Individualisation,readjustment and athlete codetermination of high-performance training in athletics and volleyball

Abstract

The study investigated the extent of individualisation, readjustment and athlete codetermination of high-performance training. Individualisation refers to the extent to which the training design is tailored to an athlete's individual wishes, interests, strengths, and weaknesses. Readjustment refers to the adjustment of day-to-day training to an athlete's current physical, health-related and psychological constitution. Athlete codetermination refers to an athlete's involvement in decision-making on the training design. German national-squad members in athletics and volleyball (n = 336) reported these variables in an online questionnaire. Analyses revealed four central findings: (1) Large parts of athletes’ training were characterised by high degrees of individualisation, readjustment and athlete codetermination. (2) These characteristics were more pronounced in athletics than in volleyball. (3) They explained large parts of the appropriateness of athletes’ training (whether the training was ‘just right’ for an athlete's performance development). Consistently, greater extents of individualisation, readjustment and athlete codetermination of training control were associated with better multi-year performance development. (4) Participation in performance analytics had no additional, specific effect on training appropriateness. All the results consistently applied to both youth age and adult age. In conclusion, greater extents of individualisation, readjustment and athlete codetermination of the training design apparently facilitate long-term training efficiency.

Keywords

Coach-athlete interaction practice design training control

In high-performance sports, a few milliseconds, centimetres, or a single action can make the difference between victory and defeat. Athletes and coaches undertake extensive efforts to optimise their training process and maximise the athlete's performance in critical competitions. The present study explores the extent to which the training of high-performance athletes is characterised by individualisation, readjustment and athlete codetermination. The study also examines whether these characteristics are associated with training appropriateness and athletes’ long-term performance development.

When designing training, coaches draw on a specific mixture of the traditional training methodology within the community of a sport, their coach education, own experience and to some part intuition and instinct.^1–4 Coaches endeavour to teach their athletes perceptual-motor and tactical skills in the most effective way;^5,6 from a physiological perspective, they aim to induce optimal functional overreaching while avoiding overtraining.⁷ It is believed that even nuanced details of the training process may significantly affect athletes’ later performance. Coaches thus not only determine the number of weekly practice sessions and the types of exercises and playing forms, they also define details such as the amount of each activity, the duration of each game format, the number of sets and repetitions per set in each exercise, the intensity and inter- and intra-set rest intervals, the exact way of executing each exercise or game and the efficient organisation of each session.^8–10 Training is mostly done as team practice, and coaches design group training plans that extend over weeks, months, and seasons.^10–13

While practice is mostly designed for groups, athletes within these groups differ individually, for example, regarding years of experience, individual strengths and weaknesses, health-related constitution, load tolerance and psychological conditions.^8–10,14,15 This suggests a certain degree of individualisation – that is, differentiation of the training design between athletes – to tailor the training to each athlete individually. Furthermore, training effects are to some degree uncertain and sometimes do not elicit the anticipated performance progress. Also, an athlete's physical and mental shape-on-the-day, health, load–recovery balance and mood may vary over time. These factors may sometimes suggest revisions of the original training plan and readjustment of the training design on a day-to-day basis.^8,10

The coach continuously observes the athlete in competition and daily practice and assesses their performance development, shape-on-the-day, health-related and mental constitution, and load–recovery balance. This subjective assessment is often supplemented with more objective tests and performance analytics. They are designed to monitor the training load, assess the athlete's health and evaluate performance components such as perceptual-technical and tactical skills, physical abilities (speed, agility, power, strength, endurance and flexibility) and psychological skills (e.g. concentration and self-regulation).^16–19 These procedures typically involve multidisciplinary perspectives including biomechanics, kinesiology, sports medicine, physiology, exercise science and psychology.²⁰ According to traditional training theory literature,^12,21 the idea of training control implies a recurrent cybernetic loop: (1) The coach defines the goals and training design for a defined period. → (2) The athlete executes the training plan. → (3) The athlete's performance components and development are evaluated based on the coach's assessment and performance analytics. → (4) This evaluation informs the training design in the subsequent period. → (5) The athlete executes the training plan … etc. Here, one might take the stance that the elite athlete's role includes more than just executing the coach's instructions. Rather, athletes themselves may be a significant factor in codetermining the training control.^22–24 Athlete codetermination has been widely neglected in traditional conceptions of training control in training theory literature. However, elite athletes have extensive experience and first-order perceptions that neither coaches’ observations nor performance analytics may capture—at least not as “fine-grained” as athletes themselves do: For example, day-to-day perception of their own body, shape-on-the-day, stress, health, load–recovery balance and psychological situation, as well as factors like how well certain exercises, playing forms, ways of executing them or load intensities suit a particular athlete, and also simply how much an athlete does or does not like doing them.^9,22,25 In this way, elite athletes may be considered as experts of themselves. Consequently, athletes’ communication of their experiences, perceptions and perhaps preferences may potentially provide a fruitful contribution to training control and effectiveness.

