Investigation of the effect of circadian rhythm on the performances of NBA teams

ABSTRACT

Professional athletes competing in the NBA are frequently exposed to time-zone-shifting travels. These time zone changes may cause circadian rhythm (CR) phase shifts and these shifts affect sportive performance. The aim of this study was to investigate the effects of CR phase shifts on the performance of NBA teams. 25016 regular season games across 21 consecutive seasons were included in the CR phase shift calculations. To examine the CR phase shift effect on team performance, teams were divided into three groups regarding Coordinated Universal Time (UTC): the same internal UTC as the local UTC (LS); the internal UTC ahead of the local UTC (LA); and the internal UTC behind the local UTC (LB). With a different approach, teams were divided into another three categories: the same internal UTC as its opponent’s internal UTC (OS); the internal UTC ahead of its opponent’s internal UTC (OA); and the internal UTC behind its opponent’s internal UTC (OB). 24985 game data were used to compare these groups in terms of 25 variables. Statistical analyses were conducted separately for home and away teams. For home games, it was found that LA and OA are the most and LB is the least successful group in winning and scoring performances. For away games, it was determined that LS is the most advantageous group with the best winning percentage. These results revealed that teams from more west may have a CR advantage in regular season home games. However, it is thought that the performance of away teams depends more on travel fatigue than CR phase shifts.

KEYWORDS:

• Circadian rhythm
• diurnal rhythm
• basketball
• NBA
• sports analytics
• sports performance
• basketball performance
• jet-lag

Introduction

The biological changes (such as body temperature and hormone release) that occur rhythmically in the human body in 24-hour periods are defined as circadian rhythms (CR; Özdalyan et al. 2021; Punduk et al. 2005). In addition to these physiological processes, mental and physical performance, as well as human’s psychological states, also vary during the day and are affected by CR (Chtourou et al. 2012; Özdalyan et al. 2021; Punduk et al. 2005). For instance, anaerobic performance peaks in the late afternoon (Chtourou et al. 2012). Therefore, disruptions in CR synchronization lead to changes in sportive performance (Roy and Forest 2018; Song et al. 2017).

Professional athletes travel frequently for competition, and some of these travels require time zone change. For example, the mainland of the United States of America has five different time zones (Pacific, Mountain, Arizona Mountain, Central, and Eastern), and professional athletes competing in this country are frequently exposed to this type of time-zone-shifting travels during a competition season (Huyghe et al. 2018). These time zone changes in a short period of time cause CR phase shifts and these phase shifts affect sportive performance (Thun et al. 2015). For this effect to disappear, the internal clock needs to fully adapt to the real (local) time, which requires an adaptation time of approximately 24 hours per one-hour time zone change (Leota et al. 2022).

The National Basketball Association (NBA) league, which is considered to be the most competitive basketball league in the world, is a league where North American teams compete. In recent years the competition has been among 30 teams in the league, but throughout history, there have been many changes in the total number and the home cities of the teams. Throughout an NBA regular season, each team plays a total of 82 games (for various reasons, in a few seasons less than 82 games were played) in 174 days, approximately. In addition to this strict game schedule, circadian phase shifts are frequently observed because each team plays at, and as a result, travels to all the abovementioned five different time zones (Dehesa et al. 2019; Paulauskas et al. 2018).

In the literature, there are few studies on the circadian advantages/disadvantages of away teams in the NBA, which are conflicting at some points (Charest et al. 2021; Cook et al. 2022; Glinski and Chandy 2022; Hasbany et al. 2023; Leota et al. 2022; McHill and Chinoy 2020; Nutting and Price 2017; Pradhan et al. 2022; Roy and Forest 2018). Furthermore, most of these studies assumed that home teams never experience a circadian phase shift and that their internal biological clock is fully synchronized with local time (Hasbany et al. 2023; McHill and Chinoy 2020; Nutting and Price 2017; Pradhan et al. 2022; Roy and Forest 2018). However, these phase shifts can be observed not only after the teams’ away trips but also after their return trips to their home cities. During the CR phase shift determination process, most of the studies made the calculations by only considering the time zone difference between the home city and the opponent’s city (Hasbany et al. 2023; McHill and Chinoy 2020; Nutting and Price 2017; Pradhan et al. 2022; Roy and Forest 2018); and some of them made the determination only from the perspective of two consecutive games (Charest et al. 2021; Glinski and Chandy 2022; Leota et al. 2022). Yet, those approaches have the potential for miscalculation as described in the methods section. In addition, CR adaptation process was not taken into consideration in most of the studies (Charest et al. 2021; Glinski and Chandy 2022; Hasbany et al. 2023; McHill and Chinoy 2020; Nutting and Price 2017; Pradhan et al. 2022; Roy and Forest 2018). Finally, the fact that most of these researches did not examine home teams in terms of circadian advantage/disadvantage stands out as another important gap in the literature (Hasbany et al. 2023; McHill and Chinoy 2020; Nutting and Price 2017; Pradhan et al. 2022; Roy and Forest 2018).

