silver to go off his rockers? This idea is silly.
 

Drebin

Heisman
Aug 22, 2012
22,953
28,645
113
Has the election caused Nate Silver to go off his rockers? This idea is silly.

You mean the guy who predicted Clinton had more than an 80% chance of beating Trump, and then wrote an article talking about how he was correct after the election?

Yeah, he's full of great ideas, I'm sure.
 

dickiedawg

All-Conference
Feb 22, 2008
4,606
1,577
113
You mean the guy who predicted Clinton had more than an 80% chance of beating Trump, and then wrote an article talking about how he was correct after the election?

Yeah, he's full of great ideas, I'm sure.

That's the guy* who had Trump's odds much, much higher than anyone else in the industry and wrote articles (before the fact) defending them against everyone saying they were too low.

(538 is actually much larger than just Nate Silver, but he did write the defense of the model when others were attacking it for "overinflating" Trump's chances)
 

tired

All-Conference
Sep 16, 2013
3,772
1,571
113
Maybe a schedule based on the previous year's rankings. That would give several months to prepare.
 
Feb 4, 2015
1,060
66
48
I do wished something could be done tochange the schedules. I for one am sick of having to play Bama,auburn & lsu every single year. I just wished the conference could rotate every year a lot of the teams and have teams play more equally the number of east teams and west teams.
 

patdog

Heisman
May 28, 2007
61,325
33,041
113
That site jumped the shark a while ago. The logistics alone with this idea would be a nightmare, if not impossible. You couldn't even predetermine your home dates with this format. What happens if both teams are "supposed" to play at home or both on the road? How are fans supposed to be expected to go to road games when they wouldn't have any idea where they're going until the week of the game?
 

Cousin Jeffrey

Freshman
Feb 20, 2011
770
50
28
You couldn't even predetermine your home dates with this format.
That is actually addressed in the article: "Each team has two home flex games and two away flex games, with the weeks designated in advance"

What happens if both teams are "supposed" to play at home or both on the road?
Also addressed: "We take the 14 Big Ten teams and split them into pools of seven home teams and seven away teams based on where they’d been assigned to play ahead of time."

How are fans supposed to be expected to go to road games when they wouldn't have any idea where they're going until the week of the game?
That's a legitimate concern.
 

patdog

Heisman
May 28, 2007
61,325
33,041
113
The first two may be "addressed" in the article, but they're not adequately addressed. It would be a freaking nightmare to implement. To the point it's purely a fantasy dream world. Which that guy seems to make a habit of living in.
 
Jul 29, 2013
211
0
0
He himself did not predict a Trump win. His formula did. His model is heavily influenced by actual poll numbers…. which were heavily influenced by previous election turnouts….. we all see how that turned out.
Like him or not, his model was actually closer than all the other models that factor in multiple (hundreds) of polling results. People don't understand that most of these "election prediction" formula's are driven by polling numbers and not political bias. The polls are the ones hammering out the political bias.
 

Go Budaw

Redshirt
Aug 22, 2012
7,321
0
36
He himself did not predict a Trump win. His formula did. His model is heavily influenced by actual poll numbers…. which were heavily influenced by previous election turnouts….. we all see how that turned out.
Like him or not, his model was actually closer than all the other models that factor in multiple (hundreds) of polling results. People don't understand that most of these "election prediction" formula's are driven by polling numbers and not political bias. The polls are the ones hammering out the political bias.

Very true. Not to mention that this year's election was an anomaly unlike any other in history, with a candidate who won the popular vote by almost 3 million votes not even coming close to winning the electoral college. The national popular vote results were within the margin of error of the model. Where those votes were was what threw it off, as no one could have predicted the Rust Belt shift based on the data that was available.
 
Last edited:

JungRebel

Redshirt
Aug 23, 2012
2,606
0
0
One clue that was available was Sanders' success in winning the Michigan and Wisconsin primaries. Not that anyone could have predicted Trump wins on that data alone, but it was a clue. Hillary racked up votes and states in the Bible Belt in primaries that should've never mattered to the Democrats because they would never win them in the general. But that wouldn't have made a difference to them, they had their candidate picked a long time ago and pushed the narrative in that direction, electorate be damned.
 
Last edited:

LesterFreeman

Redshirt
Jul 23, 2016
21
0
0
Very true. Not to mention that this year's election was an anomaly unlike any other in history, with a candidate who won the popular vote by almost 3 million votes not even coming close to winning the electoral college. The national popular vote results were within the margin of error of the model. Where those votes were was what threw it off, as no one could have predicted the Rust Belt shift based on the data that was available.

