Betting NBA Teams on a Back-to-Back

For quite a few years now, NBA teams playing a back-to-back has been a well-discussed angle for situational handicapping. A rested team is likely to outperform a less-rested team, so it’s logically sound and there are certainly profits to be made if handicapped correctly. It is something I consider in my models, and should certainly be part of your research if it isn’t already.

However, the NBA has gradually reduced these situations over the past few years. The league realized the negative effects of its grueling 82-game schedule that included back-to-backs, three games in four nights, and up until this season, the dreaded four games in five nights. These stretches caused a number of issues, such as heightened risk for player injury, lower quality games, and relatedly, star players sitting out games to rest and recover.

To help reduce these instances, the NBA season started two weeks early this year – throwing off fans and handicappers alike, myself included. The added two weeks eliminated the need for four-in-fives as noted above, which is the first time in the NBA’s 72-year history. This also decreased the average amount of back-to-back’s per team from 16.3 in the 2016-17 season to 14.4 this season – an all-time low for the third straight season.

The table below shows the downward trend for these situations on the average NBA schedule:

SEASONAverage B2Bs Average 3-IN-4sAverage 4-IN-5s
2012-1319.1312.572.53
2013-1418.6712.702.10
2014-1519.2613.032.30
2015-1617.7712.400.90
2016-1716.2711.970.67
2017-1814.379.470.00

AVG B2Bs = All games when a team played the day before. This includes instances where a team plays a third game in four nights (i.e. game/off-day/game/game).

AVG 3-IN-4s = All games after a day off, but the team played a back-to-back prior (i.e. game/game/off-day/game).

AVG 4-IN-5s = Since teams cannot play three days straight, this counts instances when teams have two back-to-back sets in five days (i.e. game/game/off-day/game/game).

When this season’s schedule was announced, I was curious to see how this would affect NBA betting and what bettors’ reactions would be. Naturally, this had to disappoint situational handicappers who capitalized on these situations, as this took away many “bad spots” for teams that the general public typically wouldn’t consider. Furthermore, this likely reduced the opportunity to use insider information about star players sitting out certain games for well-connected sharp bettors.

As someone that relies on models for my own betting though, I welcomed it. This scheduling trend will hopefully lead to less variance, which should theoretically lead to my models providing more accurate projections.

With that said, I was still curious to see if the betting market was correctly pricing these presumably tired teams. With this angle getting more and more attention, along with a generally more informed public, my hypothesis was that back-to-backs are more than priced in at this point, and possibly even over-accounted for. This in turn would create value on the less rested team, which certainly intrigued me and inspired me to write this article.

First, let’s see how NBA teams have fared overall with at least one day of rest and when playing a back-to-back. For simplicity’s sake, I’m just going to look at back-to-backs as a whole for this article.

SEASONWin-Loss with RestWin-Loss on B2B
2012-13975-911 (51.7%)254-318 (44.4%)
2013-14995-903 (52.4%)235-327 (41.8)
2014-15963-922 (51.1%)267-308 (46.4%)
2015-16997-931 (51.7%)233-299 (43.8%)
2016-171025-940 (52.2%)205-288 (41.6%)
2017-18*570-535 (51.6%)103-138 (42.7%)

*Through January 20, 2018

These results shouldn’t be surprising. A rested team has an obvious advantage over a team playing two games in as many nights when simply looking at wins and losses, and this advantage has been quite consistent over the past 5½ seasons. On average, rested teams have won 51.8% of their games, and a team on a back-to-back has won 43.6% of their games. If we take out the 2014-15 outlier for back-to-backs (46.4% win percentage), the average drops to 42.9%.

Now, there are additional situations you can add to this analysis, such as playing two back-to-back road games, already being on a long road trip, playing three games in four nights, distance traveled, and changing time zones. If you’re a situational handicapper, these are all worth considering. ESPN has even started covering the worst of these situations, which also led me to my hypothesis. If ESPN is covering a sports betting angle, the value is likely gone.

