Performance analysis based on GPS data of Olympic class windsurfing

ABSTRACT

This study examined the technical and tactical variables that determine performance before departure and during a race in all wind conditions. Data were collected from 44 Olympic-class windsurfers who participated in the World Cup and Allianz Regatta. Distance to the start line and speed 5 s before departure, distance to start line and speed when starting, and distance travelled and time at each leg were recorded via GPS (5 Hz). Average speed was calculated through distance travelled and time. Results indicated that the high-performance group had significantly (p < 0.05) less distance from the starting line and faster speed when starting than the low-performance group in heavy wind conditions. During the first or third legs with light and medium winds, the high-performance group demonstrated a significantly (p < 0.05) faster average speed in heavy wind and significantly less distance travelled, and faster average speed was observed in all legs. The first half of the high-performance group had significantly (p < 0.05) longer distance travelled than the second half. Significantly faster average speed was observed in light wind. These results may assist in preparing for the tactical and technical requirements during a race and enhance performance in every situation.

KEYWORDS:

• Olympic board sailing
• windsurfing
• performance
• tactical
• technical

1. Introduction

All Olympic sailing classes, including windsurfing, are considered complex, multidimensional sports in which tactical, technical, psychological, and physical qualities must be considered to determine a successful race outcome (Bojsen-Møller et al., 2015, 2007; Caraballo, Lara-Bocanegra et al., 2021). Since physical propulsion by pumping is possible in windsurfing (Bojsen-Møller et al., 2015; Guevel et al., 2000; De Vito et al., 1997), scientific analysis of the related technical and physical qualities is an important factor that can enhance performance. Several aspects related to performance have been identified, such as anthropometric (Caraballo et al., 2020; Cortell-Tormo et al., 2010; Doumtsios et al., 2010; Pérez-Turpin et al., 2009), technical, and physiological qualities (Andrianopoulos & Vogiatzis, 2017; Castagna et al., 2008, 2007).

Recently, to improve the understanding of race operations, the tactical and physiological capacities of sailors were investigated by monitoring or GPS tracking and heart rate during official Laser and kiteboarding races (Caimmi & Semprini, 2017; Winchcombe et al., 2021). Additionally, considering environmental conditions, interest in the sailors’ technical and physiological abilities demanded by wind conditions has increased (Caraballo, Cruz-Leon et al., 2021; Chun et al., 2021; Winchcombe et al., 2021).

Interestingly, Anastasiou et al. (2019) investigated the tactical and technical qualities of windsurfers in various wind conditions via GPS, and emphasised the first upwind leg and the position of the first mark rounding for finishing a race successfully. However, Anastasiou et al. (2019) only focused on the first half of the race and neglected the second half, other legs, and before departure. Hagiwara and Ishii (2016) confirmed that distance to the start line when starting the race differed according to performance levels during practice races. Researchers analysed the differences between the performance levels of upwind, downwind, and beam reach sections in Formula Kite (Caraballo, González Montesinos et al., 2021), Laser (Caraballo, Conde-Caveda et al., 2021), and 2.4 mR class studies (Paralympic; Caraballo, Cruz-Leon et al., 2021) to identify the variables that determine performance and emphasise the importance of specific sections. These results can help coaches and sailors design racing strategies and understand how to race better (Anastasiou et al., 2019).

Thus, the purpose of this study was to investigate the technical (speed) and tactical (distance) variables that determine performance according to the influence of before departure, each leg, and first and second halves of the race on performance in the windsurfing class under all wind conditions (Caraballo, González Montesinos et al., 2021). We hypothesised that, under all wind conditions, there would be a significant difference between the high-performance group and the low-performance group before departure, when starting, and at each leg (Caraballo, Conde-Caveda et al., 2021; Caraballo, González Montesinos et al., 2021). Furthermore, the tactical strategies of the first half in the upwind section of the race would be different from the second half (Anastasiou et al., 2019).

2. Materials and methods

2.1. Participants

The study sample consisted of 44 male windsurfers from 24 different nationalities. The analysis was conducted considering the 2020 World Cup series in Enoshima, Japan (12 races) and the 2021 Allianz Regatta in Medemblik, The Netherlands (11 races). This study was conducted according to the guidelines of the Declaration of Helsinki and approved Research Ethics Committee.

