⭐Upside Studies: “Icing the shooter” in Basketball: The Unintended Consequences of Time-Out Management When the Game is On The Line
⭐Upside Studies: “Icing the shooter” in Basketball: The Unintended Consequences of Time-Out Management When the Game is On The Line
Title: “Icing the shooter” in basketball: The unintended consequences of time-out management when the game is on the line
Authors: Nadav Goldschmied (a),*, Mike Raphaeli (b), Elia Morgulev (c,d)
a) University of San Diego, United States
b) Nanning College for Vocational Technology, China
c) Kaye Academic College of Education, Beer-Sheba, Israel
d) The Academic College at Wingate, Wingate Institute, Israel
Abstract
While the sport of basketball is usually characterized by a non-segmented and mostly uninterrupted play, the stoppage of the time due to time-outs (TOs) is widespread when the game is on the line. In the current investigation, we studied the effects of TOs on free-throw performance when NCAA (National Collegiate Athletic Association) games were close and time was winding down (n = 99,026 combined sample). We generally found that time extension before execution undermined performance but not in the last minute of play when performance deteriorated altogether. In line, traditional icing when TOs were called by opposing coaches specifically to undermine performance in the last minute of play failed to exact the intended goal of lowering free-throw performance.
On March 30, 1987, during the Championship game of the National Collegiate Athletic Association (NCAA), the Indiana University team was trailing Syracuse University by one point with time winding down. Indiana’s coach, Bob Knight then instructed his players to foul the opponent and Derrick Coleman a star freshman player was forced to shoot 1 and 1 (i.e., only if the first free-throw [FT] is successful does the player get a second shot).
Furthermore, Knight, then “iced” Coleman by calling a time-out (TO) just before Coleman was about to execute his first attempt. After the forced hiatus, Coleman who had a sound 78% FT success during the tournament and went on to an impressive NBA career, missed badly. Indiana made the last shot in the game and went on to win the national title.
Picture: March 30, 1987, Indiana University Vs Syracuse University Championship game (NCAA)
While this anecdotal evidence is memorable as well as consequential, the current investigation attempts to answer the question whether the icing strategy as well as time-extension overall do indeed undermine FT performance as a whole and under which pressure situations. This exploration is especially important in light of the research into the Zeigarnik Effect (Denmark, 2010; Seifert & Patalano, 1991) showing improved recall for a task interrupted rather than one completed without disturbance. Coaches request TOs with growing use as the game nears its end (Saavedra et al., 2012) for numerous reasons such as to devise strategy, impede an opponent’s momentum, employ game management or provide rest time (Halldorsson, 2016).
This practice is found to be beneficial (Sampaio et al., 2013) especially in the last 5 min of play (G´omez et al., 2011). More selectively coaches ask for TOs to ice (Goldschmied et al., 2010), a maneuver to extend time just before an opposing player is about to execute a routinely practiced skill of a FT when the game is on the line. This act is supposedly exercised to induce rumination and self-doubt upon the rival. Much research on self-depletion supports the position that after exerting self-control, succeeding acts of self-control and other volitional self-processes are more likely to be unsuccessful due to failure of self-regulatory resources (Baumeister et al., 1994; Baumeister et al., 1998; Freeman & Muraven, 2010). Furthermore, Muraven and Baumeister (2000) demonstrated that for a task to undermine future self-regulatory performance, it must require the overriding of a prepotent, programmed response.
Most of the research on icing in sports was conducted in American football. First, Berry and Wood (2004) showed that kickers’ performance declined in the National Football League (NFL) after being iced before pressure kicks.
The authors classified a pressure kick as one that would create a lead or tie the game for the kicking team, within three or less minutes remaining in the game (or overtime) for the 2002 and 2003 NFL regular seasons and playoffs.
Picture: NFL kicker, Adam Vinatieri
Kickers were successful 73% of the time when kicking when not iced (101 successful kicks out of 139 attempts), while managing only 67% of the kicks after having been iced (24 successful kicks out of 38 attempts). However, their conclusions were tempered by the small sample size, especially of iced instances.