In summary, any training at any time is located somewhere on the following three continua: (1) from entirely identical training design for all athletes to entirely individualised training for each athlete; (2) from complete adherence of each athlete's training to an original training plan to continuous readjustment of each athlete's training to their current constitution; and (3) from determination of the entire training design by the coach to determination of the entire training design by the athlete(s). These three dimensions were the subject of the present study. The aim was to investigate to what extent the training of high-performance athletes involves individualisation, readjustment and athlete codetermination, and whether these three characteristics are associated with training appropriateness, effectiveness and efficiency.

Typical individual and team sports may differ in the extent of individualisation, readjustment and athlete codetermination. Hence, this study takes potential differences into account and exemplifies this comparison by involving athletics and volleyball. In both sports, training is typically organised as team practice and largely focuses on perceptual-motor skills and physical conditioning.^26,27 In athletics, athletes perform individually in competitions. At the elite level, their training is focused on continuously improving nuances of each athlete's individual skill technique through exercise forms. While practicing as a group, athletes’ tasks often vary individually.²⁶ In volleyball, athletes perform jointly in competitions, and a player's actions and their quality rely on teammates’ actions. At the elite level, their training is mainly focused on continuously improving team and group tactics through playing forms and sub-phases of play: aligning actions of different players within offense and defence plays, playing systems, and offense–defence–transition, especially coordinating the positioning and timing of different players’ actions, and anticipating teammates’ decisions and actions.²⁷ These factors may limit the scope for individualisation, readjustment and athlete codetermination in volleyball compared to athletics.

The coach–athlete interaction and athlete codetermination in the training process have extensively been studied with regard to youth athletes’ autonomy perception, motivation and psychological well-being (see Nichol et al.,²⁴ for a review.). The present study begins to explore potential associations with athletes’ training appropriateness and performance development in a select sample of youth and adult elite performers.

The aim was to explore three questions: (1) To what extent is the training of high-performance athletes characterised by individualisation, readjustment and athlete codetermination? (2) Do these characteristics differ between an individual sport (athletics) and a team game sport (volleyball)? (3) Do individualisation, readjustment and athlete codetermination of training control facilitate training appropriateness and performance development? To the best of our knowledge, this is the first time these questions have been explored among high-performance athletes. We therefore formulate no directed a-priori hypotheses.

Methods

Participants

The sport directors of the German Athletics Federation and the German Volleyball Federation sent e-mails to all national-squad athletes inviting them to participate in the online questionnaire. Between October 2019 and January 2020, 336 (40%) of the national-squad athletes participated, 194 from athletics and 142 from volleyball. Table 1 describes demographic and participation variables of the athletes. The convenience sample exceeded the minimum size suggested by an a-priori power analysis, n ≥ 191 (GPower 3.1.9.7, two-tailed, Cohen's d = 0.20, α = 0.05, 1 − β = 0.80).

Table 1.
Demographic and participation variables of the participants.^a

Variable Athletics Volleyball

Juniors Adults Juniors Adults

n = 147 (50 m, 97 f)^b n = 47 (18 m, 29 f)^b n = 47 (15 m, 32 f)^b n = 95 (42 m, 53 f)^b

M SD M SD M SD M SD

Age (years)

Current 16.8 1.3* 24.8 5.8 18.0 1.0 26.8 6.1

Start practice^c 8.6 2.7 9.1 4.0 9.6 1.9 10.1 2.9

Start competitions^c 9.4 2.6 9.9 3.6 10.5 1.7 10.9 2.8

Practice amount (hours)^c

Until age 15 years 1791 854 1551 1323 1637 842 1448 1075

Age 16–19 years 2577 916 2405 1655 1906 796 2053 1033

Age 20–25 years 4913 2528 2868 1559

^a
Means and standard deviations (M, SD).

^b
m = male, f = female, no participant reported LGBT+ identity.

^c
Starting age and cumulative practice amounts in the main sport (athletics, volleyball).

