enigmaax;522713 wrote:Here, I'll tie a couple things together for you. In 2007, Kentucky beat #9 Louisville and #1 LSU. That is two top ten teams. They also beat EKU, Kent, and Florida Atlantic - all similar to several Boise opponents. They also beat Vandy, Arkansas, and Florida State - all better than most Boise opponents. And they lost to #6 South Carolina, #10 Georgia, #12 Florida, as well as Tennessee (10 wins) and Miss State (8 wins).

If you want to make it a two game season from any given year in the last five, you can make a case for the a lot of teams. The reason those couple of wins here and there for Boise don't count for much is because other teams that do the same thing can't stop with those wins and say, "See, we are clearly the best" because they have to play plenty of other teams that are similar. Boise wins its two games and then beats 10 other teams that 62 out of 66 BCS schools would beat regularly as well. Doesn't mean a thing.
You may be right about Boise having talent. But talent doesn't mean anything if you don't earn your way in life. It is kind of like a spelling bee where you are calling Boise the greatest speller around because they nail cat, fat, and hat every time.

enigmaax;522729 wrote:So, your whole premise is that you know more than anyone and that is why the polls are flawed/biased/etc. Thread dead.

And we can all have a good laugh at Hawaii and Fresno State as your point. Was it the losses to Colorado and Mississippi that really proved to you how great these programs are?

Zombaypirate;522735 wrote:No, the point which you miss entirely is the rankings are superfluous.

ytownfootball;522741 wrote:Then why...pray tell...do you use wins in the past few seasons to bolster your argument that Boise should be ranked as highly as they are?

You can't have it both ways.

Zombaypirate;522733 wrote:No, you cannot. Kentucky has changed dramatically over the last couple seasons where Boise St has not. Comparing apples to oranges.

WebFire;522748 wrote:Only 2 things to comment on...

1. I think we all agree pre-season polls are worthless. I wouldn't even release one until after week 6. That seems to be the point in the season where we truly get a good idea of who is real.

2. Why are throwing out what Boise St. did last year, but I've seen reference all season from Buckeye fans to the Rose Bowl win from last year? Two-way street fellas.

centralbucksfan;522692 wrote:That would have been 2 weeks ago. How different would that have been from the preseason poll? A little here and there maybe. I dont' disagree. But in the end, the polls typically all work itself out.

ytownfootball;522694 wrote:That will absolutely NEVER happen though...ESPN, Rivals AP...they make a living off that shit. Comes down to coin as we all know.

Zombaypirate;522696 wrote:Yes SOS is nothing. How do you arrive at your SOS? Please describe in detail how you come to your conclusion.

fan_from_texas;522779 wrote:http://www.masseyratings.com/theory/sched.htm

Good enough?

enigmaax;522744 wrote:Kentucky changed dramatically within the same season. Because they played more than two teams. That is what Boise doesn't do. All you have is that Boise looks, to you, like they could beat anyone. Well, great...but it isn't a two game season and then what you think would happen. That is exactly what you hate about the polls now, it is all based on what people think. The difference is that it changes weekly based on results and you want to lock it down for Boise after their two wins (last year, even?).

Zombaypirate;523223 wrote:If you consider circular reasoning to be somehow accurate.

Zombaypirate;523223 wrote:If you consider circular reasoning to be somehow accurate.

enigmaax;523232 wrote:I just took a quick glance at the link and maybe I need to look closer, but I actually agree that the SOS element is flawed in the explanation, though not because it is circular.

Its the fact that the schedule evaluation is relative to the perceived status of the team playing the schedule. It sounds as though Team A who is expected to win a lot of games and Team B who is not expected to win a lot of games could play the same exact schedule but Team B's schedule would be considered "stronger" because they played up.

There are simple formulas that could be based exclusively on results (and it will never be perfect) that would be a more true representation. Saying, "right now you are this good so you should beat this team so it isn't as tough for you" basically punishes teams for being good to begin with.

Zombaypirate;523227 wrote:I do not want to lock down anything. Boise St is the number one team right now. When they finish undefeated with another BCS bowl win over an alleged power team it will be shown on the field. I just hope they draw someone from the SEC or Big 12 can't wait for the excuses and rationalizations just like when Utah showed they were the best team after they defeated Alabama in 2009 Sugar bowl.

Time will tell the truth.

fan_from_texas;523238 wrote:What part of it is circular?




No, you're misunderstanding the math behind it--that's not what it's saying. The idea is to consider the number of wins a median strength team would have against a particular schedule--that's not based on record, but on expected wins, so playing a 10-0 and 0-10 team isn't treated the exact same as playing 2 5-5 teams, which makes sense.

enigmaax;523242 wrote:I get that part...and don't disagree. Thats one of the things I point out about the OSU schedule when everyone talks about how bad ass their OOC schedule is. They play 1 out of 4 games with a reasonable potential for a loss. Playing one great team and 3 gimmes isn't necessarily better than playing 2 decent teams that could pull an upset on a given day plus 2 gimmes.

