Introduction
Momentum is widely believed by the sporting community to have an influential role on the outcome of sporting events. Athletes, commentators, and fans alike often refer to ‘critical’ moments in a game that are perceived as momentum ‘starters’ or ‘shifters’ which seemingly alter the dynamics of a contest. There is little doubt the notion of momentum is inherently compelling to the sporting public and the belief in its power is widespread.
Psychological momentum is the term used by sport psychologists to describe the notion that performance is influenced by the outcome of a preceding event or events. There are two major theoretical models that support the concept of momentum in sport. They are the ‘Antecedents-Consequences Psychological Momentum Model’ and the ‘Multidimensional Model of Momentum in Sport’. These two models suggest positive and negative momentum occur because of ones response to a prior event. For example, after a successful performance/s an athlete will experience changes to cognition, physiology, and emotions, such as a rise in feelings of self-confidence, perceptions of control, motivation, satisfaction, etc, which consequently increases the chance of future success (termed positive momentum). Alternatively, after an unsuccessful performance, an athlete may experience a drop in self-confidence, perceptions of control, motivation, etc, which consequently increases the chance of future failure (termed negative momentum).
However, a contrasting model, the ‘Projected Performance Model’ disputes the actual existence/influence of psychological momentum. This model claims the ‘purple patches’ or ‘slumps’ in performance are just natural fluctuations in performance that occur over time. Athletes are unlikely to perform at the same level throughout an event/s, rather their performances are likely to vary, at times surpassing or falling short of their norms. For example, it is quite possible for a basketball player who makes on average 50% of his/her shots, to make (or miss) three or four baskets in a row (as it is likely a 50% chance event such as a series of coin tosses will produce a run of 4 heads or tails in a row from time to time). But many athletes and observers view these streaks in performance as ‘abnormal’ or ‘unnatural’ and seek causal explanations for the apparent break from the norm. Randomness is frequently disregarded while concepts such as positive and negative psychological momentum are invented in an attempt to explain the ‘abnormal’ phases of performance (e.g., making four baskets in a row), when in fact making 4 baskets in row is quite likely just a natural fluctuation within a random sequence. Koehler and Conley (2003) state “declarations of streaks are probably best viewed as a commentary on past performance rather than as prophecy about future performance” (p. 257).
Research methods investigating psychological momentum are somewhat limited because of the current difficulty in directly measuring an individual’s psychological and physiological response during actual performance. One method of researching psychological momentum has been to look for patterns in performance that would imply psychological momentum. For example, a study found that 79.1% of professional tenpin bowlers were more likely to get a strike after four successive strikes compared to after four successive non-strikes. Similar results have been found in studies involving professional horseshoe pitchers and professional pool players. These results provide behavioural evidence that ones performance is influenced by the outcome of a preceding event/performance.
Research has also shown that men and women may differ in their response to prior performance. Men were found to be more likely to experience positive momentum and women were less likely to improve their performance when losing. These findings suggest men and women react differently to performance outcome and generate different perceptions of psychological momentum.
However, there have been an equal number of studies conducted within sports such as basketball, volleyball, baseball, and golf that have found no statistical difference in ones performance after successful performances compared to after unsuccessful performances. For example, hole-by-hole scores were analysed for 35 professional golfers from rounds played in the 1997 PGA Tour season, to determine whether a score on a hole influenced the score on the following hole (e.g. is a golfer more likely to score bogey or worse after a hole where they scored bogey or worse, i.e. negative momentum).
There are a few possible reasons for the discrepancy in the psychological momentum research. It is possible that the effect of psychological momentum may be more or less prominent depending on the characteristics of each sport. For example, studies within sports that found support for psychological momentum (pool, tenpin bowling, and horseshoe pitching) are sports that may provide an environment where psychological and cognitive components have a greater impact on performance. They are all individual sports that provide competitors with ample time between performances to direct their attention towards past performance and generate perceptions of psychological momentum.
In contrast, studies which have not found support for psychological momentum within sports such as volleyball, baseball, and basketball, are sports that involve continuous performance and performance is more reactive, i.e. other competitors are involved. Therefore, more attention is required on performance and there is less time to think about prior performance and generate perceptions of psychological momentum.