The aim of this study was to investigate the effects of CR phase shifts on the performance of NBA teams. In addition, it is aimed to address the methodological shortcomings (summarized in Supplemental Material 1) of the few studies in the literature (Charest et al. 2021; Cook et al. 2022; Glinski and Chandy 2022; Hasbany et al. 2023; Leota et al. 2022; McHill and Chinoy 2020; Nutting and Price 2017; Pradhan et al. 2022; Roy and Forest 2018) and to explain the circadian advantages/disadvantages of NBA teams more clearly.

Methods

Data sample

For this study, first of all, the teams and their home cities that competed in 21 NBA seasons between 2000–2001 and 2020–2021 were determined (Table 1). All the regular season games played during this period were included in this study and data from 25,104 games was acquired from an open-access website (Sports Reference 2023) by using especially beautifulsoup package (Richardson 2007) contained in Python programming languages (Van Rossum and Drake 2009). The dataset provided the date, location, home team, away team, number of overtime, game result (win or loss; W/L) and 24 in-game team performance statistics (given in Table 2) for each game. Due to the number of analyzed metrics, the results of eight of these 25 variables (noted in Table 2) were given in the main text and the results of others were only given in the tables and supplemental materials. In addition, 88 games played during the bubble system implemented due to the COVID-19 pandemic were excluded from the study, as the teams did not travel during this period and played all their games in Orlando. Therefore, the remaining 25,016 games were included in the study. For all 1,531 games that went to overtime, the eligible variables (P, OP, PD, FGM, FGA, FG3M, FG3A, FG2M, FG2A, FTM, FTA, OR, DR, TR, AS, ST, TO, BL, and PF) were normalized to 48 minutes. The ethics committee approval for this study was obtained from Dokuz Eylul University Research Ethics Committee on 27.05.2021 with the decision number 2021/16–32.

Table 1. The time zones and the DST characteristics of the NBA teams’ home cities (for the 21 seasons between 2000–2021 years).

UTC	DST	Teams
Pacific UTC-8	In use	Golden State Warriors – Los Angeles Lakers – Los Angeles Clippers – Portland Trail Blazers – Sacramento Kings – Seattle SuperSonics – Vancouver Grizzlies
Mountain UTC-7	In use	Denver Nuggets – Utah Jazz
Arizona Mountain UTC-7	Not in use	Phoenix Suns
Central UTC-6	In use	Chicago Bulls – Dallas Mavericks – Houston Rockets – Memphis Grizzlies – Milwaukee Bucks – Minnesota Timberwolves – New Orleans Pelicans – Oklahoma City Thunder – San Antonio Spurs – New Orleans Hornets – Oklahoma City Hornets
Eastern UTC-5	In use	Atlanta Hawks – Boston Celtics – Brooklyn Nets – Cleveland Cavaliers – Toronto Raptors – Washington Wizards – Philadelphia 76ers – Indiana Pacers – Orlando Magic – New York Knicks – Detroit Pistons – Miami Heat – Charlotte Hornets – New Jersey Nets – Charlotte Bobcats

UTC: Universal Time Coordinated, DST: Daylight Saving Time.

Table 2. The abbreviations and short definitions of the variables.