Hillary needed 38 Electoral Votes.
These 3 states were the difference for Trump - about 100k votes out of 13.5 million if my quick math is right.
MI by 11k out of 4.5 million 16 Electoral Votes
PA by 68k out of 6 million 20 Electoral Votes
WI by 27k out of 2.8 million 10 Electoral Votes

Out of the number of votes FL was pretty close.
FL by 120k out of 9 million 39 Electoral Votes

Had the results been closer (say Hillary took MI and PA) then it would have been interesting to see what the 7 electoral voters who didn't vote as pledged would have done, and how nuts it really could have gotten.
Pledged toVoted for
HawaiiClintonBernie Sanders
TexasTrumpRon Paul
TexasTrumpJohn Kasich
WashingtonClintonColin L. Powell
WashingtonClintonColin L. Powell
WashingtonClintonColin L. Powell
WashingtonClintonFaith Spotted Eagle

<thead>
</thead> <tbody>
</tbody>
 

LesterFreeman

Redshirt
Jul 23, 2016
21
0
0
It's an interesting idea. Based off the way TV currently works, I think you would have to see the games set two weeks in advance even though that probably throws off his model some. Travel for visiting fans would be tough, but that only comes in to play for 2 games a year.

In the end, you will never get a perfect situation, That is part of what makes college football what it is, the fun of debating who is the best team....behind bama.
 

BiscuitEater

Redshirt
Aug 29, 2009
4,178
0
36
He himself did not predict a Trump win. His formula did. His model is heavily influenced by actual poll numbers…. which were heavily influenced by previous election turnouts….. we all see how that turned out.

errr .. WRONG. His 'formula' did NOT predict a Trump win. Here are Nate Silver's 'words' right after the election ..

Nate Silver was on The Daily Show last night, taking something of a victory lap because he wasn’t as wrong as the other statistical wizards when it came to predicting the race for president.
But make no mistake, Silver was wrong. He was wrong the entire race. He was wrong the day before the election. He was wrong when Florida went for Trump.

Now the statistical experts will tell you he wasn’t wrong because he gave Trump a chance of winning and that’s the way numbers work. And that’s true. Silver wasn’t wrong because he never said Donald Trump couldn’t win.

Read 'his' own words, from HIS 538 Site .. http://blogs.mprnews.org/newscut/2016/11/nate-silver-does-a-victory-lap/
 

Arthur2478

Redshirt
Oct 17, 2010
1,407
2
38
I know there's been talk about getting rid of the championship games because they're either a big mismatch (see almost every SECCG since 2009) or irrelevant (ahem, Penn State this year) to the national championship picture.

There has been zero talk of dropping these games by the conferences that commission them. These games produce huge television dollars for the leagues and thus aren't going anywhere. Any "talk" is just internet blotter or brainless sports radio banter.
 

patdog

Heisman
May 28, 2007
61,325
33,041
113
Agreed. Anyone who thinks dropping the championship games would even be considered for one second by the conferences is crazy. We may see them restructured so that they match up the top 2 teams instead of 2 division championships, but they ain't going away by any means.
 

QuaoarsKing

All-Conference
Mar 11, 2008
6,373
3,312
113
Well I don't even know what to say. I've been brainstorming an idea on and off for several years kinda like this, and I even thought about registering a website makecollegefootballgreatagain.com to discuss it. Now if I ever do, people would assume I ripped off Nate Silver. I snooze, I lose. 😕

In my plan, I had the paired games in weeks 1, 5, 9, and 14 to give about a month's notice, but otherwise it's pretty similar - matching up teams with similar records to create interesting matchups that fans would love and would give us a lot more data to accurately rank/rate teams.

I think people are getting too worked up about the uncertainty of the schedules. If Nate's plan were in place for us, our preseason schedule would have 8 definite games, 2 listed as "home" and 2 listed as "away." The vast majority of fans only go to home games, so there would be no problems planning ahead. If you like going to road games, it would be tough, but they could set it up where they always give a few weeks of notice, like in my plan. As far as logistics for the teams go, the SEC would just have to have reservations already made at the 7 home sites for that week and give them to the road team. After a year or two, it would be second nature to everyone.
 

Maroon Eagle

All-American
May 24, 2006
18,674
8,702
102
Agreed. Sanders winning those primaries showed that establishment candidates would have a tough time in those states & Trump ran successfully as anti-establishment candidate.
 

QuaoarsKing

All-Conference
Mar 11, 2008
6,373
3,312
113
Agreed. Anyone who thinks dropping the championship games would even be considered for one second by the conferences is crazy. We may see them restructured so that they match up the top 2 teams instead of 2 division championships, but they ain't going away by any means.

It's possible (likely?) that conferences could bring in more money from an extra playoff round than they get in CCGs, and if there's too much opposition to adding an extra game from teams, that could be the route to axing them.

Remember that the Big 12 dropped their championship game and is only bringing it back begrudgingly after being left out of the playoff twice in 3 years. I'm going to laugh hard if they have a team next year that's probably in and then gets upset in the championship game and is out.
 

