You’re sitting on the couch, checking the scores, and you see a team like Michigan or Arizona sitting at the top of the rankings. But wait—the AP Poll says something totally different. You check your phone, and some guy on Twitter is screaming about "Adjusted Efficiency Margin" and "Tempo." Welcome to the world of KenPom.
If you've spent more than five minutes following college basketball in the last twenty years, you've heard the name. Ken Pomeroy. He’s basically the Bill James of hoops. He’s the guy who took a bunch of messy box score data and turned it into a crystal ball that predicts March Madness better than almost any human "expert" with a microphone.
Honestly, the KenPom rankings aren't just another list of who's good and who sucks. It’s a proprietary algorithm that ignores the "drama" of a win and looks at the math of how that win actually happened.
What is KenPom and Why Does It Move So Weirdly?
Most people think a ranking should just be a list of teams with the best records. If Team A is 15-1 and Team B is 12-4, Team A should be higher, right? Not in Ken’s world.
Basically, KenPom cares about efficiency. It asks a simple question: If every team played exactly 100 possessions, how many points would they score and how many would they give up? This is what we call "Tempo-Free" analysis. It levels the playing field between a team like Virginia, which plays at a snail's pace, and a team like Alabama, which runs like they’re trying to catch a flight.
The Big Three Metrics
- Adjusted Offensive Efficiency (AdjO): This is the number of points a team would score per 100 possessions against an "average" Division I defense.
- Adjusted Defensive Efficiency (AdjD): The number of points a team would allow per 100 possessions against an "average" Division I offense.
- Adjusted Efficiency Margin (AdjEM): This is the big one. It’s just AdjO minus AdjD. It tells you how much better (or worse) a team is than the average squad.
In 2026, we’re seeing teams like Michigan and Iowa State dominate these metrics because they aren't just winning—they’re dismantling people. If you beat a bad team by 30, KenPom loves you. If the AP Poll sees you beat a bad team by 2 in overtime, they might drop you five spots. KenPom just sees the efficiency dip and adjusts accordingly.
Why the AP Poll Often Gets It Wrong
The AP Poll is a "clerical exercise," as Pomeroy himself has sort of hinted at before. It’s a group of 60 or so journalists who mostly look at the win-loss column. If a Top 10 team loses on the road to a Top 15 team by one point at the buzzer, the AP voters almost always drop them.
KenPom doesn't care about the buzzer-beater.
If you played a great game and lost on a fluke shot, your KenPom ranking might actually go up. Why? Because the data showed you played like a Top 5 team, even if the scoreboard didn't go your way. It’s a "predictive" model, not a "resume" model. It’s trying to tell you who will win the next game, not who deserved to win the last one.
Take the 2025-2026 season so far. We had a situation where St. John's was ranked No. 5 in the preseason AP Poll, but KenPom had them down at No. 16. After a month of basketball, St. John's fell to No. 23 in the AP. The computer saw the flaws before the humans did. It’s why Vegas oddsmakers use KenPom more than they use the AP Poll.
The Secret Sauce: Adjusted Tempo and Luck
Have you ever noticed the "Luck" column on the KenPom site? It’s hilarious and frustrating for fans. It’s a measure of the difference between a team’s actual winning percentage and their expected winning percentage based on their efficiency.
If a team is winning a lot of close games, their "Luck" rating will be high. KenPom usually sees this as a red flag. He figures that over time, luck runs out. On the flip side, if a team is 10-8 but has a "Luck" rating near the bottom of the 360+ Division I teams, they are probably much better than their record suggests.
Then there’s Adjusted Tempo (AdjT). This is just the number of possessions a team has per 40 minutes. Some coaches, like Tony Bennett or Mick Cronin, prefer a slower game. Others want a track meet. By adjusting for this, KenPom ensures that a high-scoring team isn't rewarded just because they take more shots. It’s about quality per shot, not quantity.
How to Read the 2026 Numbers
If you’re looking at the site today, you’ll see teams like Purdue and Houston consistently in the Top 5. They aren't just winning; they are efficient on both ends. To win a National Championship, history shows you usually need to be in the Top 20 for both Offense and Defense. If a team is #1 in Offense but #100 in Defense, they are almost certainly going to get upset in the second round by a mid-major that can actually guard someone.
📖 Related: NBA Trade Machine ESPN: What Most Fans Get Wrong
Actionable Steps for Using KenPom
Stop looking at the Top 25 number next to a team's name on ESPN and start looking at the "Four Factors" on KenPom’s site. If you want to actually understand why a team is winning—or why they’re about to lose—here’s what you do:
- Check the SOS (Strength of Schedule): A team might be 15-0, but if their SOS is ranked 300th, their high ranking is a total sham. They’re "padding" their stats against cupcakes.
- Look for the "Elite" Balance: Find teams that are Top 20 in both AdjO and AdjD. These are your Final Four contenders. In 2026, Michigan and UConn have been the gold standard here.
- Ignore "Luck" during March Madness: If a team is "unlucky" but efficient, they are the perfect sleeper pick for your bracket. They are due for a "regression to the mean," which is just a fancy way of saying they’re finally going to win the close ones.
- Watch the Tempo: If a fast team (high AdjT) plays a slow team, the game usually favors the team that can force the other into their style of play. KenPom's "FanMatch" feature actually predicts the final score and the number of possessions for every single game.
KenPom isn't perfect—he's admitted that preseason rankings are basically educated guesses until we get about 10 games of real data—but it’s the closest thing we have to an objective truth in a sport that is usually governed by emotion and "eye tests." Use the data to spot the frauds before they bust your bracket.