clock menu more-arrow no yes

Filed under:

IOWA HITS NUMBER 1 IN KENPOM RATINGS

New, 102 comments

Not just first in the Big Ten, but first in the Big... Everything.

Jeffrey Becker-USA TODAY Sports

There's a long way to go in the season, but mark it down in your diaries or what have you for posterity: after beating Penn State by 24 on February 3, 2016, around 8:30 p.m., the Iowa Hawkeyes reached No. 1 in the nation in the KenPom.com computer rankings.

Kenpom number 1

If you're unfamiliar, Ken Pomeroy's ratings are a tempo-free, formula-driven system universally regarded as the best and most predictive available in college basketball. If you're ever curious what the Vegas spread for an upcoming game will be (and if you're a subscriber, which we at BHGP have been for years), you can check his system's projections and get a pretty solid idea.

The nuts and bolts of Pomeroy's rankings include mountains of historical data, regressions, meticulously adjusted formulas everywhere—yeah. But it comes down to three things: offensive efficiency, which is the number of points a team scores per possession, defensive efficiency (the same, but as allowed on defense), and strength of schedule (the opponents, rated by the same two prior factors). The "adjusted" factors mean they reflect the quality of foes faced, and as you can see, Iowa's foes have been generally excellent, especially on offense. This all passes the smell test, right? It should.

Iowa started out the season ranked 36th in Pomeroy's ratings, and despite a few hiccups (most notably the two clunkers in Florida and the six-point win against lowly Drake), the Hawkeyes have steadily moved up the ranks over the course of the season. The sweeps against Michigan and Purdue propelled Iowa into the top five, where it has hovered for the last three weeks until Wednesday's win. It's also worth noting that while the polls had Iowa unranked, the Hawkeyes never dropped out of KenPom's Top 20 after the Western Illinois game (the loss at ISU actually moved Iowa up from No. 18 to 13 in KenPom's ratings, which makes sense—you have to be pretty dang good to go to Ames and only lose by 1, you know?). So even then, it was a big deal that Iowa started the Big Ten season the way it did, but it wasn't the impossibility generally assumed by casual observers.

The reason this is a more predictive statistic than the RPI is that a team's adjusted efficiency can move up or down no matter whom it plays, and the Pomeroy system factors in where the games are played. Where the RPI will give you a boost for losing to a Top 10 team by 80 and penalize you for beating a sub-200 team by 50, there's no guarantee where a team's KenPom rating will go before the tipoff of any game.

Another example: if the Golden State Warriors all up and moved to the Big Sky Conference, it would be A) the most egregious NCAA violation in college basketball history and B) not be very good for the Dubs' RPI, as they'd face a parade of below-average teams. Maybe they could even things out with some high-quality non-conference foes, but who on earth would want to face the Warriors, y'know? So even with an undefeated record, their RPI would likely be somewhere in the teens. There's virtually no chance they'd be top five. The KenPom system, on the other hand, would notice this team incinerating everyone by 70 points per game* and project it as the best team in the nation, which it would be: it's the Golden State Warriors.

*They'd be coasting, obviously.

Moreover, the KenPom ratings are pretty robust, compared to polling done by people who instinctively react more to the facts of wins and losses than performance relative to expectation. On Saturday, Purdue goes to Maryland, and if the Terps win by four, they'll probably be rewarded in the polls, and Purdue will probably take a hit. But Pomeroy projects the score to be Maryland winning 71-67, so an actual result like that isn't a cause to reward Maryland or punish Purdue; it's an affirmation of their ratings coming in. So nothing changes.

What this all means is that there's not a team in America that would—or should, anyway—be favored to win on a neutral court against Iowa, based on the 2016 season's performance as a whole. That's not to say it can't happen, obviously; Iowa would be an effective coin-flip against Oklahoma and Villanova, and only a slight favorite against the rest of the Top 10. Maybe don't buy your tickets to Indianapolis quite yet. But don't be surprised if it happens, either.

Now, it's worth pointing out that historically, Iowa's KenPom rating would be good for about fourth place on average, going back as far as the data does, and no team has finished the season ranked first in the KenPom with a rating as low as Iowa's is right now. And again, this passes the smell test: we've heard all year long there's no juggernaut in college basketball this season, nobody against whom it takes a minor miracle to win.

But that's fine. The point isn't to be better than everyone in 2007 or whatever. It's to be better than everyone whom you'll actually face come March, and thus far, Iowa has earned the right to be in that conversation as much as anyone else in the nation.

There's a lot of season to go, obviously, and they don't hand out trophies for being No. 1 in some computer rankings six weeks before the NCAA Tourney begins. And Iowa's grasp of that top spot is hardly robust, nor even a reason to assume the Hawkeyes will win out (Pomeroy projects a 24-6, 15-3 regular season). And yes, Iowa's 2014 team had gotten as high as No. 4 before The Bad Things Happened, so the team is acutely aware how tenuous this position can be.

However.

There's no real reason to cap expectations for the Hawkeyes anymore; if there's going to be a gigantic target on your back like this, you might as well recognize how high the stakes are and play accordingly. Nobody's going to be sleeping on Iowa now, not for the rest of the season. But as we've said all along: Iowa doesn't need to weather the storm. Iowa is the storm. And now the numbers prove it.

Go Iowa Awesome.