Variable	Abbreviations	Definitions
Win/Loss^#	W/L	Results of the games either win or loss
Points scored^#	P	Number of points scored
Opponent points^#	OP	Number of points scored by the opposing team
2-point field goals made*	FG2M	Number of successful 2-point shots
2-point field goals attempt*	FG2A	Number of 2-point shot attempts
3-point field goals made*	FG3M	Number of successful 3-point shots
3-point field goals attempt*	FG3A	Number of 3-point shot attempts
Points differential^#	PD	P – OP difference
Field goals made^#	FGM	Sum of FG2M + FG3M
Field goals attempt^#	FGA	Sum of FG2A + FG3A
Field goal percentage^#	FG%	The ratio of FGM/FGA
2-point percentage*	FG2%	The ratio of FG2M/FG2A
3-point percentage*	FG3%	The ratio of FG3M/FG3A
Free throw made*	FTM	Number of successful free throws
Free throw attempt*	FTA	Number of free throw attempts
Free throw percentage*	FT%	The ratio of FTM/FTA
Offensive rebounds*	OR	A rebound at the basket a team is attacking
Defensive rebounds*	DR	A rebound at the basket a team is defending
Total rebounds^#	TR	Sum of DR + OR
Assists*	AS	A pass thrown to a player who immediately scores
Steals*	ST	To intercept a pass or dribble and gain possession of the ball
Turnovers*	TO	A loss of possession of the ball
Blocks*	BL	A tipped or deflected shot belonging to opponent’s team which alters its flight so the shot misses
Personal fouls*	PF	Number of fouls committed
Assists/Turnover ratio*	AS/TO	Ratio of AS/TO

^#: Results of the variable were given in the main text, *: Results of the variable were given only in the tables and supplemental materials.

Calculation of CR phase shift

To calculate the CR phase shifts of the teams, the time zones of the cities where all games were played were identified. Then, considering the Daylight-Saving Time dates, the active Universal Time Coordinated (UTC) information of these time zones was determined. Finally, the CR phase shifts (the existence, the direction, and the amount of the shift) of each team during each game were calculated by writing code in R programming language (R Core Team 2023). For this procedure, the method used by Leota et al. (2022) was utilized. According to this method, teams were assumed to travel to the city of the next opponent immediately after a game (on the same day). It was presumed that the next day of the travel the teams were exposed to a CR phase shift in the east or west direction by the amount of time zone change, and one unit CR adaptation occurred for each following day spent in the new time zone. A few examples explaining the CR phase shifts and CR adaptations by considering only one travel are shown in Table 3. As both the examples in this table and the study by Leota et al. (2022) showed, CR adaptation is not fully achieved for every game. Therefore, in this study, calculations were made considering that when a game is played under CR phase shift, this will affect the CR phase shift and CR adaptation of the next game. For this purpose, this process was carried out by the approach of season-long rolling calculation, as in Cook et al. (2022) study. These calculations were made by applying the steps described below for each team separately for each season:

1. For the first game of the season, all teams’ internal UTCs were assumed to be adapted to the local UTC.

2. For each calendar day (including non-game days) from the first game to the last game of the season for each team, the local UTC of their default city was added.

3. Finally, for each calendar day (including non-game days) the teams’ internal UTC data according to the following conditions was written:

   1. If the local UTC of a fully adapted team changed, the internal UTC of these teams was calculated as shown and described in Table 3.

   2. If the local UTC of a team that was not fully adapted changed, the following conditions were taken into account;

      1. If the local UTCs before and after the change are both greater or less than the team’s internal UTC (in the same direction), the CR adaptation was calculated to continue without any pause (an example is given in Table 4).

      2. If only one of the local UTCs before and after the change was greater or less than the team’s internal UTC (in the opposite directions), the calculation was made to ensure CR adaptation on every day except the first day of the change (the same calculation approach in step (a); also an example is given in Table 4).

Table 3. Examples of CR phase shift and CR adaptation by considering only two consecutive games.

Examples of CR Adaptation of a Team Traveling from UTC-8 to UTC-5
Example 1	Travel Day	Rest Day	Rest Day	Rest Day	Game Day
Local UTC	−8	−5	−5	−5	−5
Internal UTC	−8	−8	−7	−6	−5
CR phase shift (direction)	0 (-)	−3 (East)	−2 (East)	−1 (East)	0 (-)
Example 2	Travel Day	Game Day	-	-	-
Local UTC	−8	−5	-	-	-
Internal UTC	−8	−8	-	-	-
CR phase shift (direction)	0 (-)	−3 (East)	-	-	-
Example of CR Adaptation of a Team Traveling from UTC-4 to UTC-6
Example 3	Travel Day	Rest Day	Game Day	-	-
Local UTC	−4	−6	−6	-	-
Internal UTC	−4	−4	−5	-	-
CR phase shift (direction)	0 (-)	+2 (West)	+1 (West)	-	-

CR: Circadian Rhythm, UTC: Universal Time Coordinated.