Goldschmied et al. (2010), expanding on Berry and Wood (2004), studied pressure kicks from six consecutive NFL seasons (2002–2008). They found that icing was successful, as in the 110 instances where the kicker was iced, he was able to score in 73 of the kicks (66.4%), while among the 163 not iced kicks, 131 kicks were converted (80.4%). In a subsequent analysis, these researchers demonstrated that if the coaches of the kicking team requested a TO (rather than the opponent) before the pressure kick, NFL kickers no longer faltered. Since the kickers were still subjected to time extension when their own coach initiated the TO, the researchers postulated that rumination and self-doubt may not be at the root of the effect. Goldschmied et al. postulated that having to go through the taxing pre-performance routine twice when the kickers did not know in advance that a TO was coming when it was called by the opponent (but not by own coach who alerted them beforehand) was the reason for the kickers’ self-depletion and failure.
While the act of icing gained some empirical support in American football as well as laboratory work (Freeman & Muraven, 2010), it did not in studies, which sought to explore this strategy in the game of basketball. McNair et al. (2020) utilized a very loose definition of icing the shooter by exploring how the most extensive hiatus from action, the 15 min long NBA halftime break, affected FT performance. Studying player shooting probabilities between the second and third quarters (who attempted at least three shots in both quarters); they found no overall undermining effect.
In a more targeted icing basketball research pertaining to significant pressure situations, Kozar et al. (1993) studied 1237 men’s NCAA Di- vision I games played between 1977 and 1989 and identified 350 relevant timeout situations of the last 5 min when the score difference was below 10 points.
Opposing coaches called 250 timeouts before an opponent shot FT. As the score differential and time remaining in the game declined, coaches utilized TOs more frequently to try to impact the outcome of the game. However, this strategy was not effective in hampering FT performance. When icing was attempted before a FT attempt, players made 72.7% of their shots while players attempting FT after their own coach called a timeout made 75.3% (n = 100) per- forming similarly. Overall, 73.4% of FT were successful when a timeout was called prior to the shot regardless of who requested it. This is in contrast to the 4138 FT in which no timeout had occurred before in which 68.7% of the FT were successful (this difference approached significance).
The researchers inferred that icing unintentionally allowed fatigue to subside allowing the tired players about to shoot a FT to restore their postural stability and motor control instead, overriding any potential increase in worry and rumination.
However, the operational definitions of stress situations the core assumption of pressure in their study may be compromised. For example, one would be hard pressed to accept being 10 points behind in the last 10 s as a pressure situation as the game has been already clearly lost. In addition, their sample was relatively small and thus possibly underpowered to test the hypothesis.
How then to reconcile the different outcomes emerging from the icing the shooter in basketball and icing the kicker in American football whereupon the former is showing no effect (and possibly a boost for the performer) while the latter is indicating performance interference in line with the original intent of the rival’s act. Indeed the two sports share a superficial similarity as team sports with an abrupt shift to individual performance effort when shooting FT or kicking field goals.
However, significant differences should be highlighted. First and foremost, a field goal attempt in football is performed by a specialized player (i.e., the kicker) whose sole game function is to kick. Since this is the case, the kicker awaits on the sidelines until called to perform. Contrastingly, in basketball, all players shoot FT when being fouled. Second, the FT attempt is always executed from the same distance without the opposing players being able to perform any physical obstruction. In contrast, in football the kicker performs from varying distances and is reliant on his entire team for success: the long-snapper must accurately snap the ball towards the holder, who must then catch the ball and place it precisely for the kicker while the rest of the team must block the rival team’s attempt to block the kick. In basketball, the FT shooter is entirely on his own from start to finish until the release of the shot. Third, football kicking is often performed outdoors when the direction of the wind may play a crucial role while basketball is played exclusively in closed arenas where FT shooters are not exposed to the elements.
Picture: Vikings’ NFL kicker (left), Nikola Jokic (Denver Nuggets) (Right)
Lastly, official TOs (i. e., stoppage in play not initiated by either teams in competition in order to allow TV commercials to be aired) is widespread in basketball. Media timeouts are usually scheduled at the first game stoppage under 16, 12, 8 and 4 min remaining in each half, including before FT attempts, with the last occurrence having a potential influence over the score of a close game. In football, media TOs may not come before a kick, a prerogative reserved only for the teams in competition.
