Weird. I don't recall even bringing up the Big Ten.... (I think they're god awful presently, fwiw.)

The SEC is the deepest conference. Whoever makes it out of the SEC championship is usually pretty damn good, if not the best team in the country. The next best team is usually very comparable to any other champion from the P5. But if it was by far and away the greatest conference to ever play football in the history of earth, then two programs who struggled to maintain relevance in the Big XII shouldn't have been able to walk right in and challenge the SEC title from day one. They should've been dumptrucked.

The SEC West is very good. But I'll stop short of saying 4 of the best 5 teams in the nation reside there.

Situation,

Based off of your rankings, are you able to make any kind of predictions week-to-week or end of season? Do you simply say the higher ranked team is going to win? Or is it not that simple?

Predicting which team will cover the spread is exponentially more difficult than picking the straight up winner. The spread is designed as an equalizer. In theory the bettor is faced with a series of coin tosses when facing the spread.

A potential exists to create a week to week advantage significantly greater than a coin toss however.

I will be noticeably less forthcoming about my spread picking methodology. But I will say the OPPA++ Rankings are the foundation.

For example,

Last week 80% of OPPA++ missed picks came on one possession point spreads. (A missed pick defined as team with a numerically lower OPPA++ ranking wins; ie #23 predicted to beat #33)

Data strongly suggests whoever wins a one possession point spread matchup (favorite or underdog) will cover the spread. (ie Favorite wins by >spread. And Underdog wins therefore cover).

I won't openly discuss what percentage of the one possession point spread matchups are upsets (although I've put enough information in this thread to glean a general idea).

What's significant in this example is a significant majority of OPPA++ upsets are by the home team underdog. And I will go further to state that their is a strong correlation between the home team underdog upset and the OPPA difference between opponents.

In other words:

I know that OPPA++ cannot pick every game straight up (this is what I've referred to as a fruitless task; an unachievable ranking system ideal). The greatest weakness lies in the one possession spread matchups (which may be obvious to some). But I do believe strong correlations exist to identify some upsets and some favorites who would likely cover. These correlations extend beyond one possession spreads, but weaken as the point spreads increase in number.

I agree it's easy to say in general home teams can upset opponents. But a significant advantage is created when the bettor can say which home teams most likely will not upset their opponent.

For weeks 8 and 9 my rankings were 10 of 10 on away teams that beat one possession spread matchups that met specific criteria. I won't disclose that criteria. In any case this is a limited sample. I have the intention to expand my analysis historically when I have the time.

But I will throw you a bone and say that the majority of road underdogs that upset are within 15 OPPA rankings of their opponent (and that's not to say all of those close matchups become upsets, only a percentage do). The data would suggest a road upset comes from a specific sect of relative competition. The home upset comes from a significantly wider range of opponents.

In other words, there are teams that can upset a specific opponent at home and on the road. There is a grouping of teams that are likely able to upset a specific opponent at home, but not on the road. And there is a grouping that is likely not able to upset a specific opponent all together.

Often the home/road matchups are too close to call. Only a few one possession spreads each week will feature an opponent who is likely unable to upset a specific opponent all together.