However, golf would be defined as an individual sport that provides ample time for competitors to reflect on their performance, but previous research conducted for this sport did not produce supporting evidence for psychological momentum. Perhaps the results was due to the sample size (35 professional golfers) being too small to get a fair representation of the general population. Or perhaps professional sportspeople are not the best participant pool to search for the effects of psychological momentum.
All previous hot hand research on sportspeople has exclusively used professional athletes as participants. However, it could be argued the effects of psychological momentum in professional sport are minimal. Professional athletes are psychologically prepared and coached to focus on the performance at hand (e.g. take one shot at a time), and therefore should be less likely to be influenced by prior performance.
The current study looked to address these issues by replicating Clark’s (2005) study but with a much larger sample size (i.e. 3214 participants compared to 35) and amateur golfers were used in preference to professional golfers.
The hypotheses of this study were: (a) a golfer’s score on a hole is influenced by their score on the previous hole at a statistically significant level, for a greater number of golfers than would be expected by chance; (b) more male golfers will perform better after an error compared to female golfers).
Method
Participants
Hole-by-hole round data was obtained from 3214 participants who were members from 9 different golf clubs. Of these participants, 2511 were male and 703 were female.
All Participants were over the age of 18.
(The current research was heavily focused on cognition and the ability to cope with performance mistakes. As adults differ in their cognitive development and coping skills compared to juveniles, the research was limited to members over the age of 18).
All Participants were required to have played at least fifteen 18-hole rounds (with a handicap less than 36.4 for men and 40.4 for ladies) within the last two years, dated from the time the members data was received.
(Requiring a minimum of 15-rounds improved the chance of attaining enough round data so that it was representative of a participant’s normal golfing performance).
Measures
Performance
Participant’s hole-by-hole nett scores were used as the measure of performance.
(Nett scores were used in preference to gross scores because nett scores are a standardised measure of performance that takes into account a player’s handicap. As the main focus of the research was on how prior performance influences subsequent performance via perceptions of psychological momentum, it was important to utilise a standardised measure of performance (such as nett score) that is likely to produce similar perceptions of success and failure for participants. Golfers are more likely to base performance success or failure in terms of their standardised ‘handicapped’ nett score rather than their gross score).
Procedure
A software program was used to examine each participant’s hole-by-hole round data, and to calculate the:
Number of times each participant followed a nett par or better hole with a nett par or better hole.
Number of times each participant followed a nett par or better hole with a nett bogey or worse hole.
Number of times each participant followed a nett bogey or worse hole with a nett par or better hole.
Number of times each participant followed a nett bogey or worse hole with a nett bogey or worse hole.
The round calculations were conducted using the following conditions:
Starting Hole. There was no way of determining which hole a player started their round on, as this information was not recorded in the databases provided. Because the vast majority of rounds played are started from hole number one, all rounds were analysed as if the player had began their round from the first hole.
Rounds. Only 18-hole rounds were included in the analysis.
Pick-ups. For handicapping purposes in New Zealand the maximum score a player can record for a hole is a nett bogey. Therefore, any hole scored as a ‘Pick-up’ was categorised as a score of nett bogey or worse, because it is known players will only pick-up their ball and record a ‘Pick-up’ after they have taken more shots than a nett bogey.
Results
Analysis
Each participant’s hole-by-hole round data was arranged in a 2 X 2 contingency table as shown in Table 1.
Table 1
Example of 2 X 2 Contingency Table
a = the number of times a participant followed a nett par or better hole with a nett par or better hole; b = The number of times a participant followed a nett par or better hole with a nett bogey or worse hole; c = The number of times a participant followed a nett bogey or worse hole with a nett par or better hole; d = the number of times a participant followed a nett bogey or worse hole with a nett bogey or worse hole (Clark, 2005).
A chi-square statistical test was performed on each golfer’s data, to determine whether a player’s frequency to follow an over par hole (error) with a score of over par (error) on the subsequent hole significantly differed from a player’s frequency to follow an over par hole (error) with a score of par or better on the subsequent hole (non-error).