Assisted by an influx of game stats brought about with technology advancement ushering the era of big data in sport (Morgulev et al., 2018), we were able to re-examine overall time-extension impact as well as deliberate icing in collegiate basketball with a substantially increased sample size of FT shots. We first conducted an extensive univariate analysis only for FT undertaken in the last 5 min when the score difference before the attempt was 5 points or less. We then further explored the last minute of play when the difference in score was 3 points or less aiming specifically at icing in the traditional sense of “aversive intent” when practiced by opposing coaches to undermine their opponent’s FT performance. Under this condition, we contend, one shot beyond the 3 point line could tip the score past the FT attempt placing significant psychological pressure on the shooter still. Lastly, we conducted a complimentary nested multivariate analysis to account for additional player related factors as well as interaction terms to possibly influence FT performance when the game was on the line
1. Method
Using web scrapping techniques, we collected data extending 13 seasons (2007–8 to 2019–20) of National College Athletic Association (NCAA) play-by- play stats. These data were obtained from ncaaRpackage of Luke Benz publicly open database at https://github.com/lbenz730/ncaahoopR. The data originated from ESPN.com. The data-set for analysis resulted in 99,026 FT from 22,224 games. ESPN is one of the leading platforms providing coverage and stats of North American professional sports leagues (e.g., basketball, baseball, hockey, football, etc.). ESPN data being widely used in sports research in general (e.g., CJ & Chakravarty, 2021) and in basketball in particular (e.g., Reich et al., 2006; Toma, 2017).
For each FT, we have documented the outcome of the shot (hit or miss). As player specific predictor variables, we tabulated the identity of the player who attempted the shot and the player’s season FT percentage. The game related predictor variables we included were: the home team, the away team, game time, score before the shot, whether a timeout was called before the shot and by whom.
Picture: Joel Embiid, Sixers (NBA)
Following the official National Basketball Association (NBA) definition of “clutch time”, we focused on shots that took place during the last 5 min of a game with a score difference of 5 points or less. Similar criteria were used in the relevant literature (Solomonov et al., 2015; Worthy et al., 2009). In order to neutralize the effect of calibration between first-second and second-third attempts (see Morgulev et al., 2020), we explored occurrences of either the first attempt in each set of FTs or a single FT which comes about when a shooter is fouled in the act of shooting but still manages to make the original shot (i.e., “And one”).
Being aware of the difference between aggregate-level and separate player analysis (see Morgulev et al., 2020; Wardrop, 1995) we tried to take the most prolific FT shooters in our sample and to examine the effect of TO on each of those players separately. However, due to the rarity of the event of interest (FT immediately after TO during the last five-minutes of the second half), only three of the players in our sample had 10 or more such attempts (collegiate career is limited to four years). Thus, individual-level analysis could not be performed reliably. Overall, we had 14,269 NCAA players in our sample.
1.1. Statistical analysis
In Table 1 we present the distribution of FT attempts and hit rate in time vs. no TO situation in 5th to last minute of the regulation time. We use Chi-square test to detect correlation between time remaining, FT attempts and hit rate. Players’ annual FT percentage also presented for attempts across the last 5 min, two-way Analysis of Variance (ANOVA) being used to detect correlation between time out, time remaining and the annual FT ability of the players who attempted the shots. In Table 2 we present FT attempts, FT hit rate and annual FT percentage of the players that attempted the shots in TO vs. no TO situation. Chi-square being used to measure the correlation between time out and hit rate, t-test being used to test differences between players’ annual FT percentage in TO vs. no TO cases.
In order to refrain from pooling together players of radically different FT ability, in Tables 3 and 4 we split the sample to better and worse FT shooters while conducting similar procedures to that presented in Table 2. Then, binary logistic analysis is used to identify correlation between TO and FT success while controlling for players’ annual FT ability and home vs. away. Finally, we conducted an extensive GEE analysis which included additional predictor variables such as the player’s experience and position as well as if the game was an elimination one (loser is knocked out) and season of play.
2. Results
First, we present in Table 1 how the variables vary across the last 5 min of the second half. We can see in Table 1 that a disproportionally large number of shots (42.30%) were attempted during the last minute (X2 = 30992.97, p = 0.000, df = 4).
Most FT subsequent to TOs (59.70%) occurred at the 4th minute to the end of the game, as media TOs are slated to be called then. Looking at the accuracy in attempts where no TO took place, we can see a significant drop during the last minute (X2 = 69.34, p = 0.000, df = 1). No such pattern is evident with FT after a TO (X2 = 2.28, p = 0.698, df = 4).
