calpreps power rankings 10-29
- Thread starter consumerman
- Start date
consumerman
New member
interesting note is moeller is ranked appx 100 spots ahead of MV who beat Moeller in Ohio by 4
???
???
Here's another head scratcher.
At the time of the game MNW and SLC were #1 and #3 in this ranking system.
Since then, SLC has had one close game and the rest blow outs. MNW has not been challenged.
We know they ave been dropped WAAAYYY down.
Interestingly, Oregon has two teams ranked #12 and #41. #12 has zero OOS opponents and #41 has one. They beat the #59 rated team in Washington, wow...
Lake Oswego #12
Port Jesuit #41
MNW #44
SLC #100
At the time of the game MNW and SLC were #1 and #3 in this ranking system.
Since then, SLC has had one close game and the rest blow outs. MNW has not been challenged.
We know they ave been dropped WAAAYYY down.
Interestingly, Oregon has two teams ranked #12 and #41. #12 has zero OOS opponents and #41 has one. They beat the #59 rated team in Washington, wow...
Lake Oswego #12
Port Jesuit #41
MNW #44
SLC #100
I know that early in the season Calpreps uses some results from the previous year. That explains the rankings early in the year.
The Texas rankings are really goofy. Katy is looking terrific and they are probably the #1 Texas team. No way Stephenville, Gilmer, Arlington Bowie and Trinity should be ranked higher than Katy.
"That explains the rankings early in the year."
It's not early in the year. A large number of prep football teams are DONE. Most are either done or within one or two weeks of being done. In fact, even among teams that play in a state tournament, many have not more than three or four games left. I think in the extreme case of the relatively minute # of teams like those in Texas that will play 16 games, they are at least half way done. And, considering recent trends, the ratings seem to be getting even more out of whack as we go along.
It's not early in the year. A large number of prep football teams are DONE. Most are either done or within one or two weeks of being done. In fact, even among teams that play in a state tournament, many have not more than three or four games left. I think in the extreme case of the relatively minute # of teams like those in Texas that will play 16 games, they are at least half way done. And, considering recent trends, the ratings seem to be getting even more out of whack as we go along.
consumerman
New member
The SLC - MNW game was early in the year
that was the reference
maybe try rereading (or reading and understanding for the first time) the post
and reference and not be so quick to argue
ams eds fan
New member
I think I see Marion Local in the top 25.
UncleBaldy
New member
I've defended the Calpreps (Freeman) rankings in the past. I still think they do a good job overall. However, their output is very, very wacky at the moment in my opinion. We'll see if that changes.
Dad: There is nothing on the calpreps site that warns us to expect less accurate resulsts this year than others. The same method used to derive the present ratings has been used in years past, with some minor exceptions the calpreps site calls improvements. Because of the interrelated nature of the data and programming the computer system uses, you can't question the validity of one seasons rankings while endorsing those from other seasons. That's not to say any one team's rating is neccesarily wrong, just to say that the ratings set as a whole is entirely unreliable.
Except . . . . . you can't compare final rankings from previous years to rankings 1/2 - 3/4 through the season this year. All of the top teams will play 14, 15, or 16 games so from that perspective the season far from over.
Still waiting to see some reasons as to why the final '06 poll was better than the the human polls.
I said the computer models were as good as the human polls and in some ways better. I've given at least one example.
We've also discussed the report card on the polls.
I've mentioned some ways that the human polls are better.
If you read the threads you'll see that the answer to your question has already been provided.
consumerman
New member
pied
daditis MO is to get the END that he wants, then justify the means
if he wants slc to come out on top and a computer model in 2006 is the best evidence then computer programs are great and take out human bias
if he wants slc to look good in 2007 and the computer models put slc at #100,
then the human polls are much better
if he wants slc to look better any time any year, he will find the poll type that does that, then try to rationalize it
UncleBaldy
New member
Trust me on this one. Marion Local does not belong in the top 25.
Consumerman, you are totally looney. Are you making this stuff up?
You know that I've always used ALL of the polls. How many times do I have to say it?
Even now, most folks agree that the Freeman model looks goofy at the present time. However, I'd still include it. SLC has been in the National Top 10 all year. I accept that. http://calpreps.com/2007/polls_view.htm
I agree. However, I still say the computer model does a very good job overall. It's remarkable that a computer can rank 15,000+ teams and produce the output that it does. The model isn't perfect (as you note) but it's pretty good.
We could find flaws in every ranking. We've all seen teams that ended up being a bust but the human rankings had them highly ranked at first.
There's really no basis for saying their system does a good job of ranking teams on a national scale.
The problem isn't with their rating method, which is probably reasonable, it's the concept of rating that many teams with that little data. Just isn't possible without getting a bunch of weird results.
Your point is well taken and this year is a good example of what you say.
However, every other year the results have been pretty good. My guess is that the model's output will be pretty good this year also as more games are played. We'll see.
Having said that, I always use ALL of the polls (both human and computer). I think that gives the best result.
consumerman
New member
Did you use ALL of the polls when you felt it necessary to start a thread on a compilation of JUST human polls?
I must have missed it.
Perhaps you are referring to your Portland Jesuit comment:
I noted this afterwards:
So, if they nailed it are you making the case that they were indeed the second best team in the country?