by Ken Pomeroy on Friday, February 25, 2011
For ever and ever, analysts have put an emphasis on a team’s performance on the road when evaluating them. But is there such a thing as an otherwise great team being inherently bad on the road, even when taking home-court advantage into account? I don’t know the answer to that nor, I suspect, does anybody else. Oh you might think you know, but I’d like to suggest those people reconsider this idea.
Don’t get me wrong, teams that play well on the road are generally going to be very good teams. That’s why one can predict a team’s home performance pretty well based on its work on the road. (More with respect to scoring margin than W-L record, though. Last season there was correlation of 0.39 between D-I teams’ home and road winning percentage. There was a correlation of 0.54 between their home and road scoring margin. This shouldn’t be surprising since outcome is less reliable than performance.)
Last season, the top five teams in terms of scoring margin in conference road games went a combined 41-1 in conference home games. You can flip it around, too. The top five teams in terms of home scoring margin went 26-14 in conference road games, though that’s dragged down by the 3-6 mark that VCU posted on the road in CAA play despite a dominant effort at home. Should one have put more weight on VCU’s road play when evaluating their postseason chances? Perhaps they were a paper tiger, and their performance was inflated in games where a friendly crowd was cheering for them.
The only way to really test that idea is to actually look at their postseason performance. It’s a small sample and one team doesn’t come close to proving anything. However, those that want to emphasize road play don’t have help in this case. VCU lost to Old Dominion in overtime in the CAA championship game, then won five in a row to win the CBI, including two road wins.
Outliers like VCU are the interesting cases. Especially the ones who played well enough at home to suggest they should also be dominant on the road, but who ended up underachieving in hostile arenas. There are still a few games left, but we have two such examples this season. One of those is Washington. With last Saturday’s loss at Arizona, the Huskies concluded their conference road schedule at 4-5. They are unbeaten at home with remaining contests against Washington State, USC, and UCLA.
Despite the poor results away from whatever they’re calling UW’s home arena these days, Washington has actually outscored its opponents by 26 over nine games. (For reference, Ohio State has outscored its eight Big Ten road opponents by 29.) Still, averaging just a bit less than +3 points per game in road margin of victory looks odd when combined with the Huskies +20.3 average win margin at home. Even accounting for the opponents’ home-court advantage, the Huskies have turned in some stinkers as the visiting team.
Kentucky is in a worse predicament than Washington, having gone just 1-6 on the road, while winning all six SEC games at Rupp Arena. The Wildcats have been outscored by a total of nine points in their seven conference road games, while winning their home games by an average of 19.5.
It’s easy to forget now, but a team with a similar profile last season was Duke. In going 8-0 at Cameron, the Blue Devils outscored ACC opponents by an average of 19.9 points per game. On the road, Duke went 5-3 including a lopsided loss to N.C. State, who won just four other conference games. Duke outscored opponents on the road by just 3.3 points per game. In addition, they lost their only two true road games in non-conference play, including a lopsided loss to eventual first-round loser Georgetown in the middle of conference play. Duke, of course, would go 9-0 in the post-season which included wins over Baylor and Butler, who both played close to home.
The most extreme recent example of excellent home play not translating to the road was 2009 Missouri. They rolled through eight Big 12 home games unbeaten, outscoring foes by an average of 19.8 points per game. On the road they were 4-4, and slightly lucky to get four wins since they were outscored by an average of 2.9 points per game. The Tigers won the Big 12 tournament and as a three-seed, fell to UConn in the national quarterfinals.
Don’t get me wrong, based on their road play, you wouldn’t consider either Washington or Kentucky an elite team. The Wildcats shouldn’t be getting taken to the wire by Ole Miss or Arkansas. The Huskies shouldn’t get handled by Oregon State. However, there’s no shame in last-possession losses at Florida or Alabama or Arizona, and the Cats’ and Dawgs’ relatively poor road play has been compounded by a dose of bad fortune.
Every team is going to play some games above their average level and some games below their average level. If you assume those games occur at random (after you’ve factored in home-court advantage, of course), then you’ll have your occasional case like Washington or Kentucky that experiences these bad performances disproportionately on the road. Is this indicative of some fundamental flaw in both of these teams? Is there any reason to ignore stellar home performance at the expense of disappointing (and somewhat unlucky) road play?
The recent examples of Duke, VCU, and Missouri would indicate that’s a mistake. Look, I’m sure it’s true that bad road teams tend to lose early in the NCAA tournament. But it’s not because there aren’t any home games in the tournament. It’s because bad road teams also tend to also be underachievers in general.
Write off Washington and Kentucky if you must. But don’t do it because they can’t win on the road. You’re arbitrarily ignoring a significant part of their schedule, which in this case happens to be the part each team has played the best in (even when accounting for home-court advantage). If the basketball committee’s job is to select the best teams, I think they should have some empirical evidence that indicates that road games deserve more weight in identifying the best teams. I’ll keep an open mind, but until such evidence is presented, I remain skeptical.