Two-way ANOVA (time vs. TO) showed that players annual FT% in no TO shots was higher as compared to shots (F = 5.998, p = 0.014, df = 1). One-way ANOVA showed that season ability of the players who attempted the shots became higher as the game progressed to its end: no TO shots (F = 271.120, p = 0.000, df = 4), TO shots (F = 5.281, p = 0.000, df = 4).
In Table 2 we compared the overall success rate in no TO vs. TO before FT for the last 5 min of play. We studied if the FT was taken at home or away game. A Chi-square test was performed to compare the hit rate in no TO vs. TO before FT In the first row we can see that the hit rate in TO shots (67.8%) is significantly lower than in shots without TO (69.5%).
In Table 3 we limited our analysis to the last-minute free-throws performance but attempted to do so separating between better FT shooters (73%> hit rate) and less proficient ones (73%< hit rate), still taking into account locale in order to obtain a more granular perspective. The detrimental effect of TO disappeared during the last minute for both types of shooters in both home and away games.
Considering the disappearance of the detrimental effect during the last minute, we present in Table 4, FT that took place during the last 5 min, excluding the last minute of play.
We observe significant decrease in overall hit rate in TO FT for NCAA basketball players. Since no such TO effect was found in last-minute (Table 3), we conclude that the overall significant difference presented (Table 2) stems from the period between the last 5 min of play until 1 min to the end of the game.
In order to examine the robustness of this effect we present in Table 5 a logistic regression with several control variables and the consequence of the FT attempt as the outcome variable.
We used Pearson correlation to affirm that there is no multicollinearity problem between the predictors. The analysis presented in Table 5 validates the detected effect in Table 4. Time (seconds) remaining until the end of the regulation time showed to be non- significant predictor for FT hit rate. Players’ annual FT% showed to be significant factor. Shooting in front of one’s home crowd was found to have a negative significant effect on success in 5th to 2nd last minutes FT. After we accounted for these variables, TO showed to have a negative significant effect on hit rate. Interaction between Home and TO showed to be non-significant, that is, the effect of TO was similar on the hit rate of home and away FT, which is also evident in Table 4.
We then focused exclusively on TO called by the teams to the exclusion of official TOs in order to assess the impact of true icing (not an act that is beyond the powers of the opposing coaches) as an undermining strategy. We limited this test to the last minute of play when the score differential was merely 3 points between the teams and thus one shot beyond the arc subsequent to the FT attempt could have at least tied the game for the team behind (in case the FT attempt was missed). The proportion of made shots when being exposed to icing (i.e., timeout by opposing coach) (n = 1428, 70.6%) as opposed to those taken following the player’s own coach calling the TO (n = 268, 67.5%) failed to reach significance, X2 (1, N = 1696) = 1.002, p = 0.32, eta = 0.024.
Another apt comparison with strategic considerations for an opposing coach contemplating icing is that of the free-throw shooting performance of players when iced vs. no TO at all. Like before, the proportion of made shots when being exposed to icing (n = 1428, 70.6%) as opposed to those with no TO (n = 23,575, 70.3%) failed to reach significance, X2 (1, N = 33,543) = 0.067, p = 0.796, eta = 0.001. Additional analysis when the shooting team was tied or behind in the last 2 min, again did not yield lower FT performance when the TO was initiated by the opposing team in comparison to no TO or TO taken by own team situations
2.1. Generalized estimating equation (GEE) supplementary analysis
We further employed a multivariate model approach to augment the previous mainly chi-squared analysis, as this approach may be a better fit with the nature of the data and variables. With GEE the interpretation of interactions and multivariate effects of player’s and game-related factors can be accounted for comprehensively. GEE accounts for the correlation between multiple observations in the dataset. Since NCAA players do not shoot individually but compete within a game setting, a traditional regression method could not be used for the risk of violating the statistical assumption of the independence of observations (Ghisletta & Spini, 2004). GEE is a conservative statistical procedure (Hardin & Hilbe, 2003) that was introduced as a method of analyzing correlated data, as it is advised not to treat observations from the same cluster as if they were independent, thus allowing for more efficient estimators of regression parameters. The main advantage of GEE is the production of reasonably accurate confidence intervals since it does not explicitly model between-cluster variation but rather estimates the within-cluster similarity of the residuals, and then uses this estimated correlation to re-evaluate the regression parameters and to calculate standard errors (Hanley et al., 2003). We used IBM SPSS version 28 to conduct the analyses.
