Past Picks

NCAAF picks are 68-58, for +4.4 units of profit (assuming 1 unit risked on every pick).

Unlock Staff Picks and more...

Subscribe now to see all picks, prediction models, and articles on both BetIQ and TeamRankings.

Get access now


Boise St. at UNLV Under 60.5 -108

Lost: 64 points

Sat Dec 2 • 3:00pm ET

More info

How it wins: Boise State and UNLV combine for fewer than 61 points in the Mountain West Title Game.

Staff notes:

  • This is our top-rated playable Over/Under for Championship Weekend.
  • So far this year, playable Over/Unders rated 55% or higher have gone 35-17 (67%). 
  • The model factors in this one include regression from a higher total for both teams, based on both having high points per game and points per play in recent games.

Pick published: Dec 1 1:34pm ET, available at that time at DraftKings.

Rot# 314


Oregon -9.0 -110

Lost: 31-34

Oregon vs. Washington

Fri Dec 1 • 8:00pm ET

More info

How it wins: Oregon wins the Pac-12 Championship Game by more than 9 points.

Staff notes:

  • This is a playable model spread pick in the Pac-12 title game.
  • Washington is undefeated, and gave Oregon their only loss of the year in a 3-point game in Seattle.
  • Washington got off to a great start, but over the last eight games, is only 2-6 ATS and has won six of those games by a single score, and none by more than 10 points.
  • Oregon, since the loss at Washington, has gone 6-0 SU and 4-2 ATS, winning by an average of 26 points, and not having a single game decided by one score.

Pick published: Dec 1 1:34pm ET, available at that time at DraftKings, BetMGM.

Rot# 305

About Our Staff Picks

We created the Staff Betting Picks feature to address several opportunities to provide more value to our subscribers:

  • BetIQ and TeamRankings offer a LOT of predictions and data, but it’s not fast or easy to parse through it all. Some subscribers just want to see a short list of our top/favorite picks.
  • Many bettors enjoy reading the rationale behind a recommended pick, as opposed to blindly trusting “because the model said so” as the reason.
  • Our algorithmic models for NFL and college football make predictions for full-game point spread, over/under, and moneyline bets. However, there’s a lot more to bet on than that.
  • Although bet size limits tend to be lower for markets like props and futures, those types of markets sometimes offer some of the biggest edges.
  • Our model predictions often change as they digest new data such as betting line movement and new game results. That approach has a lot of benefits, because the predictions shown always reflect the most up-to-date data we have. However, some subscribers just want to see a pick that doesn’t change.
  • As we do research on teams and players, sometimes we see a situational or one-time angle on a bet that we are confident provides expected value, and that angle may not be something that our models are well trained to pick up. Models typically need a lot of historical data to work well, and deep historical data simply doesn’t exist for situations that are less common (e.g. quirky injuries or weather or another more creative angle).

Some (and potentially the majority) of our Staff Picks will be drawn from top-rated model picks, but we’ll explain the data angle(s) that our models are likely seeing. Other Staff Picks may not be even be favored by our models, but we’re making a judgment call to overrule them.

Finally, some Staff Picks will be bets like player props and futures that our game models don’t currently cover, or more market-based value opportunities that we see (e.g. an off-market line offered by a particular sportsbook).

For each pick we make, we will note the sportsbook that offered it, and the associated line/payout odds at the time when we published it.