Tuesday, January 02, 2007

Bowl Challenge Update

Results for the past 2 days –

3rd tier ACC (Miami) over 2nd tier WAC (Nevada) – ACC +2

2nd tier Big 10 (Penn State) over 3rd tier SEC (Tennessee) – Big 10 +2

2nd tier SEC (Auburn) over 2nd tier Big 12 (Nebraska) – SEC +3

2nd tier Big 10 (Wisconsin) over 2nd tier SEC (Arkansas) – Big 10 +3

1st tier Big East (West Virginia) over 2nd tier ACC (Georgia Tech) – Big East +2

1st tier Pac 10 (USC) over 1st tier Big 10 (Michigan) – Pac 10 +3

1st tier WAC (Boise State) over 1st tier Big 12 (Oklahoma) – Big 12 minus 1

Our current standings -

ACC = 11

Big 10 = 5

Big 12 = 6

Big East = 6

Pac 10 = 5

SEC = 13

We have the Big 10 getting on the board finally with its first two bowl wins of 6 games played. The ACC also continues to surprise. With losses by LSU and Florida, and a win by Wake Forest, the ACC would win the entire contest.

Games remaining –

Jan. 2 Orange Bowl at Miami, Wake Forest vs. Louisville (1 ACC v 1 Big East)
Jan. 3 Sugar Bowl at New Orleans, Notre Dame v LSU (1 IND v 1 SEC)
Jan. 6 International Bowl at Toronto, Cincinnati vs. Western Michigan, (3 Big East v 2 Mid Am)
Jan. 8 BCS National Championship at Glendale, Ariz., Florida v Ohio State (1 SEC v 1 Big 10)

We will update again after the Sugar Bowl.


For our format, click here.

16 comments:

Anonymous said...

I'm still not entirely clear on what the result of this system will show. Right now it looks like its not going to determine which leagues are better but which are bigger.

If it's based on sheer totals then wouldn't that be like playing fantasy football with someone where I have 9 players and you have 5? No matter how much better your players do I'm more than likely going to win by having almost twice the ability to generate points as you.

I think this should be more like recruiting ratings that rank the "average" recruit. you can get 5 star through 1 star players but the average "stars" goes a big way in determining the best class. In this case we could see the average points per win or per game for each conference and therefore compare apples to apples.

If the result is to see the sum of points then the SEC is probably going to win. We have the most bowl teams and statistically we have a much greater chance of winning by the law of averages. But if the Big East can go 5-0 and the SEC can still win the contest despite going going 5-4 that seems like a flawed formula to me.

Either penalties for losses need to be introduced or an average "per game" determined for a better comparison.

1BigNolesfan said...

There is also a huge difference in how some of the teams are scored. It appears that some teams status have been inflated in an attempt to help the SEC teams.

For example, why is a 4 loss Penn St. a 2 tier team, but a 3 loss Tenn. team is a 3? It should have bee nthe other way around. There is also NO WAY that Houston should be a 1 tier team. Both of these scores defy logic and benefit the SEC teams.

Using a more fair and even scoring system, I have the follwing scores.

ACC 11
Big 10- 6
Big 12- 8
Big East- 5
Pac 10- 8
SEC- 12

Like the first person said, the SEC having more teams gives them a huge advantage in this system, but it appears that was your goal.

Henry Louis Gomez said...

Mergz will have to adress the the issue you raise and it's a good one. I was thinking along the lines of percentages. What if we calculated the total number of possible points each team conference could get if it went undefeated in Bowl Season then we could calculate the percentage of available points that each conference in fact got. With a max of 100 and min of 0 every conference would have an equal shot. I would require that a conference have a certain number of teams playing in bowls though. A conference with 2 teams could much more easily score 100% than a big conference with a lot of Bowl teams.

Henry Louis Gomez said...

1BNF,

Tiers 1 and 2 each have two teams in them for all the conferences being measured.

Tennessee in not better than either any of the 4 top SEC teams (FLA, LSU, Auburn, and ARK) and therefore by default is in the third tier.

The first tier for the Big 10 is OSU, Mich. 2nd Tier is Wisconsin and Penn State. Who do you suggest is better candidate for tier 2? Purdue, Iowa or Minnesota?

You need to check out the original post.

Link to original post

Henry Louis Gomez said...

I have another idea that might be easier for mergz to carry out. All the conferences have 4 teams in the top 2 tiers. It's the 3rd tier that is variable in size. We could just use scores from those 4 teams OR we could average the scores from Tier 3 and multiply that number by 2 (so that tier 3 would be on equal footing with tiers 1 and 2).