Of the entire sample, 309 participants (92%) achieved top-10 placings at national junior championships; 74 (22%) achieved a top-10 result and 30 (9%) won medals at international junior championships. Of the 142 adult athletes, 114 (80%) were placed among the top 10 at open-age national championships; 53 (37%) achieved international top-10 placings, and 34 (24%) won medals at Olympic Games, world, or European championships.

Data collection

We used a performance and participation history questionnaire that had been validated in equivalent elite samples^28–30 (retest reliability 0.80 < r_tt < 1.00; external validity 0.81 < r < 1.00). A briefing sheet informed the athletes about the purpose of the study, voluntariness and anonymity of participation and data protection. The participants consented by opening the link to the questionnaire. It took the athletes on average M = 39 minutes (standard deviation SD = 15) to complete the questionnaire. Ethical approval was provided by the authors’ faculty ethics committee.

The cross-sectional retrospective questionnaire measured, amongst others, athletes’ socio-demographic variables, competition level and placing within each age category, training volume and participation in several performance analytics procedures (see Table 2). The questionnaire was complemented by relevant questions on athlete codetermination of training control adopted from the Sport Climate Questionnaire (SCQ).³¹ In addition, we conducted preliminary interviews with nine national-squad coaches from both sports and pilot tests involving eight athletes. Based on these insights, we developed indicators of training appropriateness, the extent of individualisation and, for the athletics subsample, the occurrence of readjustment of the training design to an athlete's current constitution. For all these questions, the question stem, format and response scale was kept identical to the SCQ. Items were arranged in random order. All variables were measured for two age periods: 16 to 19 years and 20 to 25 years. Table 2 describes the recorded variables, dimensions, operational definitions and internal consistency of the multi-item scales (Cronbach's α).

Table 2.
Recorded variables, dimensions and internal consistency (Cronbach’s α).

Dimension Variables, operational definition Cronbach's α

Performance Highest championship level – Olympic Games, world, European, national, regional, district championship – and placing within each age category: under 11, 12–13, 14–15, 16–17, 18–19, 20–21, 22–25 and 26 + years. Each athlete's highest performance within each age category was computed by a 3-digit code: 1^st digit = championship level (1 = Olympic Games/world championship, 2 = European, 3 = national, 4 = regional, 5 = state, 6 = district championships, 7 = below), 2^nd and 3^rd digit = placing. n.a.

Relative performance development Improvement (plus > 3 ranks) versus stagnation/decrease of the athlete's highest championship level and placing. For two age periods: 15 to 19 years and 19 to 25 years. n.a.

Appropriateness of the training ‘What and how we have trained was exactly right for me personally’. ‘What and how we have trained was not quite the right thing for me personally’. ‘The types of exercises/training forms we did were exactly right for me personally’. ‘Certain types of exercises/training forms we did were not quite the right thing for me personally’. ‘The way we did the exercises/training forms was exactly right for me personally’. ‘The way we did certain exercises/training forms was not quite the right thing for me personally’. ‘The strength of the training load was exactly right for me personally’. ‘The strength of the training load was not quite the right thing for me personally’. In each case for two age periods: 16 to 19 years and 20 to 25 years. 5-point scale: 0 = never/almost never … 4 = always/almost always. 16–19 years:α = 0.9020–25 y.:α = 0.91

Individualisation of the training ‘The training was … exactly tailored to my individual wishes and interests; … exactly tailored to my individual strengths and weak points; … mainly aimed at expanding my individual strengths’. In each case for two age periods: 16 to 19 years and 20 to 25 years. 5-point scale: 0 = never/almost never … 4 = always/almost always. 16–19 years:α = 0.8520—25 y.:α = 0.86

Readjustment of the training design (only for athletics) ‘The training load … was exactly adjusted to my current physical constitution/shape-on-the-day; … was exactly adjusted to my current health-related constitution’. ‘I shared with my coach … how my current physical constitution was; … how my current health-related constitution was’. ‘My coach and I spoke about my current constitution’. In each case for two age periods: 16 to 19 years and 20 to 25 years. 5-point scale: 0 = never/almost never … 4 = always/almost always. 16–19 years:α = 0.8120—25 Y.:α = 0.81

Athlete codetermination of the training control (adopted from the Sport Climate Questionnaire, SCQ³¹⁾ ‘I feel that my coach provides me choices and options’. ‘My coach encouraged me to ask questions’. ‘My coach listens to how I would like to do things’. ‘My coach tries to understand how I see things before suggesting a new way to do things’. In each case for two age periods: 16 to 19 years and 20 to 25 years. 5-point scale: 0 = never/almost never … 4 = always/almost always. 16–19 years:α = 0.8420—25 y.:α = 0.81

Participation in performance analytics Whether or not an athlete participated in the following procedures and, if yes, annual frequency of participation: sports-medical health assessment, video analysis, biomechanics, kinesiology, physiology and exercise science. In each case for two age periods: 16 to 19 years and 20 to 25 years. n.a.

