Jump to content

College Football Vib Rankings Released

Tara Clover

By Tara Clover
in The Dog Pound

Recommended Posts

Druez Newbie

Druez

Posted
Posted
Just another example of why computer rankings are bogus. When you have a team that is 6 and 3 A&M ahead of a team that is 8 and 1 West Virginia, it simply baffles me. West Virginia's only loss came to VT a rival on the road. A&M gets blown out by UTAH and loses to Baylor? They played a good game vs Oklahoma in a loss. Do you lose points when you lose to a terrible team? I don't get these computer formulas. I love computers, but computers just don't get football. Wisonsin at 9 and 0 is number 12? I would rather see a Vibdog ranking where Scott ranks the college teams. I know Scott knows college football and I'm pretty sure that his rankings would be quite different then his computer one. So in closing, I say why bother with computer polls when they aren't very accurate.
Link to comment
Share on other sites
Scott Vibert Explorer

Scott Vibert

Posted
Posted
[QUOTE=Druez]Just another example of why computer rankings are bogus. When you have a team that is 6 and 3 A&M ahead of a team that is 8 and 1 West Virginia, it simply baffles me. West Virginia's only loss came to VT a rival on the road. A&M gets blown out by UTAH and loses to Baylor? They played a good game vs Oklahoma in a loss. Do you lose points when you lose to a terrible team? I don't get these computer formulas. I love computers, but computers just don't get football. Wisonsin at 9 and 0 is number 12? I would rather see a Vibdog ranking where Scott ranks the college teams. I know Scott knows college football and I'm pretty sure that his rankings would be quite different then his computer one. So in closing, I say why bother with computer polls when they aren't very accurate.[/QUOTE] The reason to bother with them is to try and get the human element and the biases inherent in human rankings out of it (sometimes you see teams more often, rooting interests, etc). The goal of the computer rankings is to provide a completely objective and unbiased weighting. This is of course an impossible goal; as the humans writing the model have their own biases (how much to weight SOS, etc.); and sometimes human judgement can be better then mathematical judgement. That said my equation is definitely still in development (I just won't be changing it until after the season in the released version, for consistency's sake). I too would agree that Wisconsin is too low, that said while they are 12th in my poll, they are 9th in the BCS composite computer polls, with the problem for them being a weaker then normal strength of schedule for the Big Ten (it hurts Michigan too). So I'm not the only one having problems getting them into the "right" place. In my latest in house revision, West Virginia has indeed moved ahead of Texas A&M (tweaked some of the factors and how their are weighted), which I agreed with, Louisville also got shifted back slightly. I'm still unable to get Wisconsin any higher no matter how hard I try, however I do think the rating will sort itself out as we get closer to Thanksgiving. Thanks for the feedback guys!
Link to comment
Share on other sites
Arlie Rahn Community Regular

Arlie Rahn

Posted
Posted
If you look at the Sagarin ratings, they have similar issues to what Druez explained: [url]http://www.usatoday.com/sports/sagarin/fbt04.htm[/url] Remember, the computer ranking is made only as one factor of many. For a true ranking system (a la BCS), you will also have human polls that help minimize the wierd CPU situations. It is simply a way to get completely numerical analysis without any human bias to add in with the multiple biased polls to help give schools that maybe aren't getting the pub they should a small bump and vice versa.
Link to comment
Share on other sites
Maximus Newbie

Maximus

Posted
Posted
I can somewhat see why computer rankings are used but they appear to be way off at times. I live in Wisconsin (trying to be unbiased here :) ) but I think they should at least be in the top 4 or 5. I also understand that the human element in the polls can lead to biased results. I wish I could remember but there was a formula that I actually did like at one time. Not sure if it was ever used but it was about 2 years ago that someone came up with a formula that was pretty decent. Having said that, I appreciate your development and time spent on doing this Scott. These are the types of things that I like to do as well. For example, I have developed formulas (for my fictional OOTP league) to determine the "manager of the year award" & how they get fired/hired. They are in excel form. Good luck Scott. I am looking forward to this. :)
Link to comment
Share on other sites
Druez Newbie

Druez

Posted
Posted
[QUOTE=Arlie Rahn]If you look at the Sagarin ratings, they have similar issues to what Druez explained: [url]http://www.usatoday.com/sports/sagarin/fbt04.htm[/url] Remember, the computer ranking is made only as one factor of many. For a true ranking system (a la BCS), you will also have human polls that help minimize the wierd CPU situations. It is simply a way to get completely numerical analysis without any human bias to add in with the multiple biased polls to help give schools that maybe aren't getting the pub they should a small bump and vice versa.[/QUOTE] Compare any human ranking vs any computer ranking. Bias or not the human ranking is closer to the majorities vote in most situations. Take it a step further, Scott said Wisconsin should have a higher ranking, but no matter what he does he can't get them ranked higher. That means that a computer poll can try and be influenced to be biased. Scott knows that Wisconsin should be ranked higher and has tried to change his formula to reflect it. Interesting thought, like one of those endless time loops that makes you go crazy :)
Link to comment
Share on other sites
Scott Vibert Explorer

Scott Vibert

Posted
Posted
Exactly maximus... in my case I've gotten to the point where I take it either as an inherent flaw, or an unbiased analysis of the Badgers. As the BCS composite ranking shows, I'm not the only computer model that has the Badgers "low", all the computers do (BCS composite ranking 9, VIB ranking 12, and Sagarin ranking 16) In this case I think it's symptomatic of a weaker then usual Big Ten and a non-conference schedule that featured a terrible UCF team, a bad UNLV team, and a bad Arizona team (whom they only got past by two points (9-7)). While my head/heart says they are better then 12th.. the raw data comes out that they aren't... and if it was say, UConn and not a well recognized team like Wisconsin, I suspect, we wouldn't be having the conversation... as I would have bought the results earlier, and people would have just said.. oh yeah it's UConn. While my gut still says they are underanked in my system, (undefeated Big Ten team, has to be higher then 12th right?) no matter how hard I tweak it, I can't get Wisconsin higher then 12th, so maybe, just maybe that's about right... (*ducks from flying cheese wedges and other things hurled from Badger fans*) ;)
Link to comment
Share on other sites
Scott Vibert Explorer

Scott Vibert

Posted
Posted
[QUOTE=Maximus]Umm, how does Wisconsin loose 12 spots. I can see maybe six or seven spots but not 12. :mad:[/QUOTE] They fell like a stone because they got blown out by a team that was previously outside the top 50 of the rankings. That coupled with a weak strength of schedule (UCF, Arizona, UNLV, OOC none of whom are very good, and having not played neither of the top 2 Big Ten teams) causes them to fall like a stone.
Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Go to topic listing
×
×
  • Create New...