Out of 3214 participants, 286 (8.9%) were statistically more likely to follow an over par hole with a score of over par on the subsequent hole (negative momentum). Table 2 shows the results for each gender.
To produce evidence for the hot hand phenomenon a sample must find more statistically significant results than would be expected by chance alone. For example, if you conducted tests on 100 golfers using an alpha level of .05, one would expect approximately five significant results from a normal distribution. So finding 5 significant results would not be surprising and within chance expectations. However, if your analysis found 20 significant results from a sample of 100, this shows a greater number of golfers are influenced by prior performance than what would be expected by chance, thus supporting the notion of the hot hand phenomenon.
Table 2
Significant Results for Gender and Psychological Momentum Type
Males were approximately four times and significantly more likely than females to have a significant tendency to follow an over par hole with a score of par or better on the subsequent hole (negative facilitation). Women were almost twice and significantly more likely to follow an over par hole with a score of over par on the subsequent hole (negative momentum).
Discussion
The hypothesis that a golfer’s score on a hole would be influenced by their score on the previous hole, for a greater number of golfers than would be expected by chance was supported. The second hypothesis that males would have a significantly higher percentage of negative facilitators compared to women was also supported.
These findings support the ‘hot hand’ phenomenon, and contradict the research conducted by Gilovich et al. (1985), Tversky and Gilovich (1989), and Koehler and Conroy (2003). One reason for the discrepancy in results could be due to the smaller number of performance attempts Gilovich et al., Tversky and Gilovich, and Koehler and Conroy analysed for each participant. The participants in the current study averaged 923 attempts and only golfers who had a minimum of 255 performance attempts were included. Whereas in Gilovich et al. controlled shooting experiment participants took only 100 attempts each. And in their NBA free-throw study participants averaged only 114 attempts each. Also in Koehler and Conley basketball shootout (which found no evidence for the hot hand phenomenon) participants had a mean of only 49 attempts each.
Stern and Morris (1993) state that finding evidence of streak shooting requires either a large effect size or a large sample, i.e. a large number of attempts. Analysing a greater number of attempts will increase the power of a study. When Wardrop (1995) reanalysed Gilovich et al. (1985) participants free-throw data as a group (effectively increasing the power and number of attempts analysed), he found participants were significantly more likely to make their second free-throw shot after hitting their first shot compared to after missing their first shot, which supports the notion of the hot hand.
A small sample size may explain why Gilovich et al. (1985), Tversky and Gilovich (1985), and Koehler and Conroy (2003) found no evidence for the hot hand. However, Albright (1993) who analysed the hitting statistics of baseball players who had at least 500 attempts at bat still found no evidence for the hot hand.
Another possible reason for the discrepancy in results could be due to the differing characteristics of golf compared to sports such as baseball and basketball. Adams (1995) suggested the impact of psychological momentum on performance might be more or less prominent depending on the characteristics of each sport.
Adams (1995), who found support for the hot hand in the game of pool, described pool as an intensely solitary game of fine motor skills. He argued these conditions provide an environment where psychological and cognitive components have a greater impact on performance. Golf is a similar game in nature to pool as it is also intensely solitary and dominated by fine motor skills with 60% to 65% of all shots played within 100 yards of the hole (Pelz, 1999).
Clark (2005) supported Adams’ (1995) concept by claiming golf is a more suitable sport to investigate psychological momentum because of the relatively long breaks between performances. A round of golf takes approximately 4 to 5 hours, but players only spend around 2 minutes of that time actually hitting the ball (Kerr, 1999). These conditions provide a golfer with ample time to reflect on past performances and generate perceptions of psychological momentum.
In contrast, generally performance in open skilled sports such as basketball, baseball, and volleyball is continuous and performance is more reactive. The ever-changing environment requires constant attention, meaning an individual will have less cognitive resources and time to reflect on past performances, and are therefore less likely to generate and be affected by perceptions of psychological momentum.
