Jump to content

Hongbit

BCS Computer Polls -- Week 4

Recommended Posts

These are 3 of the 6 computer polls that will account for 1/3 of the BCS ratings. The other 3 will become public on October 18th. The polls should sort themselves out in a few weeks, but from the looks of things right now its pretty scary these guys have any say in the BCS.

SAGARIN

1.MIAMI

2.USC

3.SOUTHERN MISS

4.TEXAS

5.COLORADO

6.OHIO ST.

7.GEORGIA

8.AUBURN

9.FRESNO ST.

10.OKLAHOMA ST.

11.ARIZONA ST.

12.UTAH

13.CALIFORNIA

14.TENNESSEE

15.MINNESOTA

16.WEST VIRGINIA

BILLINGSLEY

1.OKLAHOMA

2.MIAMI

3.USC

4.OHIO ST.

5.BOISE ST.

6.TEXAS

7.GEORGIA

8.TENNESSEE

9.AUBURN

10.LSU

11.MICHIGAN

12.FLORIDA ST.

13.MINNESOTA

14.WEST VIRGINIA

15.UTAH

16.PURDUE

COLLEY MATRIX -- This Guy is High

1.AUBURN

2.ARIZONA ST.

3.MIAMI

4.UTAH

5.USC

6.BOSTON COLLEGE

7.COLORADO

8.TEXAS

9.CALIFORNIA

10.OKLAHOMA

11.STANFORD

12.VIRGINIA

13.OKLAHOMA ST.

14.WEST VIRGINIA

15.OHIO ST.

16.ALABAMA

Share this post


Link to post
Share on other sites

The computer numbers are meaningless right now because one strange game can move a team up or down really fast. They become more meaningful later in the season when teams have played more games. I trust computers more than sportswriters and coaches.

BC rocks.

Share this post


Link to post
Share on other sites

Laughable. My Boise State Broncos 5th in the 2nd poll? That's ridiculous. That ranking isn't there because it's early; it's there because the criteria is flawed. So what if we beat two mediocre to bad teams by a wide margin and upset Oregon State. What about strength of schedule? I love my team but I know they do not deserve to be 5th. Heck, we play 8 of our 12 games at home this year. Is this factored in? If this is an example of the new BCS ranking formula it will indeed be an interesting year.

Share this post


Link to post
Share on other sites
Originally posted by IdahoBert

Laughable. My Boise State Broncos 5th in the 2nd poll? That's ridiculous. That ranking isn't there because it's early; it's there because the criteria is flawed. So what if we beat two mediocre to bad teams by a wide margin and upset Oregon State. What about strength of schedule? I love my team but I know they do not deserve to be 5th. Heck, we play 8 of our 12 games at home this year. Is this factored in? If this is an example of the new BCS ranking formula it will indeed be an interesting year.

All those things are factored if the programs are coded correctly. There hasn't been a major problem in the past and there shouldn't be this year. These strange early season blips always go away by the end of the season. By the time the season is over, there will only be minor discrepancies between the various computers and the human polls and people will get to argue about it for weeks.

Before they had the computer results, people used to complain about the final results of the human polls. Now they complain just as much. If they get rid of the computer numbers, people will still complain.

I've keep my own rankings without the use of a computer) for about 25 years and that's what I go by. I don't know what I'm doing but my results come out pretty much the same as everyone else at the end. Bc is unranked and Idaho is ranked about #15 based on a strong season last year and a continuation this year.

Share this post


Link to post
Share on other sites

Almost every major issue with these polls is a sample size issue. Right now there are simply two many teams with the exact same record and any strength of schedule calculation is meaningless because that still only looks at the teams competitions current record. After a bunch more games have been played all those calculations become much more usefull, but at this point we might as well be talking about the league leader in batting after 25 AB's.

In general, I like the computer polls better than the human polls because they aren't impacted at all by where a team was ranked before it even played a game. The fact that pre-season rankings have an impact on where you finish at the end of the year is extremely stupid.

Share this post


Link to post
Share on other sites
Originally posted by tiger337

I've keep my own rankings without the use of a computer) for about 25 years and that's what I go by. I don't know what I'm doing but my results come out pretty much the same as everyone else at the end. Bc is unranked and Idaho is ranked about #15 based on a strong season last year and a continuation this year.

