Quantifying the difficulty of winning on the road

Utah a team to look out for (0:53)

Kara Lawson explains why Jakob Poeltl and Brandon Taylor can lead the Utes to a deep run in the NCAA tournament. (0:53)

We have seen the difficulty of winning on the road first-hand this season.

The No. 1 team in The Associated Press college basketball poll has lost road games seven times this season, more than in any other season in the AP poll era, which dates to 1948-49.

The same difficulties have held true for AP top-5 and top-10 teams, which have lost an unprecedented number of road games this season.

So, no surprise -- it’s harder to win on the road than at home. But how much harder?

Employing the back-end analysis used to produce ESPN’s Strength of Record, we can provide context to answer that question. What’s more, we can provide a new framework for defining what constitutes an “elite game” -- a widely discussed topic this time of the season.

How hard is it to win on the road?

Ken Pomeroy does a good job of outlining the tiers of difficulty based on opponent strength and game site, and we can take it a step further.

Strength of Record measures the difficulty of achieving each team’s record, given its schedule. The difficulty of each game is captured from the perspective of an average 90th-percentile team, meaning it’s from the perspective of the 35th-best team (or an average NCAA tournament team) in the country. Think of it as from the perspective of a good, but not great, major-conference team such as Seton Hall or Texas.

From the perspective of this team, beating the 75th-ranked team on the road is about as hard as beating the 16th-ranked team at home this season.

That sounds crazy, right? But coincidentally, Texas has played the 83rd-ranked team in BPI -- Stanford -- on the road and 16th-ranked team in BPI -- Iowa State -- at home this season. The results are strikingly similar: a two-point win at Stanford and a three-point win in overtime at home against Iowa State.

Further, beating the 25th-ranked team at home is about equivalent to beating the 100th-ranked team on the road. Would you rather play Gonzaga at home or Oklahoma State in Stillwater? According to BPI, an average tournament team would have about a 58 percent chance of winning either game.

Instead of “top 50 wins" …

We often default to a team’s win-loss record against top-50 opponents as a way to measure the strength of a team’s résumé. That’s generally a good measure of quality wins, but if those games came disproportionately at home or on the road, it could be misleading.

Like KenPom, we can split quality wins into different tiers based on the difficulty of each game. Like with any tiers, the cutoff is arbitrary, but this version accounts for game site and tells a more complete story when evaluating a team’s quality wins. Because Strength of Record evaluates every game, it is still the best way to look at a team’s entire résumé.

Let’s call an “elite win” one in which an average tournament team would have less than a 50-percent chance of winning. That equates to beating a team ranked better than 70th on the road, better than 36th at a neutral site or better than 15th at home. There are a similar number of elite wins as BPI top-50 wins this season.

Not surprisingly, the teams with the highest winning percentage in our “elite games” mirror the SOR rankings, with Kansas leading the way with a 9-3 record in such games.

Strength of Record also measures the quality of a team’s losses. Based on opponent BPI rank, the worst loss by a current AP-ranked team is Kentucky’s loss at Auburn (No. 169 in BPI). Accounting for game site, that is only the fifth-worst loss, behind three home losses and a neutral-site loss. The worst loss, by this measure, was by Miami at home versus Northeastern. But even a loss at home to Georgetown (69th in BPI) is worse than one at Auburn.

Oregon and Utah: A case study

The Pac-12 is an interesting case study when evaluating the difference between BPI top-50 wins and “elite wins.”

Oregon is 8-2 against BPI top-50 opponents, but six were at home and none of those home wins were against BPI top-20 opponents. Oregon is 3-3 in “elite games,” with one of those wins at UCLA (66th in BPI). The Ducks rank 14th in Strength of Record, four spots lower than Pac-12 rival Utah.

Utah is 6-5 against BPI top-50 opponents, but three of those losses came on the road against BPI top-25 foes (games even the best teams aren’t expected to win). Utah is also 6-5 in those “elite games,” with their five best wins coming at road or neutral sites.

Without a dominant set of teams in college basketball this season, it’s not surprising that highly ranked teams are losing on the road. It’s hard to win on the road in college basketball. This is a fact we should continue to remember as we head toward Selection Sunday.