The golf swing is a repetitive exercise, but many golfers face some problems in that their techniques cannot always be applied in a round because the conditions for every shot are different. Unlike other sports with conditions specified in terms of court size, they must play in huge golf courses with varied terrains under various constrained environments (field of view, slope, and weather conditions). In a recent study, a total of 953 shots of 22 PGs were analyzed in 16 different courses [1]. The results showed that the average slope of all shots was 4.6°, and approximately 80% of the shots had a slope of 1°–10°. Almost all players practice at flat golf driving ranges, although the player rarely ever has to hit a shot on flat ground in a golf match. Amateur golfers (AGs) often practice in golf ranges with a fixed environment to warm up and practice their skills before the round. However, in an actual course, all shots must be executed under constrained conditions, and AGs often take shots without being able to process the interference of the environmental information, leading to poor results [2, 3]. Although unexpected situations arise during every shot in a round, a professional golfer (PG) will attempt to perform appropriate shots while successfully overcoming these various disturbances.

Golf swings involve focusing on a stationary ball in a static address position and performing a fast, dynamic motion, and almost all measurements have traditionally been performed using high-speed cameras or ultrasonic Doppler for golf equipment in small indoor flat environments [4-6]. However, recent studies have compared shots in uphill and downhill conditions using sloping mats in indoor environments [7, 8]. These studies targeted low-handicap players (3-5) and 12 average golfers (10-15) using a 6-iron golf club on a 5° slope mat. In an attempt to resemble the outdoor environment during the second shot in golf, they used a virtual reality (VR) golf simulator (consisting of a screen placed in front of the golfer’s position), and motion analysis was conducted under uphill and downhill conditions to measure the trajectory and center of gravity sway (taking the slope mat). The results showed that the distance traveled by the ball was significantly longer in the low-handicap group and that the leftward deviation from the target was considerably smaller [8]. Both studies found that the left-right deviation with respect to the target was the largest in uphill conditions, and the ball position with respect to the stance width (distance between the left and right feet) was closer to the center in downhill conditions. However, in these studies, the subjects could only recognize the slope by feeling the inclination through the soles of their feet. In addition, the VR screen was set to show the direction of flight and display the trajectory of the shot on flat ground. Therefore, these analyses did not include shots on slopes, which may interfere with the field of view and are characteristics of actual golf courses.

Recent trajectory analyses have shown that hitting the clubhead at 90° to the target contributes to most of the directionality and makes the golf ball fly straight in a horizontal direction [9-12]. From a kinematic point of view (Fitts’ law) [13], AGs with swing speeds slower than PGs could easily repeat the square conditions of the swing. However, a study has clarified that PGs have a smaller standard deviation (variation) in a square face angle and better reproducibility [11]. In other words, PGs have fast club head speeds (CHSs) and are able to achieve the seemingly contradictory motor skill of swinging the club head at a square angle. Further, as low-handicap players make the ball “fly without curving,” a remarkable difference could occur between PGs and AGs starting from the first shot. When humans perform the physical exercise for a certain purpose, they initially use visual details from the outside environment. Subsequently, depending on the perceived details, they consciously or unconsciously choose and express the optimal possible movements based on their physical function and past experiences. Gibson proposed the concept of affordance, which refers to how an individual adapts and acts on his or her internal environment in response to the information provided by the external environment [14]. While AGs tend to have an inconsistent play owing to the illusion that occurs in the retina during each shot, it is unclear whether PGs with skilled swing movements adapt to the perceptual information from the environment and then perform the same swing motion in all conditions. Therefore, confirming the difference in affordance between high- and low-skill shot movements based on the details provided by the environment is critical to understanding the basis of golf skills. In the present study, we hypothesized that PGs would perform the same stable club head movement if they shot toward uphill or downhill fairways, but AGs would have changes in the movement along with the forward fairway’s landscape. This study aimed to analyze the club head movement at ball impact of a tee shot by PGs and AGs in swinging to uphill and downhill fairway environments at an outdoor golf course using the TrackMan portable launch monitor.