Well, I'm still glad to see Boise State listed so high. I just wonder where we'll be after twelve games if we're undefeated and keep beating up on the WAC teams. We're just not playing the cream of the crop like BCS teams do. It will be interesting to see if we get some BCS bowl bid like the Fiesta if we're in the top ten and have to play a team that's had a far more demanding schedule. By the way, you keep your own data? Wow. We're expected to beat BYU by 20 or so on ESPN Friday night on the smurf turf. We scored over 50 against them on their turf last year. It will be interesting.

Share this post


Link to post
Share on other sites

That's a good analogy QS. It really is too early to tell. I have also started my own poll. I'll post it later in the year as there isn't enough data right now for it to be accurate.

The Computer polls have a few serious flaws:

1) They don't differentiate between playing at home vs. on the road -- Knowing this forum I am sure someone will step up and disagree but its always tougher to win on the road. Why the polls do not take this into account is a mystery.

2) The BCS outlawed any polls using margin of victory -- IMHO, the reason this is important not to account for how much a team wins by, conversely, I think its important to note how much a team loses by.

3) The SOS component is flawed -- If team A beats team B, that loss is not taken out of the equation when using team B's record for the SOS ranking for team A. You are essentially hurting your long term SOS by hanging a loss on the team you beat. In fact, if you lose a game to a ranked team it will help your SOS more than if you won. Additionally, They use a sliding scale and do not account for records & ranking of two teams at the time they play the game.

Here is an example:

Lets say that this weekend, Penn State goes on the road and destroys #20 Wisconsin by a score of 44-7. In Ann Arbor, UofM needs 3OT's to sneak out at 31-30 victory against #24 Iowa. These two wins will count nearly the same in most of the computer polls even though in comparison to Michigan, Penn State pounded a better team and they were on the road. The computers wont account for location or MOV.

Additionally for Iowa, this loss will look the same as the loss by the Badgers. It doesn't matter they lost by one on the road vs. a ranked opponent while Wisconsin laid an egg on its home field against an unranked team. Iowa will get a temporary SOS boost over the Badgers, but it will evened out later in the year after Iowa plays PSU and Wisconsin plays Michigan.

The computers still need tweaking.

Share this post


Link to post
Share on other sites

I agree 1 and 3 are a flaw, I don't agree that 2 is. I also am not convinced that anyone is willing to go through the trouble necessary to actually figure out the difference between a home and road win, which is not nearly as large as most people think and in other sports at least, it is largely a function of travel time (for example in college basketball generally you give a team an extra 3 points in the spread for playing at home, and most research shows that anywhere between 1 and 2 of those points is directly related to the fact that the road team has to travel while the home team does not).

My main disagreement with margin of victory is that it assumes that winning by 7 points is different than winning by 14 points. Maybe a margin of 20 points or more should be counted (then again, why should a team that is winning by 21 points in the fourth quarter but gives up a garbage TD with it's 2nd string in be hurt when they clearly dominated the game) but it is almost impossible to make a meaningfull measurement of how much margin of victory means, especially when you have coahces out there like Stoops in oklahoma who is willing to pass the ball in the 4th quarter of a 55-0 game, and other coaches like Tressell who is willing to let a 14 point lead turn into a 7 point lead if it takes an extra 5 minutes off the clock.

As far as the strength of schedule goes, they absolutely should NOT account for the rankings at the time the game is played. The way a team plays throughout the season is much more indicative of how good a team they are than what the media/coaches thought about the team prior to the season. Beating the #20 team in week 2, only to have that team finsish 6-6 should not be more impressive than beating the #24 team in week 1 that ends up 8-4.

Share this post


Link to post
Share on other sites

I find it pretty laughable that we are 1,10, and then don't even show up in the other poll....colorado, texas, and ok state make it, but I guess we haven't played well enough to crack the top 16:)

maybe we should get rid of the computer polls and allow monkeys to throw darts to determine the last third of the bcs.

Share this post


Link to post
Share on other sites
Originally posted by IdahoBert

Well, I'm still glad to see Boise State listed so high. I just wonder where we'll be after twelve games if we're undefeated and keep beating up on the WAC teams.

I think some of the second rate conferences like the WAC and the MAC are stronger than they used to be. It seems like the top teams from those conferences fare a lot better than they did in the past against stronger conference opponents. That's just based on memory though.