Table 4. CR phase shift and CR adaptation examples as a result of local UTC change without adaptation.

The Local UTCs before and after the change are in the same direction
Example 1	Game Day 1	Rest Day	Game Day 2	Rest Day	Game Day 3
Local UTC	−8	−6	−6	−5	−5
Internal UTC	−8	−8	−7	−6	−5
CR phase shift (direction)	0 (-)	−2 (East)	−1 (East)	−1 (East)	0 (-)
The Local UTCs before and after the change are in the opposite direction
Example 2	Game Day 1	Rest Day	Game Day 2	Rest Day	Game Day 3
Local UTC	−5	−7	−7	−4	−4
Internal UTC	−5	−5	−6	−6	−5
CR phase shift (direction)	0 (-)	+2 (West)	+1 (West)	−2 (East)	−1 (East)

CR: Circadian Rhythm, UTC: Universal Time Coordinated.

As an example, the calculated CR phase shift and CR adaptation schedule of Toronto Raptors team during 2010–2011 season is given in Supplemental Material 2.

Statistical analyses

As in most team sports, there is also a home team advantage in the NBA (Entine and Small 2008). Therefore, in this study, statistical analyses were conducted separately for home and away teams. To examine the CR phase shift effect on team performance, teams were divided into three groups: 1) teams with the same internal UTC as the local UTC (LS), 2) teams with the internal UTC ahead of the local UTC (LA), and 3) teams with the internal UTC behind the local UTC (LB). To investigate the interaction of the opponents’ CR phase shift situations, another dividing approach was applied and teams were divided into another three groups: 1) teams with the same internal UTC as its opponent’s internal UTC (OS), 2) teams with the internal UTC ahead of its opponent’s internal UTC (OA), and 3) teams with the internal UTC behind its opponent’s internal UTC (OB). All the statistical analyses were performed separately according to these two different grouping approaches. Finally, these groups were compared with each other in terms of the variables given in Table 2.

Of the 25,016 games used in the CR phase shift calculation, 31 were excluded from the statistical analyses due to unusual circumstances. Of these games, 22 were excluded because they were played in other countries of the world (NBA 2023; Sports Reference 2023), seven were excluded because at least one of the teams was subject to a CR phase shift of more than three hours, and two were excluded because they were incomplete (NBA 2023; Sports Reference 2023). Thus 24,985 game data were used to compare the groups. Since the W/L metric is a dichotomous variable chi-square test was used to analyse. One-way analysis of variance in independent groups was used for the remaining 24 variables. The chi-square test results were interpreted with the help of cross tables. For the variables that show significant ANOVA results, the Tukey HSD test (Tukey 1949) was applied as a post-hoc test to determine which group or groups the difference originated from. Finally, all the statistical analyses were performed with the R programming language (R Core Team 2023) and p values less than 0.05 were considered significant.

Results

This study determined that home teams won 14,810 (59.3%) and away teams won 10,175 (40.7%) of the 24,985 games included in the statistical analyses. The time zones of the games and the CR phase shift data of the teams (for both according to local UTC and the opponent’s internal UTC) during the games are given in Tables 5–7 respectively. Due to the number of analyzed metrics, the results of the main eight variables out of 25 (noted in Table 2) were given in the text and the results of others were only given in the tables and supplemental materials.

Table 5. Time zones that the games played in (n = 24,985).

Time Zone	n	%
Pacific UTC-8	4,557	18.24
Mountain UTC-7	1,676	6.71
Arizona Mountain UTC-7	839	3.36
Central UTC-6	7,099	28.41
Eastern UTC-5	10,814	43.28

UTC: Universal Time Coordinated.

Table 6. CR phase shift data of teams during games according to local UTC (n = 49,970).

CR Phase Shift Volume	Home Teams (n = 24,985)		Away Teams (n = 24,985)
	n	%	n	%
3 Hours Ahead	1	<0.01	3	0.01
2 Hours Ahead	174	0.70	185	0.74
1 Hour Ahead	1,254	5.02	1,898	7.60
Adapt	22,363	89.51	20,617	82.52
1 Hour Behind	1,089	4.36	2,061	8.25
2 Hours Behind	103	0.41	220	0.88
3 Hours Behind	1	<0.01	1	<0.01

CR: Circadian Rhythm, UTC: Universal Time Coordinated.

