This story appears in ESPN The Magazine’s Nov. 11 College Basketball Preview Issue. Subscribe today!

FLORIDA STATE’S 7-0 start put the Noles in prime position for one of four spots in the College Football Playoff. Then again, the Football Power Index (FPI), ESPN Stats & Information’s predictive measure of team strength, says FSU would be a ‘dog against four SEC teams on a neutral field. What’s a selection committee to do? At a recent mock exercise, presided over by committee chair Jeff Long, we were given a few stats that might be used to make the final picks. In the spirit of debate, here are some metrics we like more.

To Determine: OFFENSIVE/DEFENSIVE STRENGTH
The Committee Might Use This Stat: YPG/PPG, YPG/PPG ALLOWED
But Should Use This Stat Instead: OFFENSIVE/DEFENSIVE EFFICIENCY — Points each unit adds to scoring margin, adjusted for opponents.
Here’s Why: Holy Wazzu! The Cougars have racked up 535.8 ypg, No. 7 in the FBS. So obviously they outgun Georgia, with its 46th-ranked 437.1 ypg. Actually, offensive efficiency tells a truer tale. The Dawgs outpace the Cougs in points per drive (3.1 to 2.6), have four turnovers to WSU’s 14 and have faced twice as many of the 30 most efficient D’s (four). Oh, and Western Kentucky is fourth in points per game but barely cracks the top 25 in efficiency. So ppg? Not the best stat. Similar logic applies on D. Mississippi State is 92nd and tied for 29th in ypg and ppg allowed but has the 11th-most-efficient unit in the FBS.

To Determine: QUALITY WINS
The Committee Might Use This Stat: OPPONENT’S RECORD
But Should Use This Stat Instead: OPPONENTS’ FPI — How many points an opponent is above or below the FBS average.
Here’s Why: In Week 8, Maryland hosted 5-1 Iowa, while Kansas State visited 5-1 Oklahoma. Both the Terps and the Wildcats eked out W’s … and a win is a win is a win! Only when it’s not. According to FPI, which weighs factors like site (a hostile Norman crowd) and strength of opponent (see below), an average Top 25 team had a 90 percent chance to beat the Hawkeyes at home and just a 35 percent chance against OU on the road. So you can give the Terps credit, but give the Cats more (see below). For comparison’s sake, FSU will likely face just one team, Notre Dame, in FPI’s top 15; K-State could face as many as four. 

To Determine: QUALITY OF OVERALL RECORD
The Committee Might Use This Stat: TOTAL WINS; WINS VS. TOP 25
But Should Use This Stat Instead: STRENGTH OF RECORD — Chance an average Top 25 team could match or beat a team’s record.
Here’s Why: FPI says FSU is the most likely Power 5 team to finish unblemished (29 percent chance), so how will the committee distinguish between all those one-loss or even two-loss teams? For instance, after Week 8 Notre Dame and Duke both stood at 6-1. Sure, the committee could look at each team’s record vs. Top 25 foes, but it won’t find much of a difference: Duke topped then-No. 22 Georgia Tech, ND beat then-No. 14 Stanford. Enter strength of record, which draws a much starker line. Using factors like schedule strength, ND’s 6-1 lands at No. 12 in SOR, while Duke’s misses the top 25.

To Determine: DOMINANCE
The Committee Might Use This Stat: FINAL SCORES
But Should Use This Stat Instead: GAME CONTROL — How teams control games start to finish compared with an average Top 25 team.
Here’s Why: News flash — Baylor scores a lot, putting up at least 45 points in five of its first seven games while giving up more than 21 only three times. Mississippi State? The Bulldogs score in bunches too (at least 34 points each game), but they’ve also given up at least 23 five times. Glancing at final score totals, Baylor (343-161) seems to outdo MSU (296-151). But game control knows best: Baylor hung 61 on TCU, yet the Bears had only a 2 percent win probability in the fourth quarter; the Bulldogs have led on 88 percent of their plays and grabbed a 20-point cushion in six of their first seven games. Oh, and Florida State? The Noles have by far the worst game control ranking (No. 22) among the top 10 teams in the AP poll. Second-worst: Notre Dame at No. 15.