All the races in the Hempel World Cup series (four competitions) and Allianz Regatta (one competition) hosted by World Sailing from 2019–2021 were reviewed. The competitions and races were selected based on the following qualification criteria: (1) a minimum of 20 windsurfers must have participated; (2) the racecourse must be a windward-leeward course to generate a comparison between the first and second half of the upwind and downwind section. Based on the results of each race, the data of the top five positions and the bottom five positions were collected depending on the following eligibility criteria: (1) the windsurfer must have completed the race; (2) the wind and GPS data (distance, speed, and time) must be checked normally.

The performance level was classified into two groups: (1) high performance and (2) low performance. The high-performance group was defined as athletes who finished in the top five positions for each race. The low-performance group was defined as those who finished in the bottom five positions. The high-performance group included windsurfers ranked 1–10 in the World Sailing rankings. Wind conditions were categorised as light wind (LW; ≤ 8 kts), medium wind (MW; ≤ 8.1–15.9 kts), and heavy wind (HW; ≥ 16 kts; Anastasiou et al., 2019). As seen in Figure 1, the racing course was windward-leeward.

Figure 1. Windward-leeward course. The windward-leeward course consisted of two upwind (windward), two downwind (leeward), and one beam reach sections.

2.2. Data acquisition

The GPS variables were obtained through the SAP-Sailing® application. The GPS device used in this study was a compact 60 g device that consisted of a mobile connection and a battery. Using this device, distance, speed, and time data were sampled and collected at 5 Hz intervals at 5 s before departure, when starting, and during the race. The information received by the device was transmitted in real time to the TracTrac® system via mobile network.

2.3. Statistical analyses

Following the Shapiro–Wilk test for normality, the parametric or non-parametric test methods were used. Student’s unpaired t-test or Mann–Whitney U tests were performed in all wind conditions to determine the performance levels for distance to the start line at 5 s before departure (m; DS₅), speed at 5 s before departure (kts; S₅), distance to the start line when starting (m; DS₀), speed when starting (kts; S₀), distance travelled (m; DT), and average speed (m/s; AS) at each leg (first, second, third, fourth, and fifth leg). The Paired t-test or Wilcoxon signed rank test assessed differences in the distances travelled and average speeds between the first and second half of the upwind and downwind section in the high-performance group. Average speed (AS) was calculated according to Equation 1(1) $\mathrm{A}\mathrm{S}=\frac{\mathrm{D}\mathrm{T}}{\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{e}\mathrm{t}\mathrm{a}\mathrm{k}\mathrm{e}\mathrm{n}}$(1) and is shown in Tables 2 and 3. The distance travelled (DT) and time taken at each leg were used.

(1) $\mathrm{A}\mathrm{S}=\frac{\mathrm{D}\mathrm{T}}{\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{e}\mathrm{t}\mathrm{a}\mathrm{k}\mathrm{e}\mathrm{n}}$(1)

All data are presented as mean (M) ± standard deviation (SD). The level of statistical significance was set at (p < 0.05), and a 95% confidence interval (CI) was used for all analyses. The statistical tests were conducted using SPSS 23.0 software (IBM, Corp., Armonk, NY).

3. Results

Table 1 shows the results for the performance level of each windsurfer at their respective distances to the start line and speed at 5 s before departure and distance to the start line and speed when starting. The high-performance group had significantly less DS₀ (Z = −2.365, p = 0.018) and greater S₀ (T = 2.750, p = 0.008) when starting compared to the low-performance group in HW condition. No performance difference was revealed for DS₅, S₅, DS₀, and S₀ in LW and MW conditions.

Table 1. Results (M ± SD and 95% CI) for the difference according to performance levels before departure and when starting

