I_Run_The_Dave's picture

I_Run_The_Dave


MEMBER SINCE   October 20, 2012

Favorites

  • COLLEGE FOOTBALL PLAYER: JT Barrett and Happy Jamarco
  • NFL TEAM: I care about Buckeyes in the NFL, not NFL teams.

Recent Activity

Comment 15 Oct 2019

But the computers don't generate talking points and speculation throughout the season.  They are cold and emotionless.  Unlike a committee full of people with changing criteria and conflicts of interest.

Comment 15 Oct 2019

I guess SDSU, Fresno, and Ga St aren't quite as good a FAU, Miami (OH) and Rutgers, but it's close.  :)

Comment 15 Oct 2019

They'd probably bring back the Harris in lieu of the AP poll again.  Maybe keep the current committee members and give them the votes for it.  I also agree that the coach's poll is suspect and less stock should be put into it (since you have GA's filling them out -- let's not give them votes right?).

Personally let's just abolish everything, end the season today, and go with the Massey Composite.  We can still play the rest of the season in an exhibition capacity, but I'll take that trophy now, please.  :)

Comment 15 Oct 2019

That means Minnesota has wins over Wisconsin and Ohio State.  Of those, you'll have Ohio State at either 12-1 and Wisconsin at 10-2, or you'll have Ohio State at 11-2 and Wisconsin at 11-1 depending on the outcome.  Either way that's a strong resume (yes I'm discounting PSU here).  And it's stronger than anything Clemson has a chance of putting together.  So if you are going to leave out an undefeated P5, it's Clemson, not Minnesota, that gets left out.

Let's then look at Alabama's resume.  The only ranked teams they'd have played is LSU (loss) and Auburn (win) assuming Auburn stays ranked (with 3 losses at that point).  Absolutely no chance they make it in.  That's a worse resume than Clemson even because Clemson wouldn't have a loss on their record.  Likewise you'd take Boise State's garbage schedule over Minnesota with 2 top 10 or so wins?  No way.  Boise has no chance either.

The field would be LSU, Oklahoma, Minnesota, then you could debate Clemson and Oregon and Alabama.  Alabama would be better than Oregon due to Auburn having 3 losses at that point and Alabama winning that game versus Oregon's loss.  Alabama vs Clemson would be a "who is the media darling going to be?" debate.  I don't care who wins that, but let's say Clemson but both schedules would be garbage at that point.  Then don't you consider a 1 loss Ohio State in that mix who would have wins over PSU and Wisconsin?  Or a 1 loss Wisconsin that beat OSU?  Either of those could be stronger arguments than Alabama at that point.

Only 1 loss SEC team that has a chance would be an 11-1 LSU because they have strong wins against the SEC East top teams and over Texas.  You are also potentially ignoring a possible 12-1 Florida team (though I think Georgia beats them and ends up 11-2 with Florida at 10-2, eliminating both).  Florida's only loss would also be to LSU, so wouldn't Florida have a better SoS than Alabama with the extra win?

Comment 15 Oct 2019

Hot take: based on Clemson's gradual decline week to week, they may be the first unbeaten to not make the playoff.  If Oklahoma wins out and either Wisconsin or Ohio State do as well, and we have a 1 loss SEC runner up, the SEC might bump the defending national champions out.  I can already see that happening in the computer numbers, and the committee may follow suit when the time comes.

I agree with everything you said, btw.  Unless we get punched in the mouth in 2 weeks (I don't think that will happen)

Comment 15 Oct 2019

I cannot answer the question as to why the BCS chose certain computer models, other than that these were existing in 1998 when the BCS was formed.  They also were not allowed to use margin of victory and there were other controls as to how they were to be set up that probably factored into the decision.  I know some of them do conference ranking and then weight your strength of schedule based on that to influence a final rating for each team.  The Colley Matrix doesn't consider conference at all and just goes off of wins and losses (and recursively recalculating based on opponent rating until it converges).  That's also why Colley has Boise St at 3 (so does Wolfe) when Sagarin and Massey have them at 22 and 19.

There has been controversy over how it selects teams since its inception, obviously, especially when there is any disagreement between the computers and the human polls.  But...  if we are going to complain about computers having different answers than humans, why use the computers at all?  The computers are there to attempt to remove bias and poll inertia, and they do a good enough job of that to counter balance the human polls, even if they only get 1/3 of the vote.

I didn't personally start looking into these until the last few years of the BCS system, and it became a curiousity when the CFP was instituted.  It's interesting that the BCS actually owns the CFP and that the top 4 and the resulting matchups have always agreed between the committee and the old BCS formula.  When Ohio State was picked over TCU and Baylor in 2014, the BCS would have done the same at the time.  It's funny to me when writers pine for bringing it back only expanded to 4 teams when we basically have that already even if unofficially.

Comment 15 Oct 2019

You stole my thunder regarding the BCS simulated standings.  I was going to complain that they were wrong, because I hadn't added Anderson and Hester's rankings this week to my formula and last week they has LSU at #1 and it seemed unfathomable that they would drop after beating Florida.  Their "Colley" column is wrong because they copied Billingsley over it on accident, but now that I have the Anderson and Hester numbers, their computed averages are identical to mine.  Not sure if I should bother making my weekly post about it since it's here now.

Comment 14 Oct 2019

It probably will boost ratings for whatever is at the 3:30 time slot as well.  They know OSU brings the ratings regardless of timeslot so this is probably a good financial strategy for them.  They could care less about the fans tuning in, however.