And again, I did just take a quick glance. I thought there was something about a team being great playing a certain schedule vs. a team being bad playing a not bad schedule. I'll have to look closer.

fan_from_texas;523248 wrote:That's a prescription for teams wanting to beef up their SoS under the current season, not necessarily how to get a good SoS. You get a good SoS by playing tough teams. The general idea is that they look at tOSU's schedule and say the average BCS team would lose, say, 4.5 games, while some other team's schedule would yield an average of 4 losses to an average BCS team. The conclusion is that the former schedule is more difficult. The distribution of the teams (great vs okay vs bad) isn't as important--it's a way to level the playing field.

enigmaax;523240 wrote:Mmmm....no. Utah didn't show that. At all. Again, you miss this part - Boise can go undefeated against a schedule that probably 25 other teams could go undefeated against. And another 25 teams would have top 10 caliber records playing the same schedule. Baylor, one of the worst BCS programs, has gone like 10-3 against Boise-opponent type games. Would you put them in the top ten playing and winning against Boise's schedule?

Boise might get in the title game. They may even win the title game this year. Still doesn't mean they earned it. You say Boise would win this and that. I say Boise would lose at least 2 games in any major conference....in their best year. We don't know who is right because they don't play that schedule. But history has shown the demo in which Boise falls (including Boise itself, who was destroyed by Georgia and lost to a porous Washington team...and who has an overall record against BCS schools of something like 7-17) has been consistently inferior when playing BCS conference competition.

You can say, oh look Hawaii, Fresno, and Nevada are good. But I can say....Oregon, USC, Washington, Arizona, Arizona State, UCLA...LSU, Alabama, Florida, Tennessee, Auburn, Arkansas, Georgia...Ohio State, Wisconsin, Iowa, Michigan State...on and on....would ALL go undefeated in Boise's conference. And what will you counter with? Boise beat Oregon l ast year? Or Alabama lost to La-Monroe three years ago? You're going to pick one or two upsets and try and pass that off as fact. I'm going to get a real sample size over the last 50 years and show you that it just isn't going to work out for teams like Boise...or TCU (who flailed hopelessly at that level for 40 years).

fan_from_texas;523238 wrote:What part of it is circular?




No, you're misunderstanding the math behind it--that's not what it's saying. The idea is to consider the number of wins a median strength team would have against a particular schedule--that's not based on record, but on expected wins, so playing a 10-0 and 0-10 team isn't treated the exact same as playing 2 5-5 teams, which makes sense.

Zombaypirate;523316 wrote:I will make this as simple as possible. From where do you start?

Who is the best team? Who is second? third etc? How do they come to those conclusions? I keep hearing and have already seen here on this thread SOS.

Well how do you know what schedule is the toughest? From where do you start?

Where is the starting point to base who is the best team in the country? That is where it is circular reasoning.
.

Sauceda Rating System Description

This rating system is based on the Elo rating formula for chess. I developed it a few years ago for football, and recently expanded it for other sports. For each game, 1 "game point" is disputed. The winning team gets the largest part, and the loser gets the rest. The formula is:
gp = 1 - 0.4 ^ (1 + pd/11) (Note: the 11 constant is a typical value for basketball; for other sports it may be smaller).
As an example, here are sample gp values for several margins of victory.
Team wins by 1 point: gp = 0.632 (the opponent gets 0.368) Team wins by 10 points: gp = 0.826 Team wins by 20 points: gp = 0.924 Elo's formula states that a team's "winning expectancy" (we) can be determined by
we = 1 /(1 + 10 ^(Rb - Ra)/400) where Ra is the team's rating and Rb is the opponent's rating. Basicly, we gives the percentage of game points that a team should win against a given opponent.
A team's total we and gp can be obtained by summing over each of its games. The new rating is calculated by
Rn = Ro + K (gp - we)
where Rn is the new rating, Ro is the original rating, and K is an constant that determines the speed of convergence.
If a team's overall performance is better than expected (i.e. gp > we), then its rating will improve (notice that this does not necessarily require winning!). Conversely, the opponents' ratings will decline. Each team begins with the same initial rating (1000). With each iteration these ratings will adjust themselves until equilibrium is reached.
The published ratings also include a homefield correction variable that likewise reaches an equilibrium. The modified formula is:

we = 1 /(1 + 10 ^(Rb - Ra +/- h)/400) Making Predictions
As it stands, the formula for we should approximate the expected game points (gp). However, it would be more interesting to instead estimate the probability of victory. This is accomplished by replacing 400 with an appropriate prediction constant (k), which will vary among the various sports. As an example, assume that k = 300, Ra = 1100, and Rb = 900. Then the probability that A would defeat B is
1 / (1 + 10 ^ (900 - 1100)/300) = 0.8227 = 82.27%
In general, to find the probability that A defeats B use the following prediction formula:

1 / (1 + 10 ^ (Rb - Ra +/- h) / k))

fan_from_texas;523327 wrote:Did you bother reading the link I posted at all? Are you familiar with how ELO ratings work? They don't have any bias as far as starting point for teams--everyone starts the season the same. Teams that win against quality opponents (which are teams that play well against other teams) move up, while teams that don't play well move down. It doesn't matter if they're in the WAC or the SEC. It's an interative process. It's the same idea that Vegas uses, which is why their point spreads are so accurate. It's not like Alabama gets bonus points for being Alabama--all teams have the same starting point, then diverge based on results.

For a more technical explanation, read this:


The fact that good teams end up on top isn't reflective of bias; it's reflective of reality. This isn't circular at all.