Table 7. CR phase shift data of teams during games according to the opponent’s internal UTC (n = 49,970).

CR Phase Shift Volume	Home Teams (n = 24,985)		Away Teams (n = 24,985)
	n	%	n	%
3 Hours Ahead	1	<0.01	0	0
2 Hours Ahead	405	1.62	290	1.16
1 Hour Ahead	2,658	10.64	2,357	9.43
Adapt	19,274	77.14	19,274	77.14
1 Hour Behind	2,357	9.43	2,658	10.64
2 Hours Behind	290	1.16	405	1.62
3 Hours Behind	0	0	1	<0.01

CR: Circadian Rhythm, UTC: Universal Time Coordinated.

Analyzing the effect of the CR phase shift on the performance of the home teams according to local UTC; W/L (χ²(2, N = 24,985) = 6.23, p = 0.044) and FGA (F(2,24982)=[14.719], p < 0.001) variables showed LA was the most successful group and LB was the least successful. For the P (F(2,24982)=[6.219], p = 0.002), FGM (F(2,24982)=[6.809], p = 0.001), and TR (F(2,24982)=[17.263], p < 0.001) parameters it was revealed that LB was more unsuccessful than the other two groups. And in terms of the PD parameter (F(2,24982)=[5.007], p = 0.007) it was unveiled that LA has better performance than LB. The results of these analyses for all variables can be seen in Supplemental Material 3. The post-hoc test results and chi-square values of all the variables with significant differences are given in Table 8.

Table 8. Post-hoc test (Tukey HSD) results and chi-square values for variables showing significant difference for home teams according to the local UTC (n = 24,985).

	Chi-Square Values of the Groups (χ^2)
Variable	LA	LS	LB
W/L*	1.40	0.05	4.79
	Adj. Sig. (p)
Variable	LA and LS	LA and LB	LS and LB
P	0.121	0.001	0.014
PD	0.092	0.004	0.062
FGM	0.089	<0.001	0.010
FGA	<0.001	<0.001	0.003
FG3A	0.872	0.164	0.012
FG2M	0.016	0.009	0.374
FG2A	<0.001	0.016	1.000
OR	<0.001	<0.001	0.004
DR	0.766	0.027	<0.001
TR	0.078	<0.001	<0.001
AS	0.014	<0.001	0.001
TO	0.089	0.738	0.010
AS/TO	<0.001	0.032	0.941

LA: The group with the internal UTC ahead of the local UTC, LS: The group with the same internal UTC as the local UTC, LB: The group with the internal UTC behind the local UTC, UTC: Universal Time Coordinated, W/L: Win/Loss, P: Points scored, PD: Points differential, FGM: Field goals made, FGA: Field goals attempt, FG3A: 3-point field goals attempt, FG2M: 2-point field goals made, FG2A: 2-point field goals attempt, OR: Offensive rebounds, DR: Defensive rebounds, TR: Total rebounds, AS: Assists, TO: Turnovers, AS/TO: Assists/Turnover ratio, *: Chi-Square Test Applied.

When the effect of the CR phase shift on the performance of the away teams is analysed according to local UTC, it is observed that LS’s PD (F(2,24982)=[7.358], p < 0.001) and FGM (F(2,24982)=[3.138], p = 0.043) performances are more successful than LA’s; LA’s FGA (F(2,24982)=[4.032], p = 0.018) statistics were higher than the other two groups; LA’s W/L (χ²(2, N = 24,985) = 8.61, p = 0.014) and FG% (F(2,24982)=[11.17], p < 0.001) metrics were more unsuccessful than the other two groups. The results of these analyses for all variables can be seen in Supplemental Material 4. The post-hoc test results and chi-square values of all the variables with significant differences are given in Table 9.

Table 9. Post-hoc test (Tukey HSD) results and chi-square values for variables showing significant difference for away teams according to the local UTC (n = 24,985).

