CCR

Say it all here
Forum rules
NOTICE: Please be sure to check the CFP Message Board Rules and Regulations and the Read Me page before posting.
User avatar
Eric
Moderator
Moderator
Posts: 10279
Joined: Thu Dec 29, 2005 1:51 am

CCR

Postby Eric » Thu Nov 01, 2007 12:15 pm

Has anybody else noticed that the CCR struggles picking the FCS upset? I've noticed that when evenly matched teams like Delaware/Navy and North Dakota State/Minnesota play eachother, the CCR usually takes the FBS team by double-digits. I think I've had more success picking FCS/FBS games than the CCR (I took Delaware over Navy, New Hampshire over Marshall, and ND State over CMU).

My only question is why are the games predicted to be such blowouts? :?
Running bowl/MSU/OSU record '05-present: 11-32

User avatar
donovan
Athletic Director
Athletic Director
Posts: 8634
Joined: Wed Oct 04, 2006 6:41 am
Location: Pacific Northwest

Re: CCR

Postby donovan » Thu Nov 01, 2007 1:06 pm

I certainly have no idea. I think most pollsters that use the computer have the advantage of consistency. The formulas used are always used and the chips fall where they. So the better the database...or maybe the more accurate in finding the markers of prediction...the better the poll. CCR has a very good record. I think excellent is appropriate.

So the question of the blowout has to lie in the fact that the algorithms used dictate the blowout. One of the differences between the CCR poll....I am guessing because it is all a mystery for very good reasons....is they are not taking bets. They have no point spread to worry about. If you were using a computer to make a betting line, you would not want a large point spread if the money is balanced....If you worry only about wins and losses..the point spread makes not difference. Didn't this site last year have some item where you picked the spread..or score and there was a ranking on that...could be wrong.

Do I think CCR is prejudiced for or against a school or conference...absolutely not...that is one of the strong points...it is not.

Though I think the computer can track a lot of statistics that is difficult without...I think Mr. Billybud had it right..look at that stuff and let the human computer make a decision....

Because I no longer carry the CCR torch on the predictions...I made mine this weekend trying to do that....Everyone gets influenced by something....especially when you do not see all the games....So you pick your best source....I think early on watching USC, you could sense..they were not playing that well as a team.....People that watched Oregon early on knew they were better than the polls.....

On to point...the blowouts have to be a consequence of the programming......unless you think there is skulduggery.... :twisted:
Statistics are the Morphine of College Football

User avatar
openSkies
Administrator
Administrator
Posts: 1288
Joined: Wed Dec 28, 2005 4:10 pm
Location: Boston, MA, USA
Contact:

Re: CCR

Postby openSkies » Thu Nov 01, 2007 1:18 pm

Let's take something into consideration...

CCR picks the games based on the difference between the power ratings of the teams. We don't cover FCS, so, therefore all FCS teams have a power rating of 0.

If we have nothing to compare the FBS and FCS teams against each other, how would the FCS team ever be able to be picked to win?
Image

User avatar
Eric
Moderator
Moderator
Posts: 10279
Joined: Thu Dec 29, 2005 1:51 am

Re: CCR

Postby Eric » Thu Nov 01, 2007 3:54 pm

Maybe all FCS teams shouldn't have a power ranking of 0 :idea:

Look, give me Appalachian State, Northern Iowa, North Dakota State, UMass, or Montana over UNLV, Utah State, Kent State, UL Lafayette, or Duke anyday.

If the CCR wants to improve as a poll, I think it has to give some FCS teams some more love.
Running bowl/MSU/OSU record '05-present: 11-32

User avatar
Eric
Moderator
Moderator
Posts: 10279
Joined: Thu Dec 29, 2005 1:51 am

Re: CCR

Postby Eric » Thu Nov 01, 2007 3:58 pm

openSkies wrote:If we have nothing to compare the FBS and FCS teams against each other, how would the FCS team ever be able to be picked to win?



You're kind of making my point for me. It's a flaw in the rankings system. We know that FCS teams lose a lot to the FBS teams, but we also know FBS teams don't win 100% of the time. I realize the FCS is 10-63 against the next tier of competition, but in many of those games, the better team (FCS, that is) was picked to lose by close to 20 points.

I'm sort of nit-picking because it is an awesome poll, but I'm just inquiring what that's all about.
Running bowl/MSU/OSU record '05-present: 11-32

User avatar
openSkies
Administrator
Administrator
Posts: 1288
Joined: Wed Dec 28, 2005 4:10 pm
Location: Boston, MA, USA
Contact:

Re: CCR

Postby openSkies » Thu Nov 01, 2007 4:44 pm

I take it back. They don't have a power rating of 0. I've asked CFP Admin to come explain it to y'all.
Image

User avatar
CFP Admin
Administrator
Administrator
Posts: 163
Joined: Wed Dec 28, 2005 9:08 pm
Contact:

Re: CCR

Postby CFP Admin » Thu Nov 01, 2007 4:52 pm

Used to be 0, because there were so few games you could not get an accurate reading on FCS vs. FBS. With the addition of the 12th game, the number of FBS vs. FCS went up dramatically, providing a reasonable opportunity to create a mathematical evaluation. Still, however, too few games by each team to weigh those schools independently so all FCS schools are weighted by average result of all FBS vs. FCS games. This means if FIU were to play North Dakota State tomorrow, NDSU would be favored on a neutral field but there would still be very few outright upsets picked. And that's how it should be when you consider that FBS teams have won 90% of the games played against FCS schools since 2003.