Variables	High-performance group		Low-performance group		T/Z*	p
	M ± SD	95% CI	M ± SD	95% CI
Light wind
DS5 (m)	9.28 ± 3.81	7.82–10.76	8.82 ± 3.72	7.34–10.25	0.432	.668
S5 (kts)	1.65 ± 0.68	1.40–1.94	1.32 ± 0.44	1.15–1.49	−1.669*	.095
DS0 (m)	5.74 ± 3.52	4.30–7.11	5.69 ± 3.48	4.36–7.06	0.050	.960
S0 (kts)	3.80 ± 0.76	3.53–4.09	3.42 ± 0.59	3.20–3.66	−1.936*	.053
Medium wind
DS5 (m)	9.83 ± 4.80	8.48–11.23	9.68 ± 6.33	8.04–11.70	−0.456*	.648
S5 (kts)	2.95 ± 3.00	2.10–3.86	3.11 ± 2.93	2.28–3.99	−0.666*	.506
DS0 (m)	6.13 ± 3.34	5.15–7.11	6.16 ± 4.22	5.00–7.46	−0.161*	.872
S0 (kts)	7.26 ± 2.35	6.62–7.94	6.85 ± 2.50	6.09–7.60	1.976	.054
Heavy wind
DS5 (m)	32.4 ± 132.80	11.63–91.47	14.6 ± 10.6	11.85–18.15	−0.105*	.916
S5 (kts)	8.85 ± 4.11	7.71–10.03	7.32 ± 4.17	6.19–8.49	−1.783*	.075
DS0 (m)	5.48 ± 4.26	4.26–6.80	8.77 ± 6.84	6.94–10.83	−2.365*	.018
S0 (kts)	11.69 ± 1.39	11.29–12.10	10.56 ±2.38	9.84–11.24	2.750	.008

DS₅: distance to the start line at 5 s before departure; S₅: speed at 5 s before departure, DS₀: distance to the start line when starting, S₀: speed when starting, * non-parametric statistics

In contrast, in Table 2, the distance travelled and average speed showed variations according to the leg for each performance level. Under the LW condition, the high-performance group obtained significantly greater AS in the first (T = 3.307, p = 0.002) and third legs (T = 2.845, p = 0.007) compared to the low-performance group. No difference was found between performance in AS at second, fourth, and fifth legs and DT at all legs. Similarly, in the MW condition, the high-performance group obtained significantly higher AS in the first leg (Z = −2.288, p = 0.022) compared with the low-performance group. However, no differences were found between AS performance in second, third, fourth, and fifth legs and DT performance in all legs. In the HW condition, the high-performance group achieved less DT in the first (T = −9.156, p < 0.001) and second (T = −2.124, p = 0.036), third (Z = 4.923, p < 0.001), and fourth legs (T = −2.954, p = 0.004) compared to the low-performance group. No significant differences were found between performance in DT at the fifth leg. Moreover, the high-performance group had significantly greater AS in the first (T = 8.217, p < 0.001) and second (Z = −6.008, p < 0.001), third (T = 4.398, p < 0.001), fourth (Z = −6.072, p < 0.001), and fifth legs (T = 5.069, p < 0.001) compared to the low-performance group.

Table 2. Results (M ± SD and 95% CI) for the difference according to performance levels in each leg during the race

