Everybody Plays Better At Home

Jun 13, 2016; Oakland, CA, USA; Golden State Warriors forward Harrison Barnes (40) and Cleveland Cavaliers center Tristan Thompson (13) go after a loose ball during the fourth quarter in game five of the NBA Finals at Oracle Arena. Mandatory Credit: Bob Donnan-USA TODAY Sports
Jun 13, 2016; Oakland, CA, USA; Golden State Warriors forward Harrison Barnes (40) and Cleveland Cavaliers center Tristan Thompson (13) go after a loose ball during the fourth quarter in game five of the NBA Finals at Oracle Arena. Mandatory Credit: Bob Donnan-USA TODAY Sports /
facebooktwitterreddit

“Role players play better at home”.

You hear it on podcasts. You read it in articles. Once a team gets to play on its home court, all the role players will step up and play better. It fits our idea of the superstar playing well, even though the crowd boos and sneers at him, but the team fails because the role players are not matching the fire and intensity that the star brings. Hence the hunger of playing Game 7’s at home.

Now this theory is all well and good. But here at The Nylon Calculus we exist also to check adages to see if they are actually observable in practice. And while It is true that role players play better at home and worse on the road, so too do the stars.

Examining the last 12 years’ worth of play, I have found little if any evidence showing much difference between role players and stars in terms of home/road play.  If role players play significantly worse, relatively, on the road than do stars, we would expect role players to have a higher percentage of their good games at home than the stars. If the stars are less affected by being away, then they should have more of their good games away than the role players.

To test this, I looked at game logs in the playoffs for the last 12 years, this season included. For each player in each season I found the median of their game performances measured by DRE. A good game is all the games for that player that season above the median of DRE. A bad game is a game below the median of DRE for that player.

If role players play better at home and more so than the stars, we would expect more of the good games from the role players to be at home. As a simple way of classifying stars, I used a simple definition: Stars should play more than 25 minutes per game and have a usage rate above 22%. That gives around 2 stars per team. I did consider using a broader definition, but I was satisfied with the group of players this definition yielded.

Rplot215
Rplot215 /

The plot above show in each season how many percentages of the good games for stars and role players that were played on their home-court.

As we can see, there are not that big of a difference between stars and role players. The value fluctuates from year to year, but there are no indication that role players have a higher percentage of their good games at home.

In fact, if all the years are pooled together, the stars have 54.4% of their good games at home and the role players have 52.3% of their good games at home.

Why do we say that role players play better at home and not that everybody plays better at home Well, he stars are stars for a reason, so even some of their bad games would be great games for a role player. The two groups have different levels of ability. When role players have bad games, it looks really bad. So the comparison is not really fair.

My conclusion is that yes, role players are better at home. But then, so is everyone else.