Comment 09 Oct 2019

He went to a 40k per year academy and trained with tennis pros his whole life.  I wouldn't describe him as "poorer".  Maybe "less ridiculously wealthy" or "only a millionaire family" or something.

The article is also misleading in that the other kid (Grant) didn't get Adam's spot, Adam probably legitimately got his offer at cal poly pulled and it might have had nothing to do with the scandal.  The other kid got into Georgetown fraudulently, for sure, but Adam didn't even want to play there because he wanted to go to a school with a better program.

And he could have walked on any team he wanted in division 1, would have been good enough to earn playing time, and the tuition costs for his parents would have been less than his high school.  So no, I really don't feel too bad for the kid.

Comment 09 Oct 2019

Lineman can declare themselves eligible to an official prior to a play if they are the end man on the line of scrimmage.  It doesn't happen often and it basically tips your hand to the defense, but it has happened on trick plays before.

Comment 08 Oct 2019

This technically means that ESPN's FPI and SP+ are out since they use preseason data in their calculations throughout the season.  Despite the fact that SP+ remains among the best we have in predictive accuracy and advanced stats.

Comment 07 Oct 2019

Not sure if Phil Steele is accounting for the entire season or just to date.  The computer models are to date, and Ohio State is above Clemson there and is first in the computer average.  It's just not enough to overcome Ohio State's 3/4 rankings in the human polls.

Comment 07 Oct 2019

I'd say the "see ya" was worse than singing a song based off of the legend of Woody Hayes.

I'd also say that every fan base has a certain number of "those fans" who ruin it for everyone.  The larger the fanbase, the more of those you get.  They also tend to be the fair weather variety, so when your team is good they come out like roaches.  So if she doesn't understand that and lumps you into that category, that's on her.  Maybe she is the reason Ohio natives don't like other Ohio natives that choose to root for other teams besides OSU.

Comment 07 Oct 2019

I had that opinion until this weekend, but Auburn will probably lose a few more games and drop out, at which point I will remember that any post UFM Florida team is destined for failure and I should stop giving them credit for beating similarly overrated teams.

Comment 07 Oct 2019

Yeah, I used to just link that in years past, but he actually doesn't calculate everything correctly when teams are unranked (or ranked past 25th) in computer polls that rank past 25 (you are supposed to get 0 points and he would give negative points).  It doesn't make a difference in the top 10 but it always annoyed me.

He will probably start posting it this year once the initial CFP rankings come out and I'll compare my results to his and see if they match.  Or I could just get lazy and use his like I used to, but I'm a numbers guy and I like doing the calculations.

Comment 07 Oct 2019

I think part of the issue is that in past years, after 6 weeks the preseason contribution is completed phased out whereas this year he still has it as 40% of the weight.  That Ohio State started so low based on the preseason metrics and is within 0.1 points of #1 with 40% of the weight on those low preseason metrics, I think that really emphasizes how dominate this OSU team is compared to the rest of the field.

Comment 07 Oct 2019

I have OSU shirts, and I have black shirts, but none are both.

My official 2014-15 CFP Undisputed Championship shirt is black.

Comment 07 Oct 2019

I don't think they submit ballots that are scored and sorted like the AP/Coaches polls.  I think they end with consensus on each ranking one spot at a time after their deliberations are concluded.  So it's more of a "Ok, we're putting Alabama at 1, everyone say yay or nay" and if there is agreement, they move to number 2, etc.

Comment 07 Oct 2019

Based on the total number of teams, if we had a field in Football with the same percentage of teams included in the playoff as Basketball, we'd have 24 teams in the playoff.  Top 8 get a bye, higher seed gets home field advantage.

Hypothetically, you could give an automatic berth to all conference champions including G5 conferences.  Then the remaining 14 teams based on resume.  At that point, nobody could realistically complain about being left out.

Only problem with that is needing 5 weeks to play it, in an already crowded schedule in December/January.  If they got rid of the majority of the bowls (or repurposed them for the quarterfinal and up matches on a rotational basis) then it would work.

Comment 07 Oct 2019

Yep, that's one of my sources of data for the computer rankings (I pull Billingsley, Massey and Sagarin rankings from it).  It's amazing to me how low our standard deviation is for all the various rankings and yet it seems only the human polls disagree.  Every other team has a much higher variance which just strengthens our argument for being the top team.  I've been following these kinds of polls for years and I've never seen a team with such a consensus this early in a season.  I'm loving it.

Comment 07 Oct 2019

Colley has Auburn at 2 still -- a lot of it is strength of schedule.  Florida's schedule has been terrible other than the Auburn win.  But even Massey and Sagarin have Auburn above Florida still.  This will correct itself as more games are played and we'll see the SEC teams sort themselves out.  Also, hopefully next week the A&H and Wolfe rankings will be released and then outliers like Colley with Auburn at 2 will be removed. 

Colley has a mathematically correct formula for ranking but it needs a "fully connected graph" to work well, so when you only have 5-6 games per team and most of your conference schedule hasn't played yet, then it's not going to be super accurate.  Especially since margin of victory isn't factored in.  They also don't factor in who you beat versus who you lost to (other than in strength of schedule calculations).  So Auburn losing to Florida is no different for their ranking than had they beat Florida and lost to Oregon, other than that Oregon and Florida would then be ranked differently.  But if you have ever been curious, he explains the method here