Rankings

Juan Tabo

11-man fan
New guy needs some help. Been looking at Milford, zephyr and Blum. All played Oakwood with similar results (within 5 points variance). It looks like Blum and zephyr both played Granberry with similar results. Puzzled by the ranking range of the three: Milford 29. Zephyr 15. Blum 53. Thoughts???
 
Also, how are some teams who are 3-0 ranked well below other teams who are 0-3? Mt. Calm is 3-0 and has only moved up 2 spots in the ranking since the beginning of the season....makes no sense?
 
It's not all about the W/L record, in fact, it has very little to do with the W/L record. It's about how you do against the spread.

Example:
Team A is ranked #25 with a rating of 150.
Team B is ranked #40 with a rating of 130.
Team A should win by 20, right?
But, Team A wins by 10. The get the W but lose rating points because they failed to cover the spread.

Another example.
Team A is ranked #25 with a rating of 150.
Team C is ranked #199 with a rating of 25.
Team A should win by 45+.
Team A wins by 45+.
Team A gets the W, but have they really proved anything? Not really, they did what was expected. So, there won't be much of a change there.
 
Thx hornkeeper. Appears in six man the variables are many. Coach changes. Personnel - moves in/out, injuries, quitting, etc. one more question. Is there subjectivity used in "spread" determination or is is simply "modeled" statisticly??
 
You're right, It's 100% subjective. No way to accurately predict the outcome of games, because no one knows all the teams. We were picked to win by 45 this past week over Living Stones, but they had all their kids back at full strength for the first time and played way better than I was led to believe. They have three kids #2, #10, #11 who can flat out play and #12 and #13 are no slouches either. Either we were vastly over rated, or they are very underrated. We stole a win, but in no way, shape or form, were we a 45 point favorite. It's fun to talk about, but remember, these ranking are done by guys who have seen tons of football; but no matter how much they watch, they can never see all the games. I appreciate how hard they work on them, but try not to value my teams fate by the spread.
 
The subjectivity is where a team is ranked to begin the year. There is not a way in this world for anyone to know for sure where any team should be ranked. There will always be 3 or 4 teams way under ranked and 3 or 4 teams way over ranked. Happens every year. The beauty of his "toy" is that by the end of the year your team will be ranked close to where it should be. Even the under ranked teams. For you new-comers, I would suggest that you have patience, watch and see.
 
hornkeeper12":2jtjnfol said:
The spread is directly taken from the rating. There is no "subjectivity" involved.

That's only partially right, hornkeeper. Okay, I'll give you 75% correct. The preseason rankings are completely subjective. The rest of the season is a weekly adjustment to the original subjectivity.

I'll grant you that the top 10-12 is a fairly accurate assumption of who is really good. And the bottom 20 or so is probably equally as accurate. But the middle - anything between the top 15 and bottom 15 - is almost impossible to get correct from the get go.

I'm no rankings guru, nor have I spent hours perusing and studying movements in the rankings from week to week, but I would venture a substantial wager of CFS at the little cafe in Strawn that the most dynamic movements up or down week in the rankings week in and week out occur among the 40 or so teams in the middle of the rankings pack.
 
rainjacktx":wd3eijwv said:
hornkeeper12":wd3eijwv said:
The spread is directly taken from the rating. There is no "subjectivity" involved.

That's only partially right, hornkeeper. Okay, I'll give you 75% correct. The preseason rankings are completely subjective. The rest of the season is a weekly adjustment to the original subjectivity.

I'll grant you that the top 10-12 is a fairly accurate assumption of who is really good. And the bottom 20 or so is probably equally as accurate. But the middle - anything between the top 15 and bottom 15 - is almost impossible to get correct from the get go.

I'm no rankings guru, nor have I spent hours perusing and studying movements in the rankings from week to week, but I would venture a substantial wager of CFS at the little cafe in Strawn that the most dynamic movements up or down week in the rankings week in and week out occur among the 40 or so teams in the middle of the rankings pack.
I'll give you 100% on that. A more accurate statement would have been there is no subjectivity involved after the initial pre-season rankings. However, pre-season rankings are entirely subjective based on what little information we have to go on, or basically, an uneducated guess.
 
hornkeeper12":3akow3zt said:
It's not all about the W/L record, in fact, it has very little to do with the W/L record. It's about how you do against the spread.

Example:
Team A is ranked #25 with a rating of 150.
Team B is ranked #40 with a rating of 130.
Team A should win by 20, right?
But, Team A wins by 10. The get the W but lose rating points because they failed to cover the spread.

Another example.
Team A is ranked #25 with a rating of 150.
Team C is ranked #199 with a rating of 25.
Team A should win by 45+.
Team A wins by 45+.
Team A gets the W, but have they really proved anything? Not really, they did what was expected. So, there won't be much of a change there.

Oh come on. We all know it involves a drunk monkey, blindfold and a dart board.
 
By definition, the beginning rankings has bias. With very little information available, that's what you get. After that, it is 100% non-subjective.

Rankings are calculated first on how TEAM A does vs. TEAM B vs. what was expected (the spread). The system then runs through the entire system, re-rating every game based on the new results (to a lesser extent, many, many times). Remember that the ranking changes literally hundreds of times as it re-evaluates the rankings.

I will post more in a bit...
 
51eleven":2okgxcy9 said:
Oh come on. We all know it involves a drunk monkey, blindfold and a dart board.

"Pay no attention to that man behind the curtain."
-- The Wizard of Oz

Just kidding. As Granger has explained to me (before I completely lose the entire train of thought ... something about derivative of the square root of the Angolian prime rate multiplied by the spot market cost of Chilean sea bass really threw me off), initial rankings -- especially with private schools, are weighted towards their previous year's rankings. And then you have new schools -- how do you rank East Bug Tussle Christian when they're playing their first game. Sure, you guess low. But "guess" may be the operative word (and it's my word, not Granger's).

I still say, with a couple exceptions, we'll know where the private schools stand in about two weeks (for example, TAPPS D2 is an entirely wide-open race this year. As many as 8 or 10 teams in contention.)
 
Apologies to the Guru. I forget that there could be people who took my sarcasam seriously, although I think he knows it was not directed at his program. Probably 98% of us who post on here could not develop a program any where close to as accurate as his rankings get. Even given 20 years to tweak it. Or for big bucks, as opposed to for the fun of it.
 
No offense taken.... You guys that have been around here know that at least once a year we have to go through this. That is quite alright. I had hoped to type the response in one fail swoop, but class got in the way. I also needed to update some code in the new software I wrote to make this all work....

Here's the tool I use to look at teams. Try it out. It will get updated all season.

http://sixmanfootball.com/why_test.php

as for the making the big bucks.... I am without a job as I hit school for another degree at the ripe old age of 48. So, it you are so inclined, please hit those things that produce money for me often....LOL

Still cannot believe this is my 21st season doing this.

More answers later as needed, but remember, if you have an issue, develop your case first
 
granger":3dcu513m said:
No offense taken.... You guys that have been around here know that at least once a year we have to go through this. That is quite alright. I had hoped to type the response in one fail swoop, but class got in the way. I also needed to update some code in the new software I wrote to make this all work....

Here's the tool I use to look at teams. Try it out. It will get updated all season.

http://sixmanfootball.com/why_test.php

as for the making the big bucks.... I am without a job as I hit school for another degree at the ripe old age of 48. So, it you are so inclined, please hit those things that produce money for me often....LOL

Still cannot believe this is my 21st season doing this.

More answers later as needed, but remember, if you have an issue, develop your case first

You didn't say "okay" once. Stellar job, boss.
 
Back
Top