Share this post


Link to post
Share on other sites

Hongbit, I basically agree with your post. The computers are not perfect. My problem with the human polls is that I suspect many of the humans do not see most of the teams play very often and I'm fairly confident a lot of them don't spend much time analyzing the things you just talked about.I think they only notice those things when they feel like noticing those things. Humans are biased. I think the computer polls pretty much do what the humans do but eliminate a lot of the bias.

Share this post


Link to post
Share on other sites
maybe we should get rid of the computer polls and allow monkeys to throw darts to determine the last third of the bcs.

Again, I would point out that computer polls at this time are about the same as having a monkey throw darts. Thankfully, the last third of the BCS is not determined four weeks into the season, or I imagine the human polls would be just as usefull as having the monkey throw darts. Want to compare the human polls now to the human polls in 10 weeks and see how different they are?

Share this post


Link to post
Share on other sites
Originally posted by tiger337

I've keep my own rankings without the use of a computer) for about 25 years and that's what I go by. I don't know what I'm doing but my results come out pretty much the same as everyone else at the end. Bc is unranked and Idaho is ranked about #15 based on a strong season last year and a continuation this year.

Wait a second here Lee. You mean to tell me that you are such a fan that you keep your own ratings. Where the hell have you been? You post over here on the college sports forum about as much as a Gary Knotts fastball finds the catcher's target. Who is your team? Who do you like to win it all? How about a Heisman top 5 list or some all america contenders. Give us some insight.

Share this post


Link to post
Share on other sites
Originally posted by Hongbit

Wait a second here Lee. You mean to tell me that you are such a fan that you keep your own ratings. Where the hell have you been? You post over here on the college sports forum about as much as a Gary Knotts fastball finds the catcher's target. Who is your team? Who do you like to win it all? How about a Heisman top 5 list or some all america contenders. Give us some insight.

I really don't follow football closely enough to give any good insight. I used to watch it more when I was younger. The only reason I'm still able to do the ratings is because I've been doing it for so long and I've got my system down pretty good. I'll try to get over here more often though.

Share this post


Link to post
Share on other sites
Originally posted by Hongbit

2) The BCS outlawed any polls using margin of victory -- IMHO, the reason this is important not to account for how much a team wins by, conversely, I think its important to note how much a team loses by.

Screw MOV. How about important stuff, like did they cover the spread. I know we'll never see it, but I'd love to see performance against the spread used as a factor.

I agree Hong on your SOS statement. Perhaps it's because my Gators have 5-6 ranked teams on their schedule every year, but I loathe the fact that SOS doesn't carry more weight.

As far as some of the smaller conferences. They get the shaft when the Big Easy sits there with an automatic BCS Bowl game.

Share this post


Link to post
Share on other sites
I know we'll never see it, but I'd love to see performance against the spread used as a factor.

Why? The spread is simply a number set up to get as close to half the people betting on each team? Adding the spread into the polls would really just be like letting the people (and by people I mean the gambling portion of the population) vote, and is that something we think we need? I'd prefer college football not become something like American Idol personally.

Share this post


Link to post
Share on other sites
Originally posted by detroittigers

gary knotts since coming off the DL, 35.1 innings, 3.84 era not bad

He still doesn't make the Top 25.

Share this post


Link to post
Share on other sites
Originally posted by qsilvr2531

Why?

Because I think it's gives a good idea of what is expected. If a team plays a weak schedule and barely beats their opponents(doesn't cover). Than perhaps that 10-1 record ain't so impressive compared to an 8-3 team that handily beats teams they should and comes close on the road against stronger teams.

Share this post


Link to post
Share on other sites

But does it give a good idea of what should be expected, or just what is expected? Especially early in the year, how much can you read into a spread when you don't really even know how good a team is? For the first five weeks of the season even the best college football analysy/coach doesn't know enough about all of the teams to make a judgement on what should be expected, much less the average gambler. Does Fresno State deserve extra points because they were underrated going into the season?

Basically, I don't think teams should get points based on the perception of their teams. The winners and losers should be decided based on performance. That's also my problem with margin of victory, it implies that winning by 24 is different than winning by 1.

Strength of schedule is another story, and should get factored very heavily. I'm not as convinced on the home/road, only because I don't think anyone can quantify it.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...