1BigNolesfan said...

Your standing as a 1,2, or 3 tier SHOULD be placed how good of a team you are. There is no resonable explanatio for a team with a worse record to be ranked higher than a team with a better record.

The thought that every conference has the same number of teams in each tier is very flawed and simply not accurate.

Henry Louis Gomez said...

I have to disagree with you. . The whole purpose of the exercise is to see which conferences are stronger.

The fact is the best team in one conference could be a worse team than the the 3rd best team in another. When a conference is weak it's 1st tier teams will be weak. The tiers are within the conference not within all of football. In other words we know who the best teams in each conference are because there's plenty of conference play to judge by what we don't know is how teams from different conferences match up against each other.

Like it or no Wake is the ACC's top team.

1BigNolesfan said...

Do you really need some formula to tell you that UF is a better team than Houston?

I just think the way it was done is very flawed. Any system that has Houston on the same level as LSU, UF, OSU UL...etc is useless IMO.

Henry Louis Gomez said...

1BNF,

I'm going to try to stay civil because we have a long history.

First of all this isn't to determine whether Florida is better than Houston. I thought that was implicit in all the talk about conferences.

Secondly Houston is in the top tier of their conference. Do you dispute that? Which conference USA team belongs in tier 1 over Houston?

Thirdly Florida didn't play Houston, South Carolina did. South Carolina is certainly not a 1st or 2nd tier SEC team.

fourth, South Carolina only gets 2 points for beating Houston because CUSA is a non-BCS conference. If S. Carolina had lost it would have lost a point because of this.

fifth, If you don't like it, fine. Whatever. Go start your own blog and then you can write about whatever you want (even how much we suck over here). I'll give a you a link on our blog roll.

1BigNolesfan said...

First, as far as having a long history, I have no idea who you are.

Second, I agree that Houston is the best in their conference, but that DOES NOT put them on the same level as the teams that I mentioned ealrier. I dont think that having all conferences with the same # of teams in each tier is an acuurate way of judging them, when they play other conferences.

Third, agreed, but not sure of your point.

Fourth, the way I read the scoring, I thought if a 3 beat a 1, it got 5 points and if the 1 beat the 3, it only received 1? If
that is not how the scoring works, then maybe I missed something.

Fifth, I dont think I have to go start my own just because you disagree with my comments.

Henry Louis Gomez said...

1BigNolesFan is the same handle that a frequent poster at one of the Gator message boards I frequent goes by. I thought it was you since we advertise this blog on that board and that board advertises on here.

Will you agree that every conference has good teams and bad teams? That tiers do indeed exist within each conference?

If so then we only need to agree to on which teams belong to which tier.

The BCS conferences can each put two teams into a BCS Bowl. So is it safe to say that making a BCS Bowl would qualify a team as being in the top tier of their conference?

So if each conference has 2 top tier teams we need to start filling in the 2nd tier. Here it gets somewhat trickier. Mergz (and he'd probably admit) arbitrarily decided how many teams make second tier but he made it equal for all the conferences. In other words the 3rd and 4th best teams would be the 2nd tier teams in each conference. The remaining bowl teams would make up the third tier. Since we have two teams in each of the first two tiers that's why I suggested averaging the score for the remaining teams and multiplying it by 2. In essence you'd have 3 tiers with 2 teams in each tier for each conference (the two 3rd tier teams would actually be a composite of however many teams are in tier 3 for each conference).

Again the point isn't to determine whether the SEC is better than CUSA. That's why we're not measuring CUSA. We're measuring the BCS conferences. But a lot of the BCS conference teams are playing non-BCS teams so we have have a mechanism whereby we qualify those results.

Yes you missed that part, maybe because you didn't read the original post closely. That's OK. But I'm telling you now that BCS conference teams that beat top tier non BCS conference teams only get 2 points which is less than they would if their opponent was a top tier BCS team. And the BCS conference also gets penalized for losing to a non BCS team (something that doesn't happen if a BCS team loses to another BCS team). So you earn less and can actually lose points.

Just because you don't understand something doesn't mean that there isn't a method to the madness.

Mergz would probably agree that it's an imperfect system and he has changed it as a result of valid suggestions. But that fugazi trophy they give to the conference with the best winning percentage in Bowls is way more flawed. It doesn't take into account the supposed quality of the opponents. It gives each bowl game the same weight which is ridiculous. Miami winning the MPC bowl against Nevada shouldn't carry as much weight as USC beating Michigan in the Rose Bowl.