Validity and reliability

An explorative factor analysis (varimax rotation; 70% cleared total variance) extracted three factors (Kaiser criterion) with a clear dimensional structure. The model had high factor loadings of designated items and low cross-loadings on other factors: individualisation (factor loadings: 0.72 to 0.86, cross-loadings: −0.09 to 0.22), readjustment of the training design (factor loadings: 0.61 to 0.78, cross-loadings: 0.08 to 0.34), and athlete codetermination (factor loadings: 0.68 to 0.85, cross-loadings: −0.13 to 0.12). The internal consistency (Cronbach's α) of each scale was acceptable (0.80 < α < 0.92; Table 2). Leave-one-out analyses did not provide any improvement of any scale. Furthermore, a retest after 4 weeks (n = 21) showed acceptable retest reliability for all variables (0.74 < r_tt < 1.00).

The questions on the appropriateness of the training referred to appropriateness for athletes’ performance development. To test external validity, we differentiated between athletes with multi-year relative performance improvement (championship level and placing) and those with performance stagnation or decrease (see Table 2) and compared their reports on training appropriateness (5-point scale from 0 to 4). Within junior age (16 to 19 years), the values were M = 3.14 (SD = 0.58) versus M = 2.57 (SD = 0.75), t = 6.01, p < .001, Cohen's d = 0.83. For adult age (20 to 25 years), the values were M = 3.10 (SD = 0.74) versus M = 2.27 (SD = 0.74), t = 4.72, p < .001, Cohen's d = 1.13. These results corroborated that athletes’ reports on their training appropriateness reflected their performance development.

Furthermore, it is theoretically conceivable that perceiving greater extents of individualisation, readjustment and athlete codetermination results from better performance development, or vice versa—another relevant point regarding validity. In the first case, it would be expected that athletes with multi-year performance improvement differed from those with performance stagnation or decrease by showing a more positive change over time of reported individualisation, readjustment and athlete codetermination. The present data enabled the testing of this hypothesis: The groups with performance improvement and with performance stagnation/decrease from age 15 to 19 years were compared regarding their development of individualisation, readjustment and athlete codetermination from age period 16–19 years to age period 20–25 years. Repeated measures analysis of variance (ANOVA) revealed that the groups did not differ in the development of individualisation (F = 0.14, p = .709) and readjustment (F = 0.91, p = .346). In terms of athlete codetermination, athletes with performance stagnation/decrease from 15 to 19 years old even showed an increase over time (16 to 19 years: M = 2.03, SD = 0.90; 20 to 25 years: M = 2.48, SD = 0.70), while the participants with performance improvement did not (16 to 19 years: M = 2.56, SD = 0.92; 20 to 25 years: M = 2.56, SD = 0.81; F = 4.33, p = .041). The finding suggests that perceiving greater levels of individualisation, readjustment or athlete codetermination was not a function of preceding positive performance development.

Finally, it is relevant that participants with multi-year relative performance improvement did not engage in greater amounts of sport-specific practice than did their counterparts with relative performance stagnation/decrease (15 to 19 years: improvement M = 2275 h, SD = 1022, vs. stagnation/decrease M = 2331 h, SD = 1298, t = 0.25, p = .805; 20 to 25 years: M = 3554 h, SD = 2106, vs. M = 3764 h, SD = 2941, t = 0.31, p = .758). This means that differences in multi-year performance development reflected differences in multi-year efficiency of practice (Δperformance/practice amount).

Data analysis

Analyses were performed with SPSS 26.0 (IBM, Armonk, NY) and included three progressive steps:
To describe the extent of individualisation, readjustment and athlete codetermination of training control and compare them between athletics and volleyball.

To analyse whether and how individualisation, readjustment and athlete codetermination explained individual differences in the appropriateness of athletes’ training using multivariate linear regression analyses (MLR).

This step included two additional sub-analyses: (a) It is conceivable that older and/or more successful athletes were generally more likely granted greater extents of individualisation, readjustment and codetermination,^32–34 implying that individualisation, readjustment and codetermination followed success, not vice versa. To begin to investigate this hypothesis, we subsequently added athletes’ age and performance level as predictor variables. (b) To investigate whether performance analytics had an additional, specific effect on training appropriateness, we added the participation frequency of each athlete in each of the performance analytics procedures as predictor variables (see Table 2).