Picture: Steph Curry, Warriors (NBA)
The current analysis was composed of 95,697 FT attempts (some observations are excluded due to missing values) nested in 22,018 games (the game with the most FT attempts included in this sample had 29 instances). As before, we incorporated only the first attempt in the FT sequence (or the one attempt in ‘And one’ situations). The overall success rate of FT in the sample was 69.4%. In the majority of FT attempts, there was no TO before the FT (91%). In the current model we included additional variables as predictors, specifically player related variables such as year of play (freshman [first year] = 13,371 instances, sophomore [second year] = 21,262 instances, junior [third year] = 30,064 instances and senior [fourth year] = 31,036 instances) as well as player position (guards = 61,200, forwards = 30,441, center = 4056). These new variables were extracted from the sportreference.com website.
Other new predictor variables which we included in the model were whether the attempt was undertaken during the culminating National NCAA Tournament, a single elimination tournament (2.2%, n = 2088) to possibly produce even more pressure on the FT shooters as missing may end the season for their team. We also incorporated the season of play (2007–2020) in the model. Lastly, we included the interaction terms of TO and the rest of the variables. Table 6 shows the results of the regression.
We used Pearson correlation to affirm that there is no multicollinearity problem between the predictors. We find similarly to the previous analyses that time extension on its own did not affect FT shooting performance but the interaction of time extension (TO) and time remaining on the game clock did. When assessing players’ related factors, only overall FT season average made a difference with the generally high performing FT shooters, also more likely to score when the game was on the line (neither experience as their year in school nor play position impacted clutch FT performance). As indicated by past research (Goldschmied et al., 2021), FT performance was stable across seasons of play
3. Discussion
Before addressing the effects of the deployment of TOs on FT performance in the NCAA, we observe in our extensive data set, trends that are similar to past findings in this field of research.
First, as close games near their end, substantially more FT are attempted (Kozar et al., 1993; Navarro et al., 2009). In our dataset, 44% (n = 43,827) FT were undertaken during the last minute of play which accounted for only a fifth of the time span we studied.
It stands to reason that late in the game when the score is close, teams that are behind foul their opponents in order to stop the game clock and attempt to challenge those ahead in the score to make FT.
Secondly, in response to this ‘come from behind’ fouling strategy, opposing teams’ try to place the ball with their superior and most experienced FT shooters to negate this effort (Gooding & Gardner, 2009; Lorenzo et al., 2019).
Lastly and in agreement with previous finding (Cao et al., 2011; Goldschmied et al., 2021; Toma, 2017) as pressure mounted, FT performance declined. It seems that the last minute of play when the game is close and FT success is seen as directly influencing who wins the game are especially psychologically debilitating for those called for the task.
Picture: Shaquille O'Neal, LA Lakers (NBA)
Demonstrating this heavy mental burden, Sherman and McCONNELL (1996) [cited in Roese, 1999] presented to participants vignettes of basketball outcomes in which serial events were manipulated. Counterfactual thinking (i.e., what if’ scenarios) was more likely to center on the alteration of later rather than earlier events (e.g., the final FT in a close game rather than FT scored earlier in the game). This finding, once again, highlights the importance of success when the game outcome is on the balance. These three general trends underscore the unique game strategy and the importance of FT shooting under pressure paramount to winning basketball games.
Focusing on the intersection of time manipulation and FT performance, the addition of time before a shooting attempt from the line seems to undermine the performer when the game was close sans the last minute of play. Time extension regardless of who initiated the TO may allow to aggravate psychological processes of a player who is now in the limelight and is required to perform individually at the service of his team. For example, Roy et al. (2016) found that players in team sports scored overall lower on reflective rumination than did non-athletes.
Being low in rumination may be generally advantageous in basketball as interactive continuous team play demands players to swiftly adjust their actions and focus to constantly shifting events, score and opponent strategies. Simply put, players need to stay in the moment and not linger on the past in order to be successful.
Yet when a TO is taken when the score is close late in the game, this unique situation allows ample time for contemplation possibly propelling unaccustomed and damaging self-reflective state-of-mind that may overwhelm the FT shooter.
Picture: Steve Kerr, Warriors (NBA)
In line, Kr¨ohler and Berti (2019) survey study found that rumination was associated with state rather than action orientation (i.e., thoughts revolve around the future or past and not around the actual demands required during competitions) and thus not likely to yield superior performance. Nevertheless, rumination does not necessarily account for all the factors involved in the deterioration in FT performance due to time extension.