1BigNolesfan said...

I do use this handle at another site, but do not remember anyone going by your name there.

Yes, I agree that every conference has different level teams.

I understand HOW it was designed, I just didnt agree with the concept that a 3 loss team from a BCS conference could be ranked worse than a 4 loss team from another BCS conference. I think you should go by record. That way you are going to end up with a different number of teams in certain tiers in different conferences. Surely you would agree that all conferences do not have the same number of teams at each level based on records. For example, Wisconsin ended the year at 12-1, but was a second tier team. I thought all 0,1,2 loss teams from a BCS conference should have been a 1 tier, all 3 loss teams from BCS a 2 tier and all teams from BCS having 4 or more losses a 3tier.

I knew I didn't read all of the posts on the subject, but thought I had the point system down correct. I missed anything saying that non BCS teams were being scored differently.

Where I had a problem was in the UT, Penn St game and the SC game, which seamed to favor the SEC in the scoring. I now konw I was wwrong on the scoring of the SC game because of the non BCS thing, but feel the UT game was still wrong.

Henry Louis Gomez said...

My handle on Zook Free Zone is FishFan-GatorMan.

I just don't think UT is a second tier SEC team (this year) and the result of the game proved it. Anyway they couldn't have qualified for second tier given the number of slots that were open in each tier.

Yes the format is arbitrary. But what you are suggesting is just as arbitrary (if not more). As we know all losses are not equal. Besides Mergz' approach did incorporate W-L inasmuch as the order of each conferences teams were informed by the BCS poll which has a computerized component.

1BigNolesfan said...

I dont understand how using the records from teams that are both in BCS conferences is not being as fair as possible. You referenced the BCS, but the fact is that UT was #17 in the BCS and Penn St was not ranked.

All things are NEVER going to be exaclty equal when comparing such things.

Could you explain why you feel that UT was not a second tier SEC team and how Penn St quailified for one in the Big 10?

Henry Louis Gomez said...

Again the whole point is to determine the strength of the conferences. Penn State doesn't get as many points for beating for beating UT as UT would have gotten for beating PSU. You think that's unfair because PSU had a worse record.

But you can't judge the strength of a conference from conference games (you could be undefeated in a crappy conference) W-L record is not enough. Remember 2/3 of all regular season games (3/4 for Pac 10 teams) are conference games. The remainder are usually comprised of an out of conference rivalry game, and 3 patsies for the home opener, homecoming, and another filler game.

Penn State is a second tier team in the Big 10 this year because the Big 10 down. There is a big disparity between the top two teams and the rest of the conference but that doesn't mean a second tier doesn't exist in that conference. PSU is the 4th best team in their conference and as you said they are unranked. Meanwhile UT is at best the 5th best team in it's conference. The fact that UT is a third tier team in the SEC speaks to the depth of the SEC relative to the Big 10. It turns out that a 2nd tier (IN ITS CONFERENCE) Penn State is better than a 3rd tier (in its conference) Tennessee. The result is consistent with the premise. I don't see what the problem is.

Henry Louis Gomez said...

Here's another way of looking at it. If we had control of who played who in the bowl games and wanted to see which conference was better, the SEC or the Big 10 we would have the teams square off 1 vs. 1, 2 vs 2, 3 vs. 3, etc.

In this case we have the 1 vs. 1 match-up in UF/OSU. We also have a slight disparity in the next game. Wisconsin (was) the Big 10's 3rd best team going in while Arkansas was the SEC's 4th best team.

Likewise Penn State is the 4th best team in the Big 10 (they tied Purdue for 4th but beat purdue head to head so in my book they are 4th and Purdue is 5th) and UT is the 5th best team in the SEC.

Now let's assume the conferences are of equal strength throughout, meaning that the hypothetical 1 vs. 1, 2 vs. 2 match-ups would be very even. If they were truly even they would play to infinite ties. But when you take the teams and square them up out of sequence you expect the the teams that are ranked higher in their conference to beat those teams ranked lower in the other conference. So Mergz system rewards the "underdog" (because we assume going in that the conferences are equal but its the match-ups that aren't) for beating the "favorite:. Likewise the "favorite" gets less of a reward for winning the the game it "should have won." Again this assumes that the 4th best Big 10 team is better than the 5th best SEC team. Like I said, that's exactly how it turned out. This doesn't prove the system is infallible, hell it doesn't prove anything. But the result is consistent with the theory.

Man, we're taking grief for hypothetically putting the Big 10 on the same level as the SEC.