Table 1 shows the data of shots on the downhill and uphill slopes obtained from TrackMan. CHS at uphill and downhill conditions in AGs were 39.8 ± 2.8 and 39.7 ± 3.0, respectively. The significance level of normality was 0.008 and 0.001, respectively. Therefore, the CHS in the AG was not satisfied in terms of normality. The CHSs at uphill and downhill conditions in PGs were 45.0 ± 2.0 and 44.5 ± 2.0, respectively. The significance level of normality was 0.200 and 0.200, respectively. Therefore, the CHS in the PG was satisfied in terms of normality. The ANOVA for the attack angle and the first-order interaction (two-way interaction) showed a significant difference between Level × Condition (F (1, 65) = 16.439, p = 0.000, η_p² = 0.580). The main effect of Level was significant (F (1,65) = 8.723, p = 0.004, η_p² = 0.118). In addition, the main effect of Condition was significant (F (1,65) = 89.616, p = 0.000, η_p² = 0.580). Therefore, the attack angle in the uphill condition was significantly larger than that in the downhill condition in both groups, and the PG values were higher than the AG values in both slope conditions. ANOVA for the launch angle and the first-order interaction (two-way interaction) revealed a significant difference between Level × Condition (F (1, 65) = 4.342, p = 0.041, η_p² = 0.063). The Level condition had no main effect. However, the main effect of Condition was significant (F (1,65) = 40.295, p = 0.000, η_p² = 0.383). Therefore, these results revealed that both groups showed significantly larger values in the uphill slope condition.

A follow-up of the two-way interaction showed no significant differences in CHS, face angle, club path, and dynamic loft. However, the main effects of Level and Condition on CHS were significant. Hence, the CHS of PGs was significantly higher than that of AG. Furthermore, the CHS in the uphill condition was higher than that in the downhill condition in both groups. The main effect of condition on the face angle was significant. Hence, the face angle in the uphill condition was significantly larger than that in the downhill condition in both groups. The main effect of Level on the club path angle was significant. The club path angle of PGs was significantly larger than that of AG, indicating that the club path direction of PGs approached the horizontal movement to the target line in both slope conditions. The main effect of the condition on dynamic loft was also significant, demonstrating a significantly larger dynamic loft angle in the uphill condition in both groups. Table 2 shows the Pearson correlation coefficients between the attack and launch angles of the AGs and PGs. PGs had a significant correlation in both slope conditions (p < 0.05), but AGs did not show a significant correlation in the downhill condition (r = 0.366, p = 0.071).

This study examined the adaptation of motor skills to the golf course environment by comparing the differences in the golf swings between PGs and AGs in uphill and downhill conditions, which affect the players’ field of view. We also analyzed the clubhead behavior at ball impact of a tee shot by 42 PGs and 25 AGs in swinging to the uphill (+5°) and downhill (-5°) fairways toward a display yardage board at 200 yards using the TrackMan portable launch monitor. Our hypothesis was rejected, as we found that PGs always change the club head movement along the fairway-of-view condition.

Highly skilled PGs in this experiment achieved optimal ball trajectory by recognizing the affordance differently from AGs, as illustrated by the fact that the attack angle was increased by approximately 3.3° for a height difference of 30 m. The difference in the attack angle between the uphill and downhill conditions among the AGs was only 1° or less, whereas the PG had a difference of at least 3°, suggesting that PGs unconsciously changed their swing in response to the visual information. In other words, a highly-skilled PG uses details from the environment as cues to reduce the attack angle and hit the ball low in downhill conditions or to increase the attack angle and hit the ball high in uphill conditions. Hence, PGs chose to hit low balls in downhill conditions and high balls in uphill conditions. The attack angle is an indicator of the upward (+) or downward (-) motion of the club head with respect to the ground at impact and is a major factor in determining the height of a shot in the trajectory analysis [20]. The proper angle range for the driver is proposed to be -3° to 5° [9]. The launch angle is the index of the vertical trajectory angle of the ball, which is the sum of the attack angle and dynamic loft angle. In the AG, no correlation between the attack and launch angles was observed in the downhill condition, implying that they could not control the height of the ball according to the slope.