	Chi-Square Values of the Groups (χ^2)
Variable	LA	LS	LB
W/L*	7.76	0.84	0.01
	Adj. Sig. (p)
Variable	LA and LS	LA and LB	LS and LB
PD	<0.001	0.273	0.227
FGM	0.033	0.192	0.972
FGA	0.030	0.021	0.562
FG%	<0.001	<0.001	0.949
FG2%	<0.001	0.017	0.992
OR	<0.001	0.001	0.513
AS	0.624	0.019	0.017
PF	0.030	0.733	0.241

LA: The group with the internal UTC ahead of the local UTC, LS: The group with the same internal UTC as the local UTC, LB: The group with the internal UTC behind the local UTC, UTC: Universal Time Coordinated, W/L: Win/Loss, PD: Points differential, FGM: Field goals made, FGA: Field goals attempt, FG%: Field goal percentage, FG2%: 2-point percentage, OR: Offensive rebounds, AS: Assists, PF: Personal fouls, *: Chi-Square Test Applied.

The results of the comparisons of the home teams according to the opponent’s internal UTC showed that OA group was outperformed OS in terms of P (F(2,24982)=[5.119], p = 0.006) and FGM (F(2,24982)=[6.282], p = 0.002) metrics. For the FGA (F(2,24982)=[9.507], p < 0.001) performance OA was more successful than other two groups; and the TR (F(2,24982)=[3.372], p = 0.034) statistics of OA was better than OB. On the other hand, the results of the same analyses conducted for the away teams revealed that OP (F(2,24982)=[5.119], p = 0.006) performance of OS group was better than OB group. The results of these analyses for all variables of home and away teams can be seen in Supplemental Material 5 and Supplemental Material 6, respectively. The post-hoc test results of home and away teams for all the variables with significant difference are given in Tables 10 and 11, respectively.

Table 10. Post-hoc test (Tukey HSD) results for variables showing significant difference for home teams according to the opponent’s internal UTC (n = 24,985).

	Adj. Sig. (p)
Variable	OA and OS	OA and OB	OS and OB
P	0.004	0.122	0.877
FGM	0.001	0.109	0.741
FGA	<0.001	<0.001	0.766
FG3M^#	0.078	0.975	0.184
FG3A	0.030	0.883	0.008
FG2M	<0.001	0.067	0.091
FG2A	<0.001	0.026	0.045
OR	<0.001	<0.001	0.306
TR	0.079	0.035	0.478
AS^#	0.202	0.942	0.106
AS/TO	0.002	0.138	0.718

OA: The group with the internal UTC ahead of the opponent’s internal UTC, OS: The group with the same internal UTC as the opponent’s internal UTC, OB: The group with the internal UTC behind the opponent’s internal UTC, UTC: Universal Time Coordinated, P: Points scored, FGM: Field goals made, FGA: Field goals attempt, FG3M: 3-point field goals made, FG3A: 3-point field goals attempt, FG2M: 2-point field goals made, FG2A: 2-point field goals attempt, OR: Offensive rebounds, TR: Total rebounds, AS: Assists, AS/TO: Assists/Turnover ratio, #: Despite the result of the variance analysis was statistically significant, the result of the post-hoc analysis was not statistically significant.

Table 11. Post-hoc test (Tukey HSD) results for variables showing significant difference for away teams according to the opponent’s internal UTC (n = 24,985).

	Adj. Sig. (p)
Variable	OA and OS	OA and OB	OS and OB
OP	0.877	0.122	0.004
FG3M^#	0.223	0.968	0.091
FG3A	0.321	0.696	0.023
FG2M	0.508	0.287	0.003
FG2A	0.223	0.542	0.004

OA: The group with the internal UTC ahead of the opponent’s internal UTC, OS: The group with the same internal UTC as the opponent’s internal UTC, OB: The group with the internal UTC behind the opponent’s internal UTC, UTC: Universal Time Coordinated, OP: Opponent points, FG3M: 3-point field goals made, FG3A: 3-point field goals attempt, FG2M: 2-point field goals made, FG2A: 2-point field goals attempt, #: Despite the result of the variance analysis was statistically significant, the result of the post-hoc analysis was not statistically significant.

Discussion

One of the most important results of this research for the home games of the NBA teams is that while a forward (traveling to the west) CR phase shift increases the performance, a backward (traveling to the east) CR phase shift decreases the performance. Another notable finding of this study is that the success of NBA teams increases when they are fully adapted to the local time for away games. Additionally, as far as our knowledge, this is the first study: 1) to examine teams’ CR phase shifts and CR adaptations according to the “immediately after the game” travel principle throughout the entire NBA season; 2) to compare the effect of the CR misalignments by considering the CR phase shifts of two opponent NBA teams relative to each other.

