Started By
Message

re: CFB Advanced Metrics and Computer Polls, Week 12

Posted on 11/18/19 at 7:36 pm to
Posted by buckeye_vol
Member since Jul 2014
35239 posts
Posted on 11/18/19 at 7:36 pm to
quote:

If a team wins a conference then how are they undeserving? Who cares how many losses they have? They accomplished something that nobody else in the conference could do, win a championship.
Every team in the ACC Coastal division has at least 3 losses and 2 conference losses and the highest rated team in the Massey Composite is 33rd (Virginia) and there is a good chance that whoever wins the division has at least 4 losses given their remaining schedules.

So while Clemson will be heavy favorites against any Coastal opponent an upset of that magnitude or even bigger happens every year, and has happened at least twice this year (Wisconsin-Illinois; Georgia-South Carolina).

And those upsets weren’t especially “lucky” just the luck of an underdog’s better performance coinciding the the favorite’s poorer performance and a few random plays in their favor here and there (turnovers, missed FGs, etc.).

The fact that an 3 or 4 loss team, ranked worse than 30+ other teams, in a historically terrible conference, can get in based solely on a single upset won without any unusual luck outside their control is bad enough to scrap the automatic bid idea. But the possibility of some external luck in their favor like a Lawrence/Etienne injury, inexcusably one-sided officiating (Oklahoma/Oregon onside kick game, WVU-Pitt in 2007, etc) should eliminate any remaining semblance of deserving a bid.

I think if they had automatic bids, they should go back to the BCS requirement where it’s an automatic bid provided a team finishes within the top-12 or whatever.
Posted by gobuxgo5
Member since Nov 2012
10028 posts
Posted on 11/18/19 at 7:37 pm to
quote:

Wait.....what?


Basically, I look at LSU's recruiting rankings on average over 3 years. I look at who they beat and how those team's recruiting has ranked.

In my scenario, I gave LSU a 14% increase in total points because they have recruited well and have beaten well recruited teams this year.

This post was edited on 11/18/19 at 7:38 pm
Posted by Draconian Sanctions
Markey's bar
Member since Oct 2008
84860 posts
Posted on 11/18/19 at 7:38 pm to
quote:

Basically, I look at LSU's recruiting rankings on average over 3 years


And why would you do that?
Posted by xiv
Parody. #AdminsRule
Member since Feb 2004
39508 posts
Posted on 11/18/19 at 7:39 pm to
quote:

USC makes no sense though.
They have the #2 SOS in my formula in part because they’ve played 11 games as opposed to most other teams’ 10. They’re idle this week, so they’ll drop.
Posted by Draconian Sanctions
Markey's bar
Member since Oct 2008
84860 posts
Posted on 11/18/19 at 7:39 pm to
You can’t control for that?
Posted by BillyBobfan24_7
R.I.P. SGT Nelson
Member since May 2004
18065 posts
Posted on 11/18/19 at 7:39 pm to
quote:

Man yall are egging this shite on obnoxiously


Because they know we dont play them anytime soon. Easy to talk shite with nothing on the line.
Posted by gobuxgo5
Member since Nov 2012
10028 posts
Posted on 11/18/19 at 7:41 pm to
quote:

And why would you do that?



High School divisions in Ohio are based on # of boys in that school district. By and large it holds true that the bigger schools are much better than the smaller schools.

A college team who recruits nationally has a large # of men to pick from to be on their team, if you will

A team who doesn't recruit well has a small # of men to recruit from.

Having a better recruited team over 3-4 years, gives you a roster that would be liken to a higher division in high school. (bigger, faster, stronger, deeper)

It establishes logic that certain teams are more talented than others. Beyond that, you get additional kickers by beating talented teams.

If A D1 school plays only D2 schools in High School, they'd have less computer points than playing D1.

I give LSU more credit for beating well recruited teams like Alabama, Florida, Auburn, Texas A&M, etc.
This post was edited on 11/18/19 at 7:43 pm
Posted by BillyBobfan24_7
R.I.P. SGT Nelson
Member since May 2004
18065 posts
Posted on 11/18/19 at 7:42 pm to
quote:

When Michigan beats OSU this year the meltdown will be glorious.


They are going to cry. That was a low blow.
Posted by Draconian Sanctions
Markey's bar
Member since Oct 2008
84860 posts
Posted on 11/18/19 at 7:42 pm to
Considering how fricky the recruiting rankings are, that’s one of the more absurd things I’ve heard about how to rank teams
Posted by gobuxgo5
Member since Nov 2012
10028 posts
Posted on 11/18/19 at 7:43 pm to
I'm not ranking them at all. I am using typically a 2-3% bump for about 12 schools total. LSU has a crazy bump because of their wins so far.