Coming full circle, the aforementioned successful icing of Derrick Coleman by coach Knight can shed light through the description by Thomas Bonk of the LA Times of the scene: “Derrick Coleman took the inbounds pass after the basket and was immediately fouled by Smart. Knight called a timeout to let Coleman, a freshman, think about his one-and-one foul shots. Boeheim [Coleman’s coach], who left Coleman alone at the line and pulled the rest of his players back to the other end of the court, had no one to rebound the ball when Coleman’s first free throw hit the front end of the rim.” (Bonk, 1987). Hence feeling psychologically alone (Kotzer, 2007), being subjected to high expectations by others (Maher et al., 2020) especially those held in high esteem (Mertens et al., 2018) as one’s own coach or having an altered visual field (i.e., no teammates in front of the shooter to rebound the shot) [Nascimento et al., 2020] may be alternative reasons for failure to execute specifically under time delayed performance.
We find that a TO occurrence before executing a well-rehearsed task when the game is on-the line and nearing its end has a debilitating force. However, the effect is observed not in the traditional “aversive intent” icing attempts initiated by opposing coach as described in the opening paragraph of this manuscript but is manifested for time-extension more generally. In other words, added time can trigger a negative spiral, which, in turn, can tip the performer beyond the optimal psychological zone of performance.
Empirical support for this rationale was provided by Nomikos et al. (1968) classic study of “surprise” vs. “suspense.” When passively watching a short movie about work-related accidents, participants in the suspense group (i.e., those exposed to 20–30s of clue-furnishing scenes) reported feeling more stress and experienced elevated skin conductance levels, in comparison with participants who had not been informed of the accidents and experienced them as a surprise.
It seems that coaches of the players who are about to execute a crucial FT intuitively understand this undermining effect and call very few TOs when the game is close despite the need to strategize and coordinate team efforts in close games when time is winding down.
Picture: San Antonio Spurs (NBA)’s Time Out
On the other hand, the innocuous act of an official TO for which no team has control over, was found to have a substantial performance undermining effect, which was not necessarily anticipated since it is lacking in the deliberate act rationale associated with the icing act.
Generally, we observe a marked decline in FT performance in the last minute of play, which is shown both in superior as well as poor FT performers.
Similarly, Bar-Eli and Tractinsky (2000) demonstrated this trend most clearly in their basketball study whereupon the criticality of game possessions increased significantly in the last minute of play in comparison to the 4 min prior (as assessed by expert coaches) and with it, a marked decline in performance and decision making was observed (by a different group of expert coaches). More specifically and in line, G´omez et al. (2018) observed a drop in FT performance in the last minute of play based on playing position. These findings, as well as ours, align with Hardy’s catastrophe theory of anxiety and performance (Hardy & Parfitt, 1991).
Through the catastrophe prism, a model proposed by Goldschmied and Vira (2019) studying shooting performance following air-ball shots that miss altogether the basket could be instructive. These researchers argued that increased self-awareness concerns are caused by certain game situations one of which may be FTs as it is an individual challenge rather than team based. On its own, this challenge is not sufficient to pierce players’ psychological armor and cause ego-depletion, as players are experienced in the act. However, upon time-extension and the greater opportunity to ruminate, they may falter. This “one-two punch” psychological double-tax model was also observed with air-ball shooting. Yet, similarly to the current investigation, Goldschmied and Vira did not detect deterioration in performance in response to an opponent’s ploy, in their investigation the chanting of the opposing fans.
Similarly, under the current crisis conditions, icing may be a weak psychological ploy to exact a meaningful effect as players can identify easily a clear attempt to hurt them. This psychological attribution is not readily available to them when the act is innocuous (e.g., media TO) and can not be linked directly to the opposing side.
As such the “successful” icing of Derrick Coleman should remain in the purview of the anecdotal rather than the empirical Our extensive investigation can be challenged most generally by its archival nature as we attempt to infer motivational intent among coaches and inner psychological states of the players performing FT according to distinct situational factors.