Uphill shots will land on the ground immediately after launch if the launch angle of the ball is low, making it impossible to maximize the flight distance. Therefore, we surmise that, based on a long experience, PGs unknowingly increased the launch angle to avoid mistakes and maximize the flight distance. In contrast, if the launch angle of the ball is high in downhill conditions, the flight time in the air becomes longer and the ball follows a more curved trajectory than that in the uphill condition before reaching the ground. In addition, the trajectory of the ball in downhill conditions is more affected by the wind owing to the longer flight time. Therefore, shots with a high launch angle may increase the risk of failure in a round of golf. We speculate that this is the reason why the PG decreased the launch angle in downhill conditions. There appeared to be a major difference between AGs and PGs in the perception of the information received from the environment, even under the same conditions.

A study in the field of perceptual psychology has revealed that the height (vertical position) of an object in the field of vision is a clue (information source) to the depth (relative distance) and that there is a horizon-ratio relation between the relative height from the horizon and the actual horizontal distance [21]. The golf course has various fields of view. In particular, when the tee shot is downhill, the far side hill is not a limiting factor for the player’s view; the relative height from the horizon to the ball landing point can be precisely confirmed in this condition, and thus, the player can execute their affordance (Fig. 3). However, in the uphill condition, it is difficult to receive an attribution of height; thus, players have to deal with the loss of the target.

Interestingly, there was a large difference in the affordance behavior of skilled PGs in driver shots that require a high degree of skill because of differences in the visual environment. This ability is thought to be acquired during the autonomous movement stage by repeated swing practice. PGs have a vast experience playing in various golf courses, which have automated golf swings. In the process of becoming skilled in golf swings, it is believed that swings suitable for the environment gradually and unconsciously become established by repeating failures while aiming for the optimal swing for the environment. The results suggest that new affordances are induced according to the golf course’s environment, forming an interrelationship between affordances and behaviors [2].

In recent years, the method of motor learning in which discovery learning is promoted by using various constraints for sports skill acquisition is referred to as a constraint-led approach (CLA) [22, 23]. One important point when using the CLA is that coaches coordinate environmental and task constraints with the learner’s current abilities and skill level to facilitate adaptation to practice and learning opportunities [24]. If the difference between the uphill and downhill conditions adopted in this experiment is seen as a limiting factor, there is a potential to contribute to improving the performance of golfers, including AG. In particular, past problems in coaching golf have been the excessive pursuit of a perfect swing and of techniques performed on a flat, featureless golf course [2].

CLA has been developed in the field of physical education pedagogy for children to deepen the enjoyment of exercise tasks. In recent years, the range of applications has expanded to the field of motion perception in tennis and golf [2, 25-27]. In other ball sports, experts adopt modes of movement that are different from normal movements because of restrictions in the environment they perceive. A study on basketball free throws reported that reducing the diameter of the basketball rim by 0.10 m results in an increase in the ball release angle [28]. In this experiment, the diameter was made smaller than the standard rim by 0.10 m, and basketball players with advanced shooting skills increased their shooting success rate by increasing the approach angle of the ball when it goes through the rim. This investigation suggested that if the main problem in players with low free-throw success rates was a smaller ball release angle, the shooting success rate could be increased by increasing the release angle. This can be achieved by practicing a rim with a small diameter.

Furthermore, a previous study investigated the swing trajectory of equipment in baseball players [29]. With a wall projecting a VR baseball stadium, subjects in the constraint group were instructed to hit the ball over the wall, and their results were compared with those of an internal focus group who were instructed to move their arms at an upward angle. The results confirmed that the launch angle and number of home runs increased in the constraint group. Similar results were observed in an external focus group that was instructed to aim and hit the ball on the bottom half. The results of these experiments suggest that affordances may be recognized and that optimal movements may be derived even under restricted conditions among skilled participants. If behaviors and movements change depending on the awareness of the player based on the information received by peripheral vision, further advances in VR might be useful for improving complex motor learning, such as golf swing in indoor environments. However, enhancing affordances by training with all senses in a constrained outdoor environment is crucial. Moreover, the interpretation of excessive feedback of data from trajectory analysis equipment might adversely affect an individual’s ability to perform [18]. Thus, coaches must take great care when using results correctly in an appropriate learning environment so that players can achieve the desired goals.

The results of this experiment confirmed that the higher the player’s skill level, the better their recognition of the affordance of the visual field and the greater their ability to adapt to environmental constraints. Conversely, as low-skilled players were inferior in terms of affordance recognition (information assembling ability), future golf research should advance with practical interventions with constraint conditions.