Let me give you the honest top of the column first: if you came for CFB best bets today, I have five spots and zero plays. Nothing cleared my number this morning. That is not modesty, that is arithmetic.

Week 2 rearranged the sport. Texas went into Austin against Ohio State and came all the way back to win it, the poll voters handed the Longhorns No. 1 on Sunday, and Ohio State dropped out of the top five for the first time in two seasons. Oregon tumbled too. When the top of a poll moves that violently in one weekend, the market has to redraw a dozen teams at once, and redrawing in a hurry is how soft numbers get made. That is where I go looking. Not today. But soon.

What I'm eyeing in college football after Week 2

  1. Ohio State's re-pricing. One loss, at a top-five opponent, on the road, in a comeback game, and the voters knocked them out of the top five entirely. Roster did not change. Recruiting rankings did not change. Jeremiah Smith is still the most impressive player anybody talked about coming out of the weekend. What turns this into a play: their next number coming out a full touchdown worse than what I had them at before kickoff.
  2. Texas wearing the crown. New No. 1 teams get taxed. Everybody wants a piece of the story, the tickets pile up on one side, and the number puffs. What turns this into a play: a price that asks Texas to be a better team than they were 48 hours ago. They are not. They are the same team that needed a historic comeback to win one game.
  3. Oregon and the quarterback question. The draft chatter out of Week 2 is openly asking whether Dante Moore's play is a worry, and the Ducks slid in the poll. That is a tell worth chasing, not a conclusion. What turns this into a play: a second straight week of the same problem showing up on the field, priced as if it were a one-week blip.
  4. Notre Dame's projection versus their price. The model people have the Irish sitting pretty in the updated playoff projection. Futures markets are slow to catch a team that is not attached to a conference title path. What turns this into a play: a playoff number that still reflects Week 0 doubt.
  5. FPI arguing with the humans. The computer top 25 came out swinging and it does not agree with the voters. Disagreement between the poll and the model is the cleanest early-season signal there is, because the number usually splits the difference and one side is wrong. What turns this into a play: a game where the poll gap and the model gap are more than a field goal apart.

The one I lean on

The Ohio State overcorrection. Not a play, a lean, and I want to be clear about the difference. One road loss in a comeback game does not cost a team a touchdown of true quality, and the poll just said it did. Checked the almanac: 2014 Ohio State lost at home in Week 2 to Virginia Tech and finished the season as the first College Football Playoff champion. September panic has a long, embarrassing history in this sport.

One more thread I like: the loudest voice on the big broadcast spent Saturday calling weak nonconference scheduling a joke, and he is right on the merits. It also has a practical edge for us later. Cupcake games distort every rating system in September, which means the ratings that feed the numbers are carrying a lot of fake information right now.

What I'm watching next is simple: the first number posted on Ohio State's next game. That is the one that tells me whether the market panicked with the voters or held its ground.

Keep it fun, keep it 21-plus where it is legal, and remember variance does not care how good the read was. You know where to find me.