My "Math" poll

[Buck 'em up]

I got a better math formula:

Well played, Script. Well played.

 

[buckiprof]

That's kind of the point. It takes away the voter bias. It is straightforward and ranks team purely on who they beat. Not really any different than any current computer polls except there is a SOS aspect to it.

I like what you have done here...fun with numbers .

However, you have removed voter bias and replaced it with a different type of bias, BCS conference versus non-BCS conference. My years may be wrong on this, but in 2004 a win over Louisville was only worth 3 points since they were in Conference USA but in 2005 a win over Louisville would have been worth 5 points because they had joined the Big East.

I am not saying the bias between BCS conference versus non-BCS conferences is right or wrong, simply pointing out that there is a bias involved.

And I am not mentioning that as a negative, just that there is still a bias involved. A bias that was created (the BCS conferences versus the non-BCS conferences) by humans, those same damn creatures that have that voter bias.

 

[MililaniBuckeye]

So, you really think Stanford, Kansas State, Michigan, Nevada, Northwestern, UCLA, and NC State would beat us on the field? Maybe you chould change the "B" to "F" in your user name...

 

[Zander42]

I think this system is fine for playoff seeding, especially if you're talking about 8 teams or more. I think by the end of the season the ranking will sort themselves out enough to establish a top 8 or 16 teams. This is how it's used in the OHSAA.

I don't like the system for what we have now, where basically it's a one round 'playoff' for 1 and 2. I'd rather have a human component for that.

 

[GeorgiaBuck2]

What was the point of this poll anyway?

 

[FCollinsBuckeye]

The only 'point' is, it's Buck 'em up's own version of a computer ranking. I think it's interesting and look forward to following it.

What was the final rankings after the bowls last year, Buck 'em up?

 

[3074326]

Seems like a pretty unbiased and simple poll system. I like it, and I think it'll be interesting to see how it ends up looking at the end of the year. I suspect it'll be pretty accurate.

Don't think it's a good week-to-week poll, but I don't think that's what it's meant to be.

 

[Mike80]

It's an interesting concept, however I don't think the sample size is nearly big enough until there are ~6 games played by each team.

 

[Coqui]

So, you really think Stanford, Kansas State, Michigan, Nevada, Northwestern, UCLA, and NC State would beat us on the field? Maybe you chould change the "B" to "F" in your user name...

I don't think that's the point. We've only had one game featuring a team a winning record with a combined record of 4-9. Based off of our resume so far, there's no reason, with how he's setting up this poll, that we should be in the top 10. As time progresses, we should move up in the poll.

 

[Wingate1217]

So, you really think Stanford, Kansas State, Michigan, Nevada, Northwestern, UCLA, and NC State would beat us on the field? Maybe you chould change the "B" to "F" in your user name...

As the season wears on and the secondary points kick in and the Bucks keep winning we will see those teams listed above in our rear view mirror....
The system rewards you for:
1) Playing a tough schedule (assuming that you win some of the tough games)
2) Winning the majority of your contests..

The results at present are a bit skewed but as league play starts the pretenders will be left at the side of the road..

 

[MililaniBuckeye]

I don't think that's the point. We've only had one game featuring a team a winning record with a combined record of 4-9. Based off of our resume so far, there's no reason, with how he's setting up this poll, that we should be in the top 10. As time progresses, we should move up in the poll.

While I'm an advocate of no polls until mid-season to weed out obvious errors in rankings, there are some teams that obviously should be at/near the top regardless of who they may play early.

 

[Coqui]

While I'm an advocate of no polls until mid-season to weed out obvious errors in rankings, there are some teams that obviously should be at/near the top regardless of who they may play early.

But then you add either bias, or namesake into the equation. This puts everyone on an equal playing field.

So now that Boise St. is done playing it's premier games, their ranking will drop whereas our schedule is only going to get tougher so ours should rise.

 

[fourteenandoh]

It seems to me that this method is pretty useless early in the season, but will gain accuracy as there are more observations. Why don't you back test it with 2009 results and 2008 results then compare your results to the actual final rankings?

 

[Muck]

It seems to me that this method is pretty useless early in the season, but will gain accuracy as there are more observations. Why don't you back test it with 2009 results and 2008 results then compare your results to the actual final rankings?

psst...

It does work for college. If you look at who some of these teams have beaten, it is a pretty accurate representation of so far this season as opposed to a coaches intern or media members opinion. My top three at the end of the season last year (pre-bowls) were Alabama, Florida, Texas.

 

[Buck 'em up]

Thanks for all the input guys. I worked (and still do occasionally) for a newspaper in NE Ohio and I started getting into the Computer points based system that the OHSAA uses for HS football and just wondered how it would work for college. As many have noticed things are skewed a bit based on early season games. UCLA's wins over Texas and Houston are huge right now, but when UCLA starts losing those wins will lose relevance. And the SEC is a bit skewed too, since they are one of very few conferences that has had a few conference games already. It does work itself out as the year goes on. Last year my top 3 (pre-bowls) was Alabama, Florida, Texas. OSU was 6th for the record. I just think it is interesting how things worked out last year.