This theory is somewhat supported by the fact the tenpin bowling, horseshoe throwing, and pool studies conducted by (Smith, 2003; Dorsey-Palmateer & Smith, 2004) that found evidence supporting the hot hand phenomenon are all individual sports, like golf, that require proactive performance.
However, Clark (2005) also investigated the hot hand phenomenon in golf, using the same method of analysis as the current study but found no supporting evidence.
Clark (2005) only found three significant results from a sample of 35 professional golfers. The likelihood of finding at least three significant results was well within chance expectations. Perhaps the inconsistency in results is due to Clark’s small sample size, as it may not have been representative of the population. Or perhaps it was because Clark analysed the performance of professional golfers, whereas the performance of amateur golfers were analysed in the current study.
Most of the previous hot hand research has been done on professional athletes. But it could be argued professional athletes are less likely to be affected by previous performances. Presumably one of the reasons professional athletes are superior is because they are better at staying in the present, which makes them less likely to be affected by prior performance, and less likely to generate perceptions of psychological momentum compared to amateurs.
The second hypothesis that males would have a significantly higher percentage of negative facilitators compared to women was also supported, which suggests women respond to performance failure differently than men in amateur golf. It was also interesting to find women had a significantly higher chance of experiencing negative momentum, showing women perform worse after an error in performance compared to men. There may be a number of reasons why women tend to perform worse after an error compared to men but there has been little research investigating this topic.
It is possible that men and women differ in their perception of performance failure. It has been suggested that men and women view performance differently because women are socialised to be less competitive and achievement orientated than men, which may explain the gender difference in subsequent performance after failure.
Perhaps men and women differ on personal variables such as confidence. It has been claimed that an individual’s level of confidence has an impact on ones ability to recover from a performance mistake. As women have been shown to have lower levels of confidence across a number of achievement situations, it is quite possible women golfers have lower levels of confidence which makes them more prone to poor performance after an error.
Limitations
As there were no direct measures of perceptions of psychological momentum in this study, perceptions of psychological momentum have had to be inferred as the reason for changes in performance. However, this limitation needs to be considered when evaluating the implications of this study.
Future Research
The notion of momentum is entrenched in sport society. But there is still much to be learnt about psychological momentum from a scientific point of view and further investigation is required to advance our knowledge of this phenomenon.
Further research investigating involving a greater number of attempts for each participant, to increase the power and chance of finding an effect.
The present study showed that women golfers were significantly more likely to experience negative momentum and less likely to experience negative facilitation compared to men. Future research could examine possible reasons behind this apparent gender difference. It could be women have a more adverse reaction to failure than men (which has a negative impact on subsequent performance), because of differences in personal variables such as ones level of trait anxiety, fear of failure, and confidence.
Thirdly, the rounds included competition and non-competition rounds, but mainly non-competition rounds. If rounds were attained from competition play only, then one might expect an even stronger ‘hot hand’ effect, because competition would arguably provide greater pressure and greater perceptions of psychological momentum.
Conclusion
Despite a large number of sport psychologists investigating the relationship between psychological momentum and performance, few definitive conclusions have been drawn. The aim of this research was to add to the existing literature on psychological momentum by investigating the hot hand phenomenon in the sport of amateur golf.
The current study expanded on previous psychological momentum research by analysing the hot hand phenomenon on a sample of 3214 participants (a considerably larger sample size than any previous hot-hand research). The main findings are summarised below:
For a greater number of golfers than would be expected by chance, a golfer’s score on a hole was influenced by their score on the previous hole. This finding supports the notion of the hot hand phenomenon, and suggests psychological momentum does have an influence on performance.
The results of this study concur with research conducted in sports with similar characteristics, i.e. research in other closed skilled sports. It appears closed skill (and amateur) sports may lend themselves to conditions that generate greater perceptions of psychological momentum, and/or where perceptions of psychological momentum have a greater impact on performance.
Significant differences were found in the way males and females perform after an error. There were a significantly higher percentage of males who performed better after an error (negative facilitation) compared to females. Also, there were a significantly lower percentage of males who performed significantly worse after an error (negative momentum) compared to females. This finding suggests females have a more adverse reaction to an error in performance compared to males.
Comments