Projecting the Pac-12's 2014 conference race using advanced stats

Oregon has no offensive line. Stanford can't close games. Arizona can only close games. UCLA just ain't right. USC leaves its brain on the sideline for about 15 minutes per game. Arizona State is young and silly. California is younger and sillier ... and currently has a half-game lead in the North division.

Every team in the Pac-12 has played one to three conference games, but only one has yet to lose, and only two are more than one game from .500 (2-0 Arizona and 0-3 Colorado). Of the 14 finished conference games, 11 have been decided by seven or fewer points. Two have been decided by a successful Hail Mary.

This conference is a hot, delicious mess right now, an [insert regionally specific messy food/drink of your choice here -- a Pink's chili dog? Pancakes at Walnut Cafe? A delicious shot of Fireball?]. It features seven teams ranked in the top 27 of the current F/+ rankings, but all of them have what seem to be fatal flaws. It will likely be a Playoff committee nightmare, with about a 50 percent chance of producing a very strong, two-loss conference champion.

Week 6 was absurd even by the Pac-12's absurd standards. Arizona upset Oregon in Eugene as a three-touchdown underdog, Utah won at UCLA with a late field goal, California gave up 812 yards to Washington State and won on a missed field goal, and Arizona State landed a Hail Mary against USC. Plus, Stanford fell via final-minute touchdown at Notre Dame. Oregon State beat Colorado by five points in what was easily the least interesting game of the week.

After such a ridiculous week, it's a good time to take stock. There are a couple of big games on the Week 7 docket (Oregon-UCLA and USC-Arizona), and for this Win Projection Wednesday, let's focus our lens on the zaniest conference in college football.

Win projections

Once again, we'll use win projection data that I post weekly at Football Study Hall. The probabilities are culled from the F/+ ratings.

As things currently stand, the most likely Pac-12 title game will feature an Oregon-Arizona rematch.

But Stanford and UCLA have direct roles to play, and California, USC, and Arizona State are all within 1.5 projected wins of the top spot. It won't take many upsets -- if you can call any result in the conference (not involving Colorado) an upset at this point -- to quickly rearrange the hierarchy. But Stanford's own faltering has allowed Oregon to stay in the North's pole position (albeit with a projected 6-3 record), and Arizona's Hail Mary win over Cal and Thursday-night win in Eugene have given the Wildcats the early South edge.

Sure bets

Each week, I re-emphasize the point that the easy games are as important as the difficult ones. This conference has proved that tossup games are everywhere you look, and in the end, having a high number of relatively routine wins will increase your odds better than anything.

70+% games: Stanford 4, Arizona 3, Oregon 3, UCLA 3, California 1, USC 1, Utah 1, Washington State 1

Stanford's got all sorts of issues at the moment, namely that it's fielding its worst offense since about 2007. But the Cardinal defense is still dominant, and David Shaw's team still gets home games (and likely wins) against Washington State, Oregon State, and Utah. (Yes, Utah was a "likely win" for UCLA until last week, as well.) And as fun as California's story has been so far, Stanford is still going to be the favorite in that one, too.

Still, trips to Oregon and UCLA dampen the chances of a Stanford threepeat.

In the South, likely wins for Arizona (Colorado, Wazzu, and ASU at home) and UCLA (USC at home, Colorado and Washington on the road) should allow those teams to separate from the pack a bit. But if there's any sort of regression-to-the-mean for Arizona after its Oregon upset, the division could tip in UCLA's favor. And if UCLA refuses to get its act together, then the division becomes a five-way tossup.

Slog Level Delta

There is a better than 80 percent chance that the North winner will lose at least one more game this year and finish either 6-3 or 7-2, while there's a three-in-four chance that the South winner finishes 7-2 or 8-1. Oregon and Stanford are still the highest-ranked teams in the conference (per F/+), so a 6-3 North champion would quite possibly be the favorite to win at Levi's Stadium on December 5.

Odds of Pac-12 conference records

In that one-in-five chance that the South winner ends up 6-3, said South winner could be almost literally anybody. It won't be Colorado, but each of the other five South teams has at least a 12.7 percent chance of finishing 6-3, and four others are between 25 and 33 percent. Meanwhile, if the bar falls to 6-3 in the North, crazy California's odds improve, but Oregon and Stanford are still the clear favorites.

This week

Last week, we took a look at the SEC West on the eve of an incredible Week 6 slate. The Week 7 Pac-12 lineup isn't quite as significant, but it has two games in particular that will determine a lot moving forward.

• Friday
  Washington State at No. 25 Stanford (9:00 p.m. ET, ESPN). Win probability: Stanford 82%.
• Saturday
  No. 12 Oregon at No. 18 UCLA (3:30 p.m. ET, Fox). Win probability: UCLA 57%.
  Washington at California (6:00 p.m. ET, Pac-12 Network). Win probability: California 79%.
  USC at No. 10 Arizona (10:30 p.m. ET, ESPN2). Win probability: Arizona 66%.

If Oregon beats UCLA, the odds of a 6-3 North champion go down significantly, and Oregon's own odds of winning the division improve quite a bit.

Odds of going 7-2 or better

For better and worse, Oregon is a better-known quantity than UCLA -- we know what ails the Ducks (a banged-up offensive line and a defense that is only decent, not elite), while the cause of the Bruins' issues seems to change from week to week -- and this one's a tossup. Home field gives UCLA a slight edge, but not a significant one.

Meanwhile, Arizona is the Pac-12's Mississippi State, undefeated but in desperate need of a home win this weekend. The Wildcats are young but seem far more stable than USC, and F/+ says they would win two of three in Tucson against the Trojans. Win, and there's a nearly two-in-three chance the Wildcats finish 7-2 or better in conference; granted, that wouldn't guarantee a South title, since UCLA could tie and might hold the tie-breaker edge with Arizona visiting the Rose Bowl on November 1. Still, it's a pretty good place to be.

Lose, however, and Arizona's South chances diminish dramatically ... at least, unless UCLA also loses.

Appendix

Here are the F/+ win probabilities for the rest of the Pac-12 slate.