WK 4 Computer vomit

[rmilia1]

Yep, and that's why the Colley looks like this...

1. LSU
2. Northwestern
3. Florida
4. Utah
5. UCLA
6. Texas A&M
7. Oklahoma
8. Indiana
9. Ohio St
10. TCU
11. Michigan St
12. Temple
13. NC State
14. Mississippi
15. Notre Dame
16. Georgia
17. Florida St
18. West Virginia
19. Iowa
20. California
21. Toledo
22. Navy
23. Memphis
24. Kentucky
25. Alabama

Now that's an ok top 25

 

[ellupo]

Ohhh, the super negative because your team isn't perfect fan.

Nice.

Not super negative. Perhaps you should read and comprehend before you reply.

 

[4down20]

Not super negative. Perhaps you should read and comprehend before you reply.

You called anything that ranks Ohio St #1 bullshit.

What did I fail to comprehend? Seems a bit extreme to me.

 

[nddulac]

Not true. The colley matrix does no such thing. Everybody starts the season even.

Just because they start the season with equal ratings does mean there is no initial bias. Any assumption that allows one to give the teams initial ranking constitutes an initial bias.

 

[rmilia1]

Nothing Ohio State has done THIS year merits a number 1 ranking. That's not being negative, it's being realistic

 

[4down20]

The fact Ohio St has been winning while seemingly playing poorly and is likely to only improve for the future just as they did last year is more than enough reason to rank them #1.

They have a schedule that is very friendly and gives them plenty of time to get things together.

 

[rmilia1]

The fact Ohio St has been winning while seemingly playing poorly and is likely to only improve for the future just as they did last year is more than enough reason to rank them #1.

They have a schedule that is very friendly and gives them plenty of time to get things together.

Their future schedule has nothing to do with where they should be ranked right now. If they win those games they'll be rewarded when they win them. You seem to want to reward teams for wins they don't have yet and then if they lose you seem to not want to penalize them for losing because you still consider them to be better. Very strange

 

[podsox]

good thing northwestern in behind Stanford

 

[4down20]

Their future schedule has nothing to do with where they should be ranked right now. If they win those games they'll be rewarded when they win them. You seem to want to reward teans for wins they don't have yet and then if they lose you seem to not want to penalize them for losing because you still consider them to be better. Very strange

In my computer rankings I predicted future games and gave them credit for 50% of the score.

It made for better and more accurate results overall despite the 25% of the time it was wrong.

 

[TROJAN-MAN]

Lol. That is me. And the posting there...that is mostly for my entertainment. People there are would so tight! This is a totally different atmosphere here. People are more down to earth and knowledgeable. BR is all about getting people all up in arms and spouting out a bunch of BS. Lol. If that makes sense. Who are you there?

I don't post there, I go to bleachers for the entertainment value also.

 

[ellupo]

You called anything that ranks Ohio St #1 bullshit.

What did I fail to comprehend? Seems a bit extreme to me.

Do you think they have played liked the #1 team in the country this year?

 

[nddulac]

In my computer rankings I predicted future games and gave them credit for 50% of the score. It made for better and more accurate results overall despite the 25% of the time it was wrong.

That's really a crazy methodology. It sounds to me like a mathematically guaranteed version of "confirmation bias." At this point, that means that half* of the data you are using in your fit is based on speculation, rather than established facts.

*half because you have 1/3 of the season complete. 2/3 of the season to go, of which you are using 50% weighting. So you have 1/3 + 0.5(1/3) = 2/3. And the 1/3 that is actual data is half of that total 2/3.

 

[4down20]

Do you think they have played liked the #1 team in the country this year?

Power ranking wise? No.

 

[4down20]

That's really a crazy methodology. It sounds to me like a mathematically guaranteed version of "confirmation bias." At this point, that means that half* of the data you are using in your fit is based on speculation, rather than established facts.

*half because you have 1/3 of the season complete. 2/3 of the season to go, of which you are using 50% weighting. So you have 1/3 + 0.5(1/3) = 2/3. And the 1/3 that is actual data is half of that total 2/3.

Confirmation bias by taking the existing stats and extending them out into the future?

Do you know what confirmation bias is?