As I said above though, we’ll stick to the basics here. Let’s now see how teams fared against the spread with rest and playing the second game of a back-to-back:

SEASONATS with RestATS on B2B
2012-13922-927 (49.9%)285-280 (50.4%)
2013-14934-923 (50.3%)272-283 (49.0%)
2014-15907-936 (49.2%)293-268 (52.2%)
2015-16940-948 (49.8%)268-260 (50.7%)
2016-17972-972 (50.0%)238-250 (48.8%)
2017-18*550-541 (50.4%)110-122 (47.2%)

*Through January 20, 2018

Well according to the data I gathered above, my hypothesis was wrong! There appears to be value fading teams on a back-to-back since 2016, albeit small. Although I may have been trying to be a shade too contrarian, we may have come across another interesting trend. Let’s put the schedule situations and B2B ATS tables together:

SEASONAverage B2Bs Average 3-IN-4sAverage 4-IN-5sATS on B2B
2012-1319.1312.572.53285-280 (50.4%)
2013-1418.6712.702.10272-283 (49.0%)
2014-1519.2613.032.30293-268 (52.2%)
2015-1617.7712.400.90268-260 (50.7%)
2016-1716.2711.970.67238-250 (48.8%)
2017-1814.379.470.00110-122 (47.2%)

Can you see it? The amount of back-to-backs peak in the 2014-15 season, as does the ATS performance for teams playing in this situation. It’s not a huge sample, but there appears to have been some value betting ON teams playing a back-to-back that season. My initial hypothesis may have been correct if I wrote this three years ago.

But since then, the NBA has made a strong effort to reduce these situations as noted earlier, which can be seen in the table. Oddly enough, the ATS performance for teams playing a back-to-back has also declined in this time frame. I would have expected the opposite with a slightly less grueling schedule, but it makes sense when you think about it.

With less back-to-backs on the schedule, these situations become more likely spots for coaches to rest key players. Not counting extraterrestrial players such as Russell Westrbrook and LeBron James who rarely take a night off, coaches not named Tom Thibodeau do their best to mix in games for otherwise healthy players to sit out depending on their age and history of injuries.

Typically, this is done on the road to lessen the fatigue of traveling and to not disappoint their home crowd. Since back-to-backs mainly occur on the road, these are ideal spots to rest players as coaches have a decent reason to do so and there is a lesser chance to win these games already (remember that 43.6 win percentage from earlier?).

For anyone that partakes in DFS (Daily Fantasy Sports) such as DraftKings or FanDuel, you probably know that announcements for players resting or sitting out due to injury don’t always happen until later in the day. This makes a bit of a guessing game of who will play in these situations, so these are not always reflected in the spread right away. Since a good chunk of money has already been taken on these games when these are announced, including some sharp action already with this information, it can be difficult for books to fully adjust. And since the NBA is such a star-driven league with only five guys on the court for each team, a player’s impact is much greater than in other team sports.

This could possibly help explain the recent dip against the spread for team’s playing a back-to-back. It could also just be short-term variance, but it makes sense and we may continue to see some value fading these spots in the future. It’s too early to say for sure, but it’s definitely worth considering.

Now, this isn’t a huge revelation at the basic level I discussed; but it’s a great starting point. I’m sure if you do more digging, such as finding back-to-backs on long road trips or identifying four-in-six or five-in-seven situations (yes, those still exist), you can likely improve the win percentage from past results.

Another thing you can do, which I am currently doing with my new NBA model based on individual stats (rather than team stats), is to learn coach and player tendencies. It’s well known that Gregg Popovich will almost automatically rest his older players like Ginobli and Parker during rough stretches of their schedule (he practically invented everything we’ve talked about in this article), but what about other coaches? Similarly, do certain players “take the night off” in certain spots and not give their maximum effort? These are all things the experienced situational handicapper knows very well.

As is the theme of my blog and website, I’ll stick to the basics today and allow you to digest what we covered. If you’re looking to dive deeper into this topic and possibly use this in your handicapping, a great resource I have used over the years is NBAstuffer.com*. The site offers plenty of information, articles and datasets to help you learn more about the effect of rest and NBA analytics in general.

*This is not a referral link and am not getting paid to include this. I genuinely use this site for my own handicapping and believe the content is fantastic.

That’s all I have for this week. If you enjoyed this article and haven’t already subscribed to my newsletter, you can do so here to have my next post sent straight to your inbox! And if you’re looking to take the next step in your sports betting journey with a trusted guide, check out the details for my picks service – Fast Break MVP. You’ll get access to all my models’ picks for NBA, WNBA, MLB, NCAA Football and any other sports I expand to in the future. Learn more here!

Thanks for reading!

Leave a Reply

Your email address will not be published. Required fields are marked *