Home teams

When the effect of the CR phase shift on the performance of the home teams is examined, it is clearly seen that the LA group is advantageous, and the LB group is disadvantaged. The most important data revealing this result is that the LA group won a higher rate of games compared to the LB group. The other two most important results supporting this finding are that the LB group scored fewer points than the LA and LS groups and had a smaller outscores over their opponents compared to the LA group. In addition, the significant results of FGM, FGA, and TR metrics support these important results. On the other hand, home teams experiencing a backward CR phase shift cannot show any increase in performance, and they also play a less successful game in terms of almost all performance metrics. It was determined that these teams did not encounter any negativities about the defensive statistics. However, it was observed that most of the decreased performances were only related to offensive end of the game (P, FGM, and FGA). And other important decreased statistics were related to rebound performance (TR). For these reasons, it can be suggested that the home teams who will be exposed to such a CR phase shift should be mindful of these potential performance detriments when constructing game plans.

Away teams

It is clearly seen that when away teams are exposed to a CR phase shift in any direction (forward or backward), the teams are generally negatively affected by this situation, but this negative effect is much more evident in teams experiencing a forward CR phase shift. The most important finding is that the LA group wins games with a lower percentage compared to the other groups. Another important result supporting this finding is that the opponents outscore the LA group more than the LS group. In addition, some other metrics (such as FGM and FG%) support these important results. It is thought that the most important reason why this group wins the game with the lowest percentage is their low shooting success (FGM and FG%). On this occasion, it is clearly revealed why the LA group was the group that wins with the lowest percentage. In light of all the aforementioned points, it can be said that this scenario may result in decreased offensive performance; hence it can be suggested coaches should bear this in mind during the game preparation period.

Relative to opponent’s internal UTC

Best of the authors’ knowledge, this is the first study in which teams were grouped based on the internal UTCs of two opposing teams by comparing each other’s. The analyses using this grouping approach revealed that home teams are advantageous when their internal clock is ahead of their opponent’s. However, such an obvious effect of CR phase shifts on the performance of the away teams did not appear. Besides, in this section of the discussion, it should be noted that OA group of away teams played against OB group of home teams, and vice versa. And, OS groups of home and away teams were the opposing groups to each other’s.

The main result of this approach showed that OA group of home teams is the most scorer group, and naturally, OB group of the away teams is the worst in terms of OP metric. This finding was also supported by the outcomes of FGM, FGA, and TR metrics for home teams. Hence it can be said that OA group of home teams is outperforming their opponents not only in the offensive side of the game, this group is also very successful in rebounding. Due to the interpretation of the results regarding OS groups, it is seen that home teams are the worst scorers, and again naturally, away teams are the most successful in OP statistics. It is thought that low FGM and FGA performances of home teams explain the reason these teams had the least P statistics to a certain extent.

Overall

In this research, for both OA and LA groups, it can easily be said that home teams are generally the most advantageous when they experience a forward CR phase shift. This situation is called “asymmetric jet-lag hypothesis” in the literature (Leota et al. 2022). The basis of this hypothesis is that it is easier to adapt to a long day than to a short day. For example, the LB group may have difficulty sleeping at a suitable time according to local time (due to early internal clock). They may then experience sleep problems the next day and, as a result, performance problems. On the contrary, it can be said that the fact that the LA group teams are not exposed to such a process is the main reason underlying the CR advantages of these teams. Therefore, teams from western time zones (e.g., Pacific time zone teams: Golden State Warriors, Los Angeles Lakers, Los Angeles Clippers, Portland Trail Blazers, Sacramento Kings, Seattle Supersonics and Vancouver Grizzlies) are considered to have a CR advantage in regular season home games than other teams. For example, it is almost certain that the teams in the Pacific time zone will play their home games either by adapting to the local UTC (LS group) or by experiencing a forward CR phase shift (LA group). On the other hand, when the away teams were examined, any advantages/disadvantages from the perspective of OA, OS, and OB groups were barely revealed. However, for the grouping approach according to local UTC, it is seen that the LS group is more successful than the others. Teams must stay in the new time zone long enough to adapt and to be included in the LS group. There are two possibilities for this to happen: 1) There is enough time between two games: In this way, teams can go to the city of the next game earlier and have the opportunity to rest more before the game. 2) Playing consecutive games in different cities in the same time zone: Since travels without changing time zones are shorter and therefore less tiring, teams may have the opportunity to play these games in a more vigorous manner. For these reasons, it can be thought that the main reason, why the LS group is more successful than the others in the away games, is not the advantage or disadvantage of CR but the fact that the teams play the game less tired. Therefore parallel with Cook et al. (2022) study, it is thought that the performance of away teams in the NBA depends more on travel fatigue than CR phase shift.