QuaoarsKing

All-Conference
Mar 11, 2008
6,373
3,312
113
538 deserves some credit for at least acknowledging the possibility of a Trump win, which very few pundits, including those on the right, were willing to do.

And considering how Trump pulled out an inside straight, flipping 3 blue states by razor thin margins to take the electoral college despite a 3 million vote deficit nationally, I don't think his 70/30 odds were bad. The national result was within a point of the national polling average and no swing state was outside the margin of error of the polling average. I wish pundits (other than 538 ) hadn't decided to ignore all that.
 

LilSebastian

Redshirt
Sep 13, 2012
175
0
0
errr .. WRONG. His 'formula' did NOT predict a Trump win. Here are Nate Silver's 'words' right after the election ..

Nate Silver was on The Daily Show last night, taking something of a victory lap because he wasn’t as wrong as the other statistical wizards when it came to predicting the race for president.
But make no mistake, Silver was wrong. He was wrong the entire race. He was wrong the day before the election. He was wrong when Florida went for Trump.

Now the statistical experts will tell you he wasn’t wrong because he gave Trump a chance of winning and that’s the way numbers work. And that’s true. Silver wasn’t wrong because he never said Donald Trump couldn’t win.

Read 'his' own words, from HIS 538 Site .. http://blogs.mprnews.org/newscut/2016/11/nate-silver-does-a-victory-lap/

None of those are his words....
 

Go Budaw

Redshirt
Aug 22, 2012
7,321
0
36
errr .. WRONG. His 'formula' did NOT predict a Trump win. Here are Nate Silver's 'words' right after the election ..

Nate Silver was on The Daily Show last night, taking something of a victory lap because he wasn’t as wrong as the other statistical wizards when it came to predicting the race for president.
But make no mistake, Silver was wrong. He was wrong the entire race. He was wrong the day before the election. He was wrong when Florida went for Trump.

Now the statistical experts will tell you he wasn’t wrong because he gave Trump a chance of winning and that’s the way numbers work. And that’s true. Silver wasn’t wrong because he never said Donald Trump couldn’t win.

Read 'his' own words, from HIS 538 Site .. http://blogs.mprnews.org/newscut/2016/11/nate-silver-does-a-victory-lap/

His model predicted a popular vote win by Hillary of 4% with a + / - 2% margin of error. Hillary won the popular vote by 2% (within the margin of error). His model also assumed that the nationwide popular vote trend would translate to the swing states and states that were "leaning" one way or the other to solidify them. His model assumed that because that is what has happened in pretty much every presidential election throughout history.

Was he wrong in predicting the eventual winner? Of course he was. So was everyone else. But, if you walked into a casino and said you wanted to place a bet that the incumbent party candidate would win the popular vote by a margin that is roughly equal to the entire population of Mississippi, yet still lose every swing state and also lose 3 safe states that were carried by the incumbent party in the last 2 elections (including one that was carried for about 30 years by the same party), you would have gotten 10,000 / 1 odds. No one saw it coming. Predicting the aggregate political sentiment of 130 million+ individuals is not an easy thing to do. Silver and his team do it a lot better than just about everyone else. I say that as a stats geek, not as an apologist for either candidate or political party.
 

JungRebel

Redshirt
Aug 23, 2012
2,606
0
0
Exactly. Silver woke up November 9th as one of the few people who actually said it could happen. Working nights I watched the whole thing as it went on. Trump seemed as surprised as all else.
 
Aug 15, 2011
786
403
63
Screw that, I want a 64 team playoff like basketball.....

Reduce the regular season to 11 games. For the eleventh game, the division winners play in their respective conference championships while everybody else plays a pre-determined opponent from their conference's other division. Conference championships get home field advantage in the playoff. The rounds of 64, 32, and 16 are played on campus, quarter and semi-final games are played at the current New Year's six bowls. Teams that lose in the rounds of 64, 32, and 16 could accept a bowl invitation to one of the lesser bowls (the higher the round you make, the better the bowl). Winning your bowl game would count into next year's selections to give the teams something to play for. The season would only be extended to the max of 17 games for 2 of the teams. Selection Sunday for college football would blow away the basketball version. It'll never happen though.
 

Hoover Dawg

Redshirt
May 16, 2016
1,584
0
0
I like the idea of reducing the season to 11 games and having everyone play on championship Saturday. But I'd like that to be the 12th game (that's probably what you meant). That way, everybody has a 12 game record to compare, and teams that play a round robin like the Big 12, can just keep the 12 game schedule they currently have. 'Conference Champion' would be all but obsolete for the playoff committee.

Other than that, I see no reason to make the playoff any bigger than it is. Maybe expand to 6 games with the #1 and #2 teams getting byes for a reward. But that's it, IMO.
 

