Looking over the SEC PSEGs

[Blue Hen]

The SEC has a relative match-up disadvantage in 8 of the 10 games.
... a relative 'home' advantage in 7 of the 10 games, and is favored in 7 of the 10 games.

No point here....just an observation.

[GatorGrad]

The SEC has a relative match-up disadvantage in 8 of the 10 games.
... a relative 'home' advantage in 7 of the 10 games, and is favored in 7 of the 10 games.

No point here....just an observation.

In the two games where the SEC does not have the matchup disadvantage, what is the other game other than Texas vs Alabama? Obviously that game is even with both teams being 9-0 in conference play. Does the SEC have the relative matchup advantage in any games at all? Too lazy to look.

[Blue Hen]

UGA (.500) vs TA&M (.375)

[NCT]

UGA (.500) vs TA&M (.375)

And Georgia has to travel well over twice the distance as the Aggies.

[JamesHowell]

> The SEC has a relative match-up disadvantage in 8 of the 10 games.

The SEC has a relative match-up advantage in 6 of 10 games.

[GatorGrad]

> The SEC has a relative match-up disadvantage in 8 of the 10 games.

The SEC has a relative match-up advantage in 6 of 10 games.

James, we're talking about comparing conference winning %. Not rankings or point spreads.

[JamesHowell]

That would be a non-relative comparison. It is like saying an artichoke is a "relatively" larger edible vegetable than a tomato. When, in reality, although the actual vegetable is bigger, the edible portion is much smaller. Comparing conference records can, by definition, only be "relative" when the teams are in the same conference. Comparing in this way is a huge distortion of the matchups.

[aufan59]

That would be a non-relative comparison. It is like saying an artichoke is a "relatively" larger edible vegetable than a tomato. When, in reality, although the actual vegetable is bigger, the edible portion is much smaller. Comparing conference records can, by definition, only be "relative" when the teams are in the same conference. Comparing in this way is a huge distortion of the matchups.

The purpose isn't to compare teams, it is to compare conferences...

[JamesHowell]

Again, comparing a conference (just a collection of teams) with disparate criteria is a huge distortion. Conference records are only relative WITHIN a conference. Now if you were comparing an SEC bowl team to, lets say, a sunbelt conference team, for a relative comparison using conference records, you would somehow have to magically infer what the SunBelt team's conference record would be were they in the SEC or the SEC teams record were they in the SunBelt.

If a 7-1 SEC team plays an 8-0 SunBelt team, the SEC has the RELATIVE advantage (by quite a bit).

[GatorGrad]

That would be a non-relative comparison. It is like saying an artichoke is a "relatively" larger edible vegetable than a tomato. When, in reality, although the actual vegetable is bigger, the edible portion is much smaller. Comparing conference records can, by definition, only be "relative" when the teams are in the same conference. Comparing in this way is a huge distortion of the matchups.

Apples to oranges. We're talking about two different things. We're talking in relation to drawing conclusions on conference strength. For example if a 7-1 SEC team beats a 3-5 ACC team, that doesn't tell me much. But if the 3-5 ACC team wins, that is telling. It would lead someone to wonder how a team that only went 3-5 in ACC play could beat a team that went 7-1 in SEC play and some would question how tough the SEC really is. Not by just one single result of course, but if there is a pattern that develops, that is interesting data.

[GatorGrad]

Again, comparing a conference (just a collection of teams) with disparate criteria is a huge distortion. Conference records are only relative WITHIN a conference. Now if you were comparing an SEC bowl team to, lets say, a sunbelt conference team, for a relative comparison using conference records, you would somehow have to magically infer what the SunBelt team's conference record would be were they in the SEC or the SEC teams record were they in the SunBelt.

If a 7-1 SEC team plays an 8-0 SunBelt team, the SEC has the RELATIVE advantage (by quite a bit).

Still talking about two different things here.

[JamesHowell]

I realize we are talking about to different things. However, your "thing" is not "relative" advantage or disadvantage.

Furthermore, even if it were relative, it would only be useful in comparing conferences if the teams in the respective conferences played all other teams in that conference. I would hold that an SEC team that didn't play Florida, Georgia, and Tennessee, is not equal to an SEC with the same conference record that didn't play Vandy, Kentucky, and South Carolina.

[GatorGrad]

I disagree. More often than not there is a large enough sample size to make the conference winning % relevant. In the SEC, ACC, and Big 12, teams play 8 of a possible 11 opponents, sometimes 9 of 11 if they advance to the CCG. In the Big Ten, teams play 8 of a possible 10 opponents. In the PAC 10 and Big East, they play all of them.

If a team goes 4-4 (.500) in Conference A but defeats a team that went 7-1 (.875) in Conference B, that is relevant data when comparing Conference A to Conference B, IMO. It tells me that a team could win .875 of their games in Conference B, but lost to a team that only went .500 in Conference A. So not only did the .875 team lose head to head to a team from Conference B, but they lost to a team that only could win half of their games in Conference B. Makes you wonder what the .500 team's conference record would have been if they were in Conference B instead and could support the idea that the competition was tougher in Conference A.