Nylon Calculus: Why you shouldn’t worry too much about your team’s schedule

Don’t worry about your favorite team’s schedule. Research shows the NBA’s regular-season schedule isn’t going to have a major impact on the final standings.

Every August, the NBA releases its schedule for the upcoming year. It’s an exciting time as fans pour over the games to see which stretches will be particularly hard or forgiving for their team.  Everyone anticipates the Christmas Day match-ups and wonders what message the league is sending by choosing the teams that it did.

Looking at the schedule is fun! But don’t worry too much about its effect on competitive balance.

How hard would each schedule be for an “average” team?

It is by no means an original observation that the NBA schedule does not have a large effect on the final standings.  Kevin Pelton, for example, wrote about this in 2014. What I sought to do was flesh out this idea by quantifying just how hard or soft each team’s schedule was for the upcoming season. I did this by estimating the expected number of wins that a team of “average strength” would achieve if it played a particular team’s schedule.

A more complete description of the methodology follows in the next two paragraphs, but this can be skipped and you can go right to the results table if you want.

To begin, I collected the regular-season win total betting line for each team from the 2006-07 season to the 2019-20 season (excluding the lockout-shortened 2011-12 campaign). I got this data for past seasons from Basketball-Reference, which stores information collected by sportsoddshistory.com. For this season, I took the WestGate lines from this article by The Action Network, which were current as of Aug. 27.  On a slightly technical note, I next scaled the data for each season by subtracting the mean betting line of that season, so each of the preseason totals would have the same number of wins.

Then, I used a simple linear regression to predict each team’s final regular-season point differential by their betting line. After doing this, I created a simple model of the likelihood of winning an NBA game based on the point differentials of the teams playing, who is at home, and whether either one of the teams is on a back-to-back. Finally, I computed how many wins an average team would be expected to win against their schedule. To do this, I plugged in ‘0’ for the point differential of our team of interest and gave each of their opponents a projected point differential based on the particular opponent’s preseason win total betting line. I then added up the expected wins of our average strength team over their full 82 game schedule to get a projection for wins based on schedule strength alone.

The results for the upcoming 2019-20 regular season are below:

It is here that I should note a caveat with my method — I am overstating the difficulty of the Eastern Conference teams’ schedules at least a little. This is because I assess opponent strength by using the preseason win total betting line, and this in itself overrates the strength of Eastern Conference teams.

With that being said, we can still see from the table that the Eastern Conference squads have a bit more forgiving schedules on average than their Western Conference counterparts. On the whole, this difference in strength-of-schedule between teams in the two conferences is not super large.  The average Eastern Conference schedule yields 41.4 wins while the average Western Conference schedule yields 40.5 wins, a difference of only 0.9 wins.  Though, as I noted in the previous paragraph, this method is at least somewhat understating the difference in schedule strength between the two conferences. The Western Conference teams have a little bit of a rawer deal than what I calculated here.

Next, we see that the disparity between the easiest and hardest schedule in each conference is not really all that much. The Milwaukee Bucks, owners of the easiest schedule in the East, have a slate which yields about 1.8 more wins on average than the hardest schedule Eastern Conference team, the Cleveland Cavaliers. In the West, this difference is only 1.3 wins between the easiest and hardest schedule.

Moreover, the main difference in schedule strength between teams in the same conference is something within their control — how strong they are on the basketball court!  As Pelton pointed out in 2014, and others have noted as well, better teams typically have slightly easier schedules because they don’t have to play themselves. The reverse is true for weaker teams. The Bucks don’t have to guard Giannis once this season in any game that counts towards their win-loss total, though I’m sure Bucks players will have plenty of opportunities to guard their talented Greek teammate in practice.  On the other side of the spectrum, Trae Young doesn’t get the benefit of playing any games against the Hawks’ porous defense.

The effect of back-to-backs

All basketball fans (and sports fans in general for that matter) know that playing games on limited rest is disadvantageous to a team. Thus it makes sense that playing on a back-to-back has a statistically measurable negative effect on a team’s performance. My model found the back-to-back penalty to be roughly 1.7 points; that is, a team on a back-to-back plays 1.7 points worse on average than would be expected based on the point differentials of the two teams playing.

For each team, I computed what I called their schedule back-to-back advantage. This is simply the number of games in which their opponent is on the second night of a back-to-back minus the number of games in which they themselves are on a back-to-back.  What somewhat surprised me is that there is, in fact, some spread in schedule back-to-back advantage, as one can see in the table below.

Before Wizards and Magic fans rue their misfortune and Knicks and Bulls fans celebrate their luck, however, it should be noted that the disadvantages or advantages of playing more or fewer back-to-backs are small when considered over a full 82 game schedule. With some basic math, we can see that Wizards and Magic are expected to lose about 8.5 points (1.7 times 5) over the course of the entire regular season by virtue of playing on five more back-to-backs than their opponents.  That works out to about 0.1 points per game, a negligible disadvantage.

The bottom line: some teams will play in more back-to-backs than others, but it won’t have a large impact on the final standings. Pelton reached the same conclusion in his 2014 piece.

Is a front-loaded or back-loaded schedule advantageous?

Some of the offseason commentaries typically focus on how strong or weak a team’s schedule is in the first few weeks of the season. Those squads with a tough opening stretch are talked about as needing to “survive” the beginning of their season until the games get easier. On the flip side, those teams with a tough closing slate are sometimes thought of as being at a disadvantage when it comes time for the crucial playoff push.

But does it really matter if a team’s schedule is harder in the beginning (front-loaded) or at the end (back-loaded)? I will admit my prior belief was a strong ‘no’. Everyone plays their entire schedule; whether the hard games are at the beginning or end they are all eventually played.

To test whether schedule distribution has any effect on performance, I used a similar methodology to the one I used when computing schedule strength for each team.  I considered all seasons from 2006-07 to 2018-19 (excluding the shortened 2011-12) and broke each team’s schedule in half.  I then computed the total expected wins of an average strength team in games 1 to 41 and compared this to expected wins in games 42 to 82. I subtracted the expected first-half wins from the expected second-half wins and called this Strength Difference.  A positive Strength Difference means that a team’s second-half schedule is easier and a negative Strength Difference means that their first-half schedule is easier.

Next, I found how many wins each team won above or below their preseason expectation by taking their regular-season win total and subtracting their preseason win total betting line. Finally, I took the correlation between Strength Difference and wins above preseason expectation.  Thus I am correlating how front-loaded a team’s schedule is with how much it outperforms its expectations.

This correlation was about 0.03, which is very low. This means that having a front-loaded schedule is not associated with exceeding or failing to meet preseason expectations. Teams with a particularly front-loaded schedule do not consistently win or lose more games than we expect at the beginning of the season. The same can be said for those with a back-loaded schedule.  I suppose one could argue that there is a disadvantage (or advantage) to a front-loaded/back-loaded schedule that Vegas is already pricing into their lines, but I find this doubtful.

Long story short, NBA fans can rest easy knowing that it is their team’s play, rather than their schedule, that will decide their fate in the upcoming season.