Hoover Dawg

Redshirt
May 16, 2016
1,584
0
0
At some point, you're going to have to realize that statistics can be interpreted a number of ways. Nate Silver got it dead wrong, because his statistics were wrong, because what they were based on was wrong. Just like all the other people.

They ignored half of the country. The census does this too. I have direct experience with this. Just because you have a bunch of numbers, doesn't make you smart or doesn't mean you have good data.
 

FQDawg

Senior
May 1, 2006
3,076
618
113
My crazy idea...

Get rid of regional conferences and set up ten 13-team divisions. You play everyone in your division once for a 12-game schedule. At the end of the season, each division would have a champion (based on record with appropriate tie-breakers as necessary) and you would have promotion/relegation for the top/bottom three teams. Sort of like how soccer is handled everywhere in the world except the U.S. That way, teams could work themselves up through the divisions over time based on what happens on the field. And, conversely, teams wouldn't be artificially ranked because of name-recognition. If you lose enough games, you'd drop down to the next lower division.

Also, you could easily incorporate 1-AA (or whatever they're called now) as Divisions 11 through whatever, though you may want to do that more regionally since they have less money for travel.

I'm sure someone smarter than me (or a computer) could figure out scheduling so that each team had six home games and six road games a year.

With this system, there'd be no need for a post-season (because the regular season would be all that matters, along with the color maroon) but you could keep the bowl system in place as a set of post-season exhibition games.
 
Last edited:

patdog

Heisman
May 28, 2007
61,325
33,041
113
My crazy idea is similar, but probably more workable. Except it's 8 groups of 9 teams each in the top flight, for 72 teams total. The 8 group winners advance to the playoffs. Last place in each group is relegated and the top 8 from the 2nd division are promoted. You play the other 8 teams in your group and have 4 games to schedule against traditional rivals, or whoever you want to. Bowls could continue to exist much like they do now, taking the non-playoff teams.
 

JungRebel

Redshirt
Aug 23, 2012
2,606
0
0
It's hard to reach 'half the country' when less than ten percent of people return your calls. Not that I blame them, I wouldn't talk to pollsters over the phone.
 
Last edited:

Go Budaw

Redshirt
Aug 22, 2012
7,321
0
36
At some point, you're going to have to realize that statistics can be interpreted a number of ways. Nate Silver got it dead wrong, because his statistics were wrong, because what they were based on was wrong. Just like all the other people.

They ignored half of the country. The census does this too. I have direct experience with this. Just because you have a bunch of numbers, doesn't make you smart or doesn't mean you have good data.

Which "half of the country" was ignored? Last I checked, 50% of the voters in Michigan, Pennsylvania, and Wisconsin does not equal half of the country. The statistics, polling, and model were dead on pretty much everywhere else, including the nationwide popular vote.
 

Hump4Hoops

Redshirt
May 1, 2010
6,611
13
38
You couldn't be more wrong. Silver had the best statistical model.

He pulled as much data as he possibly could into his model, not "ignoring half of the country." If anything, it was the people that pollsters were trying to reach either lying or refusing to answer that was a large part of the results not matching the predictions based on polls. If he had "predicted" a Trump win, it would have been with a lack of any evidence supporting that.
 

Hoover Dawg

Redshirt
May 16, 2016
1,584
0
0
Exactly. The data was wrong. So why would anybody listen to it?

It was the only data you say? BS. I saw the turnouts in the primaries and suspected that it would be very close, not a Clinton blow out.

People like you also need to realize that just because there is not a study-specific set of data backing something up, doesn't mean it is wrong, no matter what it is.
 
Last edited:

JungRebel

Redshirt
Aug 23, 2012
2,606
0
0
To me, Trump winning that election is like me flipping a coin heads ten times in a row. The nature of probability is that even the person who gets the right answer never gets it for all the right reasons, otherwise the model would be mechanical and fully explicable. If I flip a coin and call it correctly I never 'knew' which way it would go, my answer is a guess. The problem for pollsters is we hold them responsible for their guesses, but they make money doing what they do so I don't mind ridiculing them. Their models should be more accurate than a coin flip, and most of the time they are, but there have been some famous misses. As you said, no one will ever be able to say 'Why' fully.
 
Maroon and White

Latest

  1. 01 Maroon and White "I started something I didn't finish": Mississippi State QB Kamario Taylor joins Sports Center to talk 2027 return
  2. 02 Maroon and White Basketball Bulldogs prepare to level up competition, starting with Xavier in neutral site meeting
  3. 03 Maroon and White Football Mississippi State Week 11 Preview: Who carries the load?
  4. 04 Maroon and White Football Mississippi State Week 11 Preview: Run vs. Run
  5. 05 Maroon and White Basketball Mississippi State 79-57 Troy: Bulldogs build big early lead in home win over Trojans