Variables	High-performance group		Low-performance group		T/Z*	p
	M ± SD	95% CI	M ± SD	95% CI
Light wind
First leg DT (m)	948.20 ± 155.65	888.66–1006.09	1020.44 ± 176.76	948.65–1090.55	−1.795*	.073
First leg AS (m/s)	1.97 ± 0.18	1.90–2.04	1.79 ± 0.18	1.72–1.86	3.307	.002
Second leg DT (m)	554.28 ± 114.25	510.13–598.24	547.44 ± 104.76	507.73–586.35	−0.233*	.823
Second leg AS (m/s)	2.41 ± 0.27	2.3–2.52	2.25 ± 0.31	2.13–2.38	−1.775*	0.76
Third leg DT (m)	668.08 ± 131.51	641.10–736.84	738.56 ± 137.52	682.73–789.36	−3.587	.191
Third leg AS (m/s)	1.84 ± 0.22	1.75–1.93	1.67 ± 0.18	1.6–1.75	2.845	.007
Fourth leg DT (m)	530.92 ± 57.94	511.69–552.48	536.16 ± 67.49	509.01–561.27	−0.204*	.839
Fourth leg AS (m/s)	2.43 ± 0.29	2.31–2.54	2.34 ± 0.35	2.2–2.48	−0.941*	.347
Fifth leg DT (m)	160.80 ± 22.65	152.48–168.76	163.96 ± 20.38	155.48–171.56	−0.894*	.371
Fifth leg AS (m/s)	2.72 ± 0.57	2.5–2.94	2.66 ± 0.62	2.41–2.89	0.387	.700
Medium wind
First leg DT (m)	1962.67 ± 463.97	1833.08–2099.91	2080.18 ± 628.03	1900.66–2257.99	−0.569*	.569
First leg AS (m/s)	4.53 ± 1.57	4.1–5.02	3.89 ± 1.46	3.42–4.32	−2.288*	0.22
Second leg DT (m)	1251.80 ± 297.21	1163.91–1336.27	1250.69 ± 332.65	1156.52–1354.56	0.017	.987
Second leg AS (m/s)	6.21 ± 1.98	5.64–6.79	5.79 ± 2.26	5.06–6.46	−0.835*	.404
Third leg DT (m)	1668.58 ± 446.94	1544.55–1802.71	1804.80 ± 623.49	1624.79–1996.11	−0.799*	.424
Third leg AS (m/s)	4.81 ± 1.77	4.32–5.34	4.20 ± 1.76	3.68–4.7	−1.804*	.071
Fourth leg DT (m)	1246.42 ± 202.71	1191.96–1301.92	1308.96 ± 265.64	1224.78–1388.05	−0.295	.770
Fourth leg AS (m/s)	6.69 ± 1.71	6.19–7.18	5.98 ± 1.87	5.38–6.5	−1.949*	.051
Fifth leg DT (m)	180.62 ± 17.63	175.85–185.89	180.07 ± 18.54	175.04–185.35	−0.562*	.574
Fifth leg AS (m/s)	6.46 ± 1.28	6.07–6.84	6.03 ± 1.89	5.49–6.56	0.387	.700
Heavy wind
First leg DT (m)	2717.71 ± 185.96	2663.32–2777.33	3164.67 ± 269.55	3081.64–3238.34	−9.156	.000
First leg AS (m/s)	6.69 ± 0.26	6.61–6.77	5.94 ± 0.54	5.78–6.09	8.217	.000
Second leg DT (m)	1868.96 ± 148.14	1823.09–1911.97	1947.51 ± 199.03	1889.00–2003.21	−2.124	.036
Second leg AS (m/s)	10.55 ± 0.67	10.36–10.75	9.37 ± 0.93	9.09–9.63	−6.008*	.000
Third leg DT (m)	2574.71 ± 251.00	2500.66–2649.55	2922.73 ± 370.44	2822.00–3034.87	−4.923*	.000
Third leg AS (m/s)	6.68 ± 0.28	6.6–6.77	6.25 ± 0.59	6.08–6.43	4.398	.000
Fourth leg DT (m)	1845.78 ± 148.80	1800.60–1888.90	1946.18 ± 172.70	1896.49–1993.97	−2.954	.004
Fourth leg AS (m/s)	10.50 ± 0.48	10.35–10.64	9.55 ± 0.92	9.29–9.85	−6.072*	.000
Fifth leg DT (m)	259.40 ± 22.68	253.02–266.15	262.38 ± 21.24	256.36–268.18	−1.050*	.294
Fifth leg AS (m/s)	10.31 ± 0.76	10.08–10.54	9.50 ± 0.74	9.28–9.72	5.069	.000

DT: distance travelled; AS: average speed; * non-parametric statistics

Finally, Table 3 shows the DT and AS of the high-performance group according to the first and second half of upwind and downwind sections. In the upwind section, under the LW condition, the first half of the race had significantly greater DT (Z = −4.049, p < 0.001) and AS (T = 3.824, p = 0.001) compared to the second half of the race. For the MW condition, the first half of the race had greater DT (Z = −4.645, p < 0.001) than the second half, and AS did not reveal any differences. Similarly, under the HW condition, the first half of the race had greater DT (T = 4.790, p < 0.001) than the second half, and AS did not exhibit any differences. No significant differences were observed in the downwind section for all wind conditions.

Table 3. Results (M ± SD and 95% CI) for difference between the first half of the race compared to the second half in high-performance group

Variables	First half		Second half		T/Z*	p
	M ± SD	95% CI	M ± SD	95% CI
Upwind
LW DT (m)	948.20 ± 155.65	888.66–1006.09	688.08 ± 131.51	641.10–736.84	−4.049*	.000
LW AS (m/s)	1.97 ± 0.18	1.9–2.04	1.84 ± 0.22	1.75–1.93	3.824	.001
MW DT (m)	1962.67 ± 463.97	1833.08–2099.91	1669.58 ± 446.94	1544.55–1802.71	−4.645*	.000
MW AS (m/s)	4.53 ± 1.57	4.1–5.02	4.81 ± 1.77	4.32–5.34	−0.988*	.323
HW DT (m)	2717.71 ± 185.96	2663.32–2777.33	2574.71 ± 251.00	2500.66–2649.55	4.790	.000
HW AS (m/s)	6.69 ± 0.26	6.61–6.77	6.68 ± 0.28	6.6–6.77	0.178	.860
Downwind
LW DT (m)	554.28 ± 114.25	510.13–598.24	530.92 ± 57.94	511.69–552.48	−1.641*	.101
LW AS (m/s)	2.41 ± 0.27	2.3–2.52	2.43 ± 0.29	2.31–2.54	−0.632*	.527
MW DT (m)	1251.80 ± 297.21	1163.91–1336.27	1246.42 ± 202.71	1191.96–1301.92	0.171	.865
MW AS (m/s)	6.21 ± 1.98	5.64–6.79	6.69 ± 1.71	6.19–7.18	−0.073*	.942
HW DT (m)	1868.96 ± 148.14	1823.09–1911.97	1845.78 ± 148.80	1800.60–1888.90	1.634	.109
HW AS (m/s)	10.55 ± 0.67	10.36–10.75	10.50 ± 0.48	10.35–10.64	0.485	.630