In a few studies, it is examined the effect of CR disruptions on sportive performance for sports other than basketball (Bishop 2004; Roy and Forest 2018; Song et al. 2017). For American football and ice hockey, it was revealed that traveling westward is disadvantageous (Roy and Forest 2018). In another study conducted on netball, it was concluded that two hours of time zone change (regardless of direction) decreases points scored (Bishop 2004). As is seen, the results of these two researches are not coherent with the present study. It is thought by the authors that, these inconsistencies may be the result of the different demands of different sports, and/or because of the methodological differences between the studies. First of all, these two studies (Bishop 2004; Roy and Forest 2018) only focused on away games and accepted home teams as fully adapted to the local hour. Furthermore, in these studies, CR adaptation was not taken into consideration. The study which has a similar methodological approach to this research (Song et al. 2017), unveiled that in baseball league winning percentage of home teams is lower when they experience an eastward CR misalignment.

Lastly, success in a basketball game requires performing different technics (such as jumping, running, dribbling, defending, shooting, passing, changing direction, etc.) with various intensities (such as walking, jogging, sprinting, etc.) (Narazaki et al. 2009). For implementing such performances successfully, advanced anaerobic and aerobic capacity are needed (Narazaki et al. 2009; Ozdalyan et al. 2022). Besides these motoric abilities, mental performance is also important for success in a basketball game (Gómez et al. 2008; Jakovljević et al. 2015). For these reasons, to achieve the best performance requires a combination of those skills (Kılınç 2008; Mancha-Triguero et al. 2019; Ostojic et al. 2006; Özdalyan et al. 2022). It is obvious, with the current study design that is not achievable to state which performance gets affected and changes the game performance as a result of CR disruption dominantly. However, in a basketball game, most of the crucial technics (such as scoring, rebounding, defending, etc.) rely on anaerobic performance (Hoffman et al. 1996; Khalifa et al. 2010; Ozdalyan et al. 2022; Ziv and Lidor 2010). Even though all the abovementioned performances are being affected by CR, the effect on anaerobic capacity is one of the most evident, and anaerobic performance peaks later in the day (Chtourou et al. 2012). Therefore, the reason for the positive performance of the home teams in the LA group may be enhanced anaerobic performance due to CR effect.

Limitations

The limitation of this research is that the actual traveling schedules of the teams are not known. Since this information was not available, it was not possible to determine how long the teams stayed in which city/time zone, how much they adapted to the local UTC and what extent they were exposed to a CR phase shift with real data. Therefore, as in similar studies in the literature (Cook et al. 2022; Leota et al. 2022), we tried to predict the traveling plans and CR adaptations of the teams by following the rules determined by the researchers. Another limitation of the current study is that the games were not separated according to teams’ ability differences. Thus, the different effects of CR desynchronization on different types of games in compliance with this fact were not revealed. To unveil this matter, future studies are needed. However, since team ability fluctuates throughout a season (Cook et al. 2022), it is suggested to develop novel methods to determine the teams’ date-specific ability levels for the dates of each game. The third limitation of the study is that the games were not separated by the schedule-related dynamics, such as back-to-back games and the day difference between consecutive games. Even though these variables affect CR adaptation and also as a result affect CR disruption, they are related to the effect of travel fatigue on the performances as well. Nevertheless, separating the games according to teams’ travel fatigue exposure and rest days between consecutive games was not preferred by the authors because of the selected statistical design. Because of the statistical design of the present study, which relies on a comparative aspect instead of a regressional one, it was preferred to work on an extended number of games (all games through 21 seasons) rather than excluding a substantial amount of acquired data.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/07420528.2024.2325641.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.