It is time to consider Memphis
The plausible deniability of the good wins, bad stats team
Does winning games matter? Last time I checked across the entire history of sports, it does. We remember winners; we celebrate them, over and over and over, with losers relegated to the dustbin. This goes for basically anything in life, but it feels more pertinent in college sports than anything else, at least in terms of US sporting options.
One of the fun features of college basketball, more than any other sport I’m aware of, is the combination of a relatively small sample size that induces lots of wild swings with the idea that 30 games is enough of a sample to tell us how good they are. Doubt the latter? Well, bro, you don’t know ball. You don’t even watch the games, bro. How else are we supposed to tell if a team is good or not unless they’re winning the games in front of them?
Such is why it’s fun to have outliers: teams that seem way higher on KenPom than they should with a weak W-L record, and teams that are much lower than their record would indicate. There are famous outliers of years past of the former you may remember:
2007-08 Arizona, a 19-14 team that ranked 19th at KenPom.
2013-14 Tennessee, a 21-12 First Four team that ranked 11th. (Sweet Sixteen!)
2020-21 Wisconsin, a 17-12 9 seed who ranked 10th.
Generally, these have easy explanations. For one, they’re all Power Five teams, and the seven teams across history who were a top 25 team despite winning less than 60% of their games was one. They play hard schedules. Pretty much everyone who gets this high in a predictive ranking despite losing games either A) ran up the score in their wins, B) kept the margin really tight in losses, or C) both. As an example, that ‘20-21 Wisconsin side played a top-10 schedule, suffered just four of their 12 losses by double digits, and ran up a 10+ point margin in 12 of their 17 wins.
The flip side of the outliers are more fun to discuss: teams that have gaudy win-loss records and are beloved by AP Poll voters, yet rank relatively poorly in the committee’s predictive metrics of choice. Sometimes, you go 25-5 against the 72nd-toughest schedule in the nation. Not bad at all! Let’s go see what you’re ranked at KenPom.
Oh.
By KenPom’s own Luck measure, Memphis isn’t the luckiest top-50 team on the site. That honor actually belongs to Oregon, who is 22-8 despite having just a +5.5 schedule-adjusted scoring margin per 100 possessions. Yet I’d personally argue for Memphis as the less-explicable team. Oregon’s scoring margin has been nuked thanks to getting blown out a few times. Memphis hasn’t lost a game by more than 14 all season, and that was to overall 1 seed Auburn on a neutral floor. What gives?
Throughout history, there’s been inexplicable teams made pretty explicable by March results. You can remember 2023-24 South Carolina, in fact, just last year: a team who ran up a fabulous W-L record yet at no point played like their record indicated. They were demolished in the first round by, of all teams, Oregon. Or 2022-23 Missouri, who won a game in the Tournament then proceeded to lose to a 15 seed.
And yet: I almost think Memphis might be the most intriguing of all of them. None of those teams had an 86% win percentage in their Quad 1 opportunities. I’m not sure any of them had three Q1-A wins at any point of a season. The explanation here is simple, yes; they don’t beat lesser competition by nearly as much as they should. But it’s also pretty complicated, involving remarkable late-game luck and good fortune. How have teams like Memphis fared in March? Well, the answer: about how you’d expect.