Picture: Miami Heat (NBA)’s Time Out
As such, these data interpretations may represent speculations as to what is really happening with FT shooters’ mentalities. Since the work is non-experimental and no attempt to query the performers directly was undertaken, we cannot conclude with overriding certainty. While acknowledging the limitations that exist through interpretation of patterns found in datasets of professional sports performance, we can all witness the lively debate that such studies stimulate within the domain of psychology. Notable examples, among others, are the discourse around the question of momentum (Csapo et al., 2015; Gilovich et al., 1985; Miller & Sanjurjo, 2018; Morgulev & Avugos, 2020) and the controversy concerning the first-mover advantage that allegedly stems from pressure in penalty shoot-outs (Apesteguia & Palacios-Huerta, 2010; Kocher et al., 2012).
While these studies may fail to allow clear causal relationship and require inference, they certainly shed light on human functioning in ecologically valid settings and provide a fertile ground for theoretical discussion and subsequent experimental work.
We argue that with the assistance of past self-report (Maher et al., 2020; Wang et al., 2004) and archival work (Toma, 2017) as well as the crucial aspect of game stoppage before FT which “isolate” the shot from the on-going play, we can reasonably assume general thinking patterns that make sense under the circumstances. In regards to icing coaching decisions, a future complimentary approach, albeit less practical, might be to watch film of basketball games to try to determine why timeouts were called. Even better, and probably even less practical, would be to ask coaches after the games why each timeout was called.
More specifically the current investigation is amiss also in the following: while we collected a large sample of FT, our investigation focused solely on the first (or only) attempt undertaken from the line, we did so because past research demonstrated dependency between shooting attempts (Morgulev et al., 2020). We took this approach as it makes sense that the impact of the TO would be most observed in the first immediately following shooting attempt. Yet from the overall game perspective, all shots from the line make similar contribution to the game outcome. Second, our sample is decidedly smaller when studying the effect of traditional icing (i.e., when the TO is called by the opposing coach) than the official TOs studied (albeit greater than past studies and statistically robust) and future research should strive to collect more data to substantiate our tentative conclusions. Lastly, while our multivariate analysis did take into account NCAA players’ individual differences such as years of play, since a college career is limited in time and extends only to 4 years at the maximum, this factor was limited in range.
Similar exploration into the National Basketball Association (NBA), for example, would be superior as to whether the findings are qualified by intra (vs. inter) influences.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgements
No organizations funded this research.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi. org/10.1016/j.psychsport.2023.102440.
References
Apesteguia, J., & Palacios-Huerta, I. (2010). Psychological pressure in competitive environments: Evidence from a randomized natural experiment. The American Economic Review, 100(5), 2548–2564.
Bar-Eli, M., & Tractinsky, N. (2000). Criticality of game situations and decision making in basketball: An application of performance crisis perspective. Psychology of Sport and Exercise, 1(1), 27–39.
Baumeister, R. F., Bratslavsky, E., Muraven, M., & Tice, D. M. (1998). Ego depletion: Is the activeself a limited resource? Journal of Personality and Social Psychology, 74, 1252–1265.
Baumeister, R. F., Heatherton, T. F., & Tice, D. M. (1994). Losing control: How and why people failat self-regulation. San Diego: Academic Press. Berry, S. M., & Wood, C. (2004). A statistician reads the sports pages: The cold-foot effect. Chance, 17(4), 47–51.
Bonk, T. (1987). Hoosiers get smart when it counts : Shot at :04 lifts Indiana past Syracuse. Retrieved from https://www.latimes.com/archives/la-xpm-1987-03-31-s p-1337-story.html.
CJ, D., & Chakravarty, A. (2021). Team contingent or sport native? A bayesian analysis of home field advantage in professional soccer. Journal of Business Analytics, 4(1), 67–75.
Csapo, P., Avugos, S., Raab, M., & Bar-Eli, M. (2015). The effect of perceived streakiness on the shot-taking behaviour of basketball players. European Journal of Sport Science, 15(7), 647–654.
Denmark, F. L. (2010). Zeigarnik effect.The Corsini encyclopedia of psychology, 1–1.
Freeman, N., & Muraven, M. (2010). Don’t interrupt me! Task interruption depletes the self’slimited resources. Motivation and Emotion, 34(3), 230–241.
Ghisletta, P., & Spini, D. (2004). An introduction to generalized estimating equations and an application to assess selectivity effects in a longitudinal study on very old individuals. Journal of Educational and Behavioral Statistics, 29, 421–437. https://doi. org/10.3102/10769986029004421
Gilovich, T., Vallone, R., & Tversky, A. (1985). The hot hand in basketball: On the misperception of random sequences. Cognitive Psychology, 17(3), 295–314.