DT: distance travelled; AS: average speed; LW: light wind; MW: medium wind; HW: heavy wind; * non-parametric statistics

4. Discussion

This study aimed to examine the effect of technical (speed) and tactical (distance) variables on performance before departure, each leg, and the first and second half of a windsurfing race in all wind conditions. The hypotheses of this research were partially confirmed as the results showed significant differences in DS₀ and S₀ performance levels under heavy wind condition, AS performance in the first and third legs under light and medium wind conditions, and also DT and AS performance in almost all legs under heavy wind condition. Additionally, in the first and second halves of the race during upwind section, the AS was different under light wind condition, and the DT was different under all wind conditions.

The novel findings of this study showed that the high-performance group had a different start and race strategy compared to the low-performance group in heavy wind conditions. Taking the lead during the first upwind leg minimised and contained interference from other competitors and allowed strategies and tactics to be used. Consequently, it can positively impact competitive performance (Anastasiou et al., 2019). Thus, being ready with a before-departure strategy could help windsurfers take the lead and thereby gain an advantage. However, the results obtained during light and medium wind conditions in this study did not show a difference based on performance level, and the results when starting are different from those of previous studies, which confirmed variation in performances of windsurfers according to light and medium wind conditions (3–5 m/s; approximately 6–10 kts; Hagiwara & Ishii, 2015, 2016). The results before departure may have been due to the prioritisation of maintaining a position at the start line and then starting rather than starting by making room for acceleration (Gladstone, 2002). The results when starting could be attributed to the differences in the starting point used in the analysis. In the previous study, the average speed and distance to the start line 0 to 5 s after starting were used for analysis (Hagiwara & Ishii, 2015). However, in this study, the speed and distance to the start line when starting were used. Windsurfers tried to get away from the group when starting from the start line; however, board speed may not have increased quickly (within a second) because of position maintenance. Nonetheless, pumping was an effective way to increase board speed, allowing windsurfers to quickly pull away from the group at the starting line under light or medium wind conditions (Castagna et al., 2007; Vogiatzis & De Vito, 2015). Therefore, the advantages of pumping should be emphasised for a successful start under light and medium wind conditions. When starting under heavy wind conditions, the DS₀ of the high-performance group was smaller than that of the low-performance group, and S₀ was faster. This indicates that the heavy wind conditions may have helped in quickly pulling away from the starting group and sailing with the best combination of speed and sailing angle (velocity made good; VMG; Caraballo, González Montesinos et al., 2021). Additionally, considering that the high-performance group was located 32.4 m from the start line 5 s before departure and was moving at a speed of 8.85 kts; it may have been more important to make room for acceleration in heavy wind conditions compared to light or medium wind conditions. Thus, these results indicate that there was a difference between the starting strategy in light or medium wind conditions and that in heavy wind conditions. It can be concluded that technical and tactical variables played an important role when starting in heavy wind conditions.