Given the disadvantage in scholarships, general across-the-board talent level, and other factors, any win by a FCS school over a FBS school has to be considered an upset. While the computer does pretty dang well at picking upsets between FBS teams, consider the number of games it can sample vs. what's at its disposal for cross-subdivision play.

Hope that explains it.

User avatar
Eric
Moderator
Moderator
Posts: 10279
Joined: Thu Dec 29, 2005 1:51 am

Re: CCR

Postby Eric » Thu Nov 01, 2007 6:52 pm

Is there any way you could fix it though?

Take Montana for instance. They haven't played an FBS team yet they're one of the best FCS out there. The computer would take Florida International on the neutral field, wouldn't it?
Running bowl/MSU/OSU record '05-present: 11-32

User avatar
Spence
Administrator
Administrator
Posts: 20243
Joined: Wed Dec 28, 2005 11:52 pm
Location: Chillicothe, Ohio (Ohio's First Capital)
Contact:

Re: CCR

Postby Spence » Thu Nov 01, 2007 8:33 pm

I would think that adding 1-AA would be a crap load of extra work.
"History doesn't always repeat itself but it often rhymes." - Mark Twain

User avatar
donovan
Athletic Director
Athletic Director
Posts: 8634
Joined: Wed Oct 04, 2006 6:41 am
Location: Pacific Northwest

Re: CCR

Postby donovan » Thu Nov 01, 2007 9:34 pm

I see it a superfluous in the obtaining of the goal....which I think is to rank the Division I football teams in some order. There could be sub division teams that are better than the upper division...not sure of the relevance. There may be a college team better than a Professional team...especially arena or the soon to be forgotten AAFFLLAKGB league.....So the question is how do you count the stats into the picture with those games....you don't. The get no credit, they get no penalty other than a win or a loss....
Statistics are the Morphine of College Football

User avatar
openSkies
Administrator
Administrator
Posts: 1288
Joined: Wed Dec 28, 2005 4:10 pm
Location: Boston, MA, USA
Contact:

Re: CCR

Postby openSkies » Thu Nov 01, 2007 11:00 pm

Donovan, I think that pretty much sums it up.

Our job isn't to deal with I-AA (and yes, it would be way too much work). We deal with I-A. And we're a computer ranking first and foremost. Hence the website name.
Image

User avatar
Spence
Administrator
Administrator
Posts: 20243
Joined: Wed Dec 28, 2005 11:52 pm
Location: Chillicothe, Ohio (Ohio's First Capital)
Contact:

Re: CCR

Postby Spence » Thu Nov 01, 2007 11:46 pm

It would be my wish that the D1A schools decide that they shouldn't play the 1-AA schools. I see no good reason for a 1A school to ever play 1-AA. I agree, that if they do decide to play that the stats from the game shouldn't count. I would go one further and say that playing a 1-AA school shouldn't count towards bowl eligibility. If they are going to play a 12th game, then it should be an interdivisional game. If they don't want that, then dump the 12th game.
"History doesn't always repeat itself but it often rhymes." - Mark Twain

User avatar
CFP Admin
Administrator
Administrator
Posts: 163
Joined: Wed Dec 28, 2005 9:08 pm
Contact:

Re: CCR

Postby CFP Admin » Fri Nov 02, 2007 12:16 am

donovan wrote:I see it a superfluous in the obtaining of the goal....which I think is to rank the Division I football teams in some order.


Bingo! That's precisely why - when there weren't that many 1-A teams playing 1-AA teams - that all 1-AA teams counted at "zero" - it was to punish those teams for playing what we used to refer to as "the weak sisters". Mike Mitchell would include, in his weekly pre-game column, a section that focused on such games that was titled "Pick On Someone Your Size" or "Poisoned Cupcakes". The point being that the 1-A (FBS) team has nothing to gain and everything to lose by scheduling one of these schools. It's merely an easy way for the Athletic Directors to get an extra home game ($$$$) without having to reciprocate, with the idea that you are also "scheduling a win". And keep in mind, the FBS teams win 90% of those games, and most of the losses have come from the worst conferences in the FBS. In 2005, 1-A teams went 52-2 in these games.

User avatar
Eric
Moderator
Moderator
Posts: 10279
Joined: Thu Dec 29, 2005 1:51 am

Re: CCR

Postby Eric » Fri Nov 02, 2007 11:36 am

Well the $ is what the schools have to gain like you said.

I just think that if the FCS teams like Delaware, North Dakota State, and New Hampshire were given the proper credit, then the CCR would be correct in those instances.

Like I've said though, it's only 10 games out of 73 that they've won.
Running bowl/MSU/OSU record '05-present: 11-32

User avatar
Eric
Moderator
Moderator
Posts: 10279
Joined: Thu Dec 29, 2005 1:51 am

Re: CCR

Postby Eric » Fri Nov 02, 2007 11:37 am

openSkies wrote:Donovan, I think that pretty much sums it up.

Our job isn't to deal with I-AA (and yes, it would be way too much work). We deal with I-A. And we're a computer ranking first and foremost. Hence the website name.


Well, not many, but some people would be turned off by a computer poll that picks North Dakota State to lose by 20 when they win by 30. That was about a 50 point swing right there.
Running bowl/MSU/OSU record '05-present: 11-32


Return to “General Discussion”

Who is online

Users browsing this forum: No registered users and 16 guests