Goldschmied, N., Nankin, M., & Cafri, G. (2010). Pressure kicks in the NFL: An archival exploration into the deployment of TOs and other environmental correlates. The Sport Psychologist, 24(3), 300–312.
Goldschmied, N., Raphaeli, M., Moothart, S., & Furley, P. (2021). Free throw shooting performance under pressure: A social psychology critical review of research.
International Journal of Sport and Exercise Psychology, 1–19. Goldschmied, N., & Vira, D. (2019). The experience of ego threat in the public arena: A study of air ball shots performance in collegiate basketball. Journal of Applied Social Psychology, 49(7), 409–415.
G´omez, M.´A., Avugos, S., O˜noro, M.´A., Lorenzo, A., & Bar-Eli, M. (2018). Shaq is not alone: Free-throws in the final moments of a basketball game. Journal of Human Kinetics, 62(1), 135–144.
G´omez, M. A., Jim´enez, S., Navarro, R., Lago-Penas, C., & Sampaio, J. (2011). Effects of coaches’ timeouts on basketball teams’ offensive and defensive performances
Psychology of Sport & Exercise 68 (2023) 102440 7according to momentary differences in score and game period. European Journal of Sport Science, 11(5), 303–308.
Halldorsson, V. (2016). Coaches use of team timeouts in handball: A mixed method analysis. The Open Sports Sciences Journal, 9(1), 143–152.
Hanley, J. A., Negassa, A., Edwardes, M. D. D., & Forrester, J. E. (2003). Statistical analysis of correlated data using generalized estimating equations: An orientation.
American Journal of Epidemiology, 157, 364–375. https://doi.org/10.1093/aje/kwf215
Hardin, J. W., & Hilbe, J. M. (2003). Generalized estimating equations. Boca Raton, FL: Chapman & Hall/CRC.
Hardy, L., & Parfitt, G. (1991). A catastrophe model of anxiety and performance. British Journal of Psychology, 82(2), 163–178.
Kocher, M. G., Lenz, M. V., & Sutter, M. (2012). Psychological pressure in competitive environments: New evidence from randomized natural experiments. Management Science, 58(8), 1585–1591.
Kotzer, R. D. (2007). The social facilitation effect in basketball: Shooting free throws. The Huron University College Journal of Learning and Motivation, 45(1).
Kozar, B., Lord, R. H., Whitfield, K. E., & Mechikoff, R. A. (1993). Timeouts before free-throws: Do the statistics support the strategy? Perceptual and Motor Skills, 76(1), 47–50.
Maher, R., Marchant, D., Morris, T., & Fazel, F. (2020). Managing pressure at the free- throw line: Perceptions of elite basketball players. International Journal of Sport and
Exercise Psychology, 18(4), 420–436. McNair, B., Margolin, E., Law, M., & Ritov, Y. A. (2020). The hot hand and its effect on theNBA. arXiv preprint arXiv, 2010.15943.
Mertens, N., Boen, F., Vande Broek, G., Vansteenkiste, M., & Fransen, K. (2018). An experiment on the impact of coaches’ and athlete leaders’ competence support on athletes’ motivation and performance. Scandinavian Journal of Medicine & Science in Sports, 28(12), 2734–2750.
Miller, J. B., & Sanjurjo, A. (2018). Surprised by the hot hand fallacy? A truth in the law of small numbers. Econometrica, 86(6), 2019–2047. Morgulev, E., & Avugos, S. (2020). Beyond heuristics, biases and misperceptions: The biological foundations of momentum (hot hand). International Review of Sport and Exercise Psychology, 1–21.
You may also like:
⭐⚽ Upside League Profile: Brazilian National Soccer Team & Chat with Guilherme Passos, Sports Scientist On His Favorite Technologies, The Best Players/Coaches He Worked With.
🔥 Upside Chat: Dave Hancock, CEO, Apollo (Leading Athlete Management Systems (AMS) vendor)
🔥Upside Chat: Alexi Pianosi, Strengths and Conditioning Coach, Pittsburgh Penguins (NHL)
🔥Upside Chat: Pierre Barrieu, High Performance Director, Toronto FC (MLS)
🔮 2022 Upside Top Sports Tech Predictions (NFT/Metaverse, Sports Performance, IPOs/M&As..)