If you look at the original top 25 I used and "tweaked" there wasn't a whole lot of movement.

But a lot of teams are just now getting to the meat of their schedules and could see more bumps.
This post was edited on 11/18/19 at 7:46 pm
Posted by xiv
Parody. #AdminsRule
Member since Feb 2004
39508 posts
Posted on 11/18/19 at 7:47 pm to
quote:

You can’t control for that?
Why would I? Playing 11 games is harder than playing 10.

The best part of it is that by the end of 12/7, 13 games will have been harder than 12, putting an organic emphasis on playing for and winning a conference championship.
Posted by OldmanBeasley
Charlotte
Member since Jun 2014
9696 posts
Posted on 11/18/19 at 7:48 pm to
quote:

Not important, and i don’t spend much time thinking of it in those terms. Any opinion I can give you can hear screamed at you by Stephen A or Skip or whatever, and the next day I might flip flop on it.

I appreciate the no screaming haha, but do you ever change or tweak your system? Penn state and Cincinnati seem way too high ( I understand why they are ranked so high in your model)
Posted by Draconian Sanctions
Markey's bar
Member since Oct 2008
84860 posts
Posted on 11/18/19 at 7:48 pm to
okay man. Silly af
Posted by KosmoCramer
Member since Dec 2007
76519 posts
Posted on 11/18/19 at 7:57 pm to
quote:


I appreciate the no screaming haha, but do you ever change or tweak your system? Penn state and Cincinnati seem way too high ( I understand why they are ranked so high in your model


As he stated earlier, once you go down the rabbit hole of "perfecting" the computer poll, you've done so many adjustments that a "simple" calculation is just as good if not better in it's simplicity.
Posted by xiv
Parody. #AdminsRule
Member since Feb 2004
39508 posts
Posted on 11/18/19 at 8:02 pm to
I’ve never tweaked this formula; it’s based on the Golden Ratio (long story). I have several other systems of my own for shits and giggles, and I tweak them for said shits and giggles, but the one I submit to Massey has been consistent.
quote:

Cincinnati
#7 in my rankings, #17 in Massey composite. Several ranking have them at 6, and they’re as high as 4. Go figure. AAC is having its best year ever, and I think their top teams are a bit underrated.
quote:

Penn State
#4 for me, #5 in the Massey Composite, as high as 3. They’ll either win out and be in the top 4 where they belong or they won’t, and they’ll drop.
Posted by KosmoCramer
Member since Dec 2007
76519 posts
Posted on 11/18/19 at 8:06 pm to
quote:

I’ve never tweaked this formula; it’s based on the Golden Ratio


That's what I love about some of these polls. They're based on mathematical proofs/constants.

Thanks for what you do.
Posted by gobuxgo5
Member since Nov 2012
10028 posts
Posted on 11/18/19 at 8:07 pm to
quote:

okay man. Silly af


I don't think it's silly. If high school school district's population is a good indicator of a division, than the overall quality of recruiting of each college team determines a nice baseline for a division. After the Top 20 recruiting classes, you see a significant drop in program talent, like you would see when you drop in division for high school.

All you need to do is run aggregate rankings over the 3-4 year periods so you can get a snapshot of which programs are top 20, top 40, top 60, etc.

I'm just saying you can reward teams for beating those talented teams with perennially talented rosters.

For instance, beating a 6-4 Texas team should have some reward even though beating a 9-1 Cincinnati team is deemed better by the committee. The Cincy win is now "better" but the Texas win would have been much "harder" so I think you get a bump for that. It's harder because nearly every Texas player would start over the Cincy roster.
This post was edited on 11/18/19 at 8:12 pm
Posted by xiv
Parody. #AdminsRule
Member since Feb 2004
39508 posts
Posted on 11/18/19 at 8:09 pm to
quote:

Thanks for what you do
Felt pretty damn good to read that
Posted by OldmanBeasley
Charlotte
Member since Jun 2014
9696 posts
Posted on 11/18/19 at 8:09 pm to
You definitely have done your research. What do you do for a living if you don’t mind me asking
Posted by gobuxgo5
Member since Nov 2012
10028 posts
Posted on 11/18/19 at 8:10 pm to
quote:

Felt pretty damn good to read that



Yeah, xiv posting here has been a treat. His posts have been the truth.

This post was edited on 11/18/19 at 8:11 pm
Jump to page
Page First 17 18 19 20 21 ... 26
Jump to page
first pageprev pagePage 19 of 26Next pagelast page

Back to top
logoFollow TigerDroppings for LSU Football News
Follow us on Twitter, Facebook and Instagram to get the latest updates on LSU Football and Recruiting.

FacebookTwitterInstagram