Our results regarding the speed and distance of each leg in light and medium wind conditions are in line with past findings regarding the first legs (Anastasiou et al., 2019). Furthermore, the significance of the third leg, which corresponds to the upwind section, was validated. For light and medium wind conditions, windsurfers rely on pumping to propel their boards (Chamari et al., 2003). Pumping is known to reduce the distance travelled and increase board speed during races (Castagna et al., 2007; Vogiatzis & De Vito, 2015). However, pumping is a rhythmic manoeuvre, and speed may vary depending on the performance level and pumping method (Castagna et al., 2008; Ouadahi et al., 2014; Vogiatzis & De Vito, 2015). Moreover, the influence of environmental variables such as tides, wind, and waves should be considered (Castagna & Brisswalter, 2007; Gladstone, 2002). Otherwise, the board speed may lose its efficacy or may become physically demanding (Ouadahi et al., 2014). Additionally, it was suggested that anaerobic efforts involved in windsurfing are mainly related to pumping (De Vito et al., 1997). As mentioned in a study that elucidated the physiological requirements of Olympic-level windsurfers for pumping (Vogiatzis et al., 2006), pumping requires high metabolic ability during upwind sailing (77% of VO₂max and 87% of HRmax). Windsurfers compete for 30 to 40 minutes in three races a day (Guevel, 1999); therefore, constant pumping is required to propel the board in light and medium wind conditions, which necessitates high aerobic capacity in windsurfers (Castagna et al., 2008). In this study for heavy wind conditions, the high-performance group was confirmed for fast speeds and short distances travelled in almost all legs. The short distances and fast speeds strategically affected the outcome of the race (Gladstone, 2002). Under heavy wind conditions of 15 kts or more, boards are propelled by wind power; under such conditions, board speed is also related to board control ability (Andrianopoulos & Vogiatzis, 2017; Caraballo, González Montesinos et al., 2021). When board control is used to reduce the propulsion angle in all sections (upwind and downwind), it can decrease speed or (vice versa) increase the distance travelled (Ferretti & Festa, 2019; Hagiwara & Ishii, 2015). Therefore, this indicates that it is important to control the board to achieve optimal performance while maintaining high speed under heavy wind conditions; it also indicates that the low-perf group may need to be improved. Furthermore, the technical and tactical variables for optimal performance differ depending on the wind conditions, just as the physical capacities of sailors for optimal performance differ according to wind conditions (Day, 2017).

The results of this study, confirmed by each section in the first and second halves of the race, show that the tactical strategies of the first half of the race in the upwind section are different from those in the second half. These results are related to the results of previous studies that confirmed the heart rate and blood lactate response in the first half and the second half in the practice race (Guevel, 1999). In the study by Guevel (1999) it was observed that heart rates were considerably high in the first half of the race compared to the second half. This may have been attributed either to the above-mentioned synergistic effect of the start, or to strategic reasons (Guevel, 1999). In the first half of the race, it seems that the players fight fiercely to get a position as close as possible to the lead (Anastasiou et al., 2019). Subsequently, the second half of the race aids in consolidating their position with a somewhat defensive attitude (Guevel, 1999). Therefore, results of this study for all wind conditions indicated that the high-performance group moved longer in the first half of the upwind leg than in the second half; this seems to be related to strategic reasons. These results consolidate the importance of the race in the first half of the upwind section (Anastasiou et al., 2019).

In all wind conditions before departure or during the race, technical and tactical variables were different according to the performance level. Furthermore, the technical and tactical variables showed differences according to wind conditions. These results suggest that the low-performance group could use the advantage of technical and tactical variables to overcome constraints depending on the wind condition. However, it is noteworthy that the professional effect of technical and tactical quality is not the only distinction that can explain successful performance in the field.

As this study used public data for analysis, information on weight, age, and experience of players were not included. However, a more interesting analysis would have been possible if this data had been included (Caraballo, González Montesinos et al., 2021). Future research could further this research to quantify the experience of athletes in iQFoiL, an Olympic event in a new windsurfing class. Furthermore, analysis of visual patterns and internal loads such as GPS-based action sports cameras (Pluijms et al., 2016, 2013) and heart rate data, may yield clearer results for tactical, technical, and physical qualities.

Author contributions

conceptualization, S.C. and J.P.; methodology, S.C. and T.K.; software, S.C.; validation, S.C. and Y.K.; formal analysis, S.C. and J.P.; investigation, S.C. and T.K.; data curation, S.C. and J.P.; writing original draft preparation, S.C.; writing review and editing, J.P., S.C., T.K.; supervision, J.P. and T.K.; project administration, Y.K. and T.K. All authors have read and agreed to the published version of the manuscript.

Informed consent statement

Patient consent was waived since all the data used in this study are publicly accessible and can be found in SAP-Sailing®.

Disclosure statement

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

Data availability statement

This study used public data for analysis. (data/20/10/2021) https://www.sapsailing.com/gwt/Home.html#EventsPlace:

Additional information

Funding

No financial support has been received.