To compare athletes who showed multi-year relative performance improvement with those who had performance stagnation/decrease (i.e., competition level: highest championship level and placing; from 15 to 19 years and from 19 to 25 years) regarding the extent of individualisation, readjustment, and athlete codetermination of training control.

The performance developments (improvement vs. stagnation/decrease) from 15 to 19 years old and from 19 to 25 years old were widely unrelated (χ² = 0.13, p = .803). Likewise, training appropriateness, individualisation, readjustment and athlete codetermination, respectively, during age periods 16 to 19 years and 20 to 25 years were only modestly correlated (0.25 < r_tt < 0.46). That is, these variables were not stable characteristics of an athlete but varied considerably in the course of their career, demanding separate analyses for each age period. Male and female participants did not significantly differ from each other in any of the variables (0.03 < t < 1.99; all p > .05), and there were no interactions of sex and performance development (0.01 < F < 2.49; p > .05). Both sexes were thus pooled in analyses. Furthermore, there was no interaction of sport (athletics, volleyball) and performance development (0.00 < F < 1.58; p > .05). Thus, both sports were pooled for analyses 2 and 3 described above.

Descriptive data include means and standard deviations (M, SD). Group comparisons (athletics vs. volleyball; performance improvement vs. stagnation/decrease) were done using an unpaired t-test. MLR analyses were conducted with backwards variable inclusion (inclusion criterion p < .10). Effect sizes are expressed as Cohen's d for group comparisons and as adjusted R² (R² _adj ) for MLR. All hypothesis testing was two-tailed. A value of p < .05 was considered statistically significant.

Results

Table 3 shows the extent of individualisation, readjustment and athlete codetermination of training control, and the comparison between athletics and volleyball. Sample and subsample mean values ranged from M = 2.17 to M = 2.98 on the 5-point scale (0 = ‘never/almost never’ to 4 = ‘always/almost always’). That is, both during youth and adulthood, large parts of the participants’ training were highly individualised, readjusted to the athlete's current constitution, and codetermined by the athlete. Individualisation and athlete codetermination were more pronounced in athletics than in volleyball, both during youth and adulthood. Effect sizes were medium to large (Table 3).

Table 3.
Extent of individualisation, readjustment and athlete codetermination of training control^a of high-performance athletes during age periods 16 to 19 and 20 to 25 years and comparison between athletics and volleyball.

Variable Overall Athletics Volleyball Sports comparison

M SD M SD M SD t p d ^b

Individualisation

Age 16–19 years 2.49 0.86 2.71 0.80 2.17 0.84 5.93 <.001 0.67

20–25 years 2.41 0.92 2.75 0.86 2.24 0.90 3.21 .002 0.58

Readjustment

Age 16–19 years 2.98 0.77

20–25 years 2.95 0.78

Codetermination

Age 16–19 years 2.60 0.94 2.87 0.89 2.23 0.87 6.57 <.001 0.73

20–25 years 2.49 0.77 2.89 0.64 2.29 0.76 4.60 <.001 0.83

^a
5-point scales: 0 = ‘never/almost never’ … 4 = ‘always/almost always’.

^b
d = Cohen's d.

The results of the MLR analyses are displayed in Table 4. More than half of the variance of the training appropriateness (0.53 < R² _adj < 0.58) was explained by either individualisation alone (Model 3: adult training, both sports), individualisation and athlete codetermination (Model 1: youth training, both sports; Model 4: adult training, athletics), or individualisation, readjustment, and athlete codetermination combined (Model 2: youth training, athletics).

Table 4.
Results of multivariate linear regression analyses (MLR). Dependent variable: Appropriateness of athletes’ training during two age periods, 16 to 19 years (top) and 20 to 25 years (bottom). Backward inclusion method (inclusion criterion p < 0.10); predictor variables in Models 1 and 3: Individualisation and athlete codetermination; in Models 2 and 4: Individualisation, readjustment and athlete codetermination. R² _adj = adjusted R².

β p

Dependent variable: training appropriateness at 16–19 years

Model 1: F_{(2, 314)} = 198.28, p < .001, R²_adj = 0.56

Individualisation 16–19 years 0.66 <.001

Athlete codetermination 16–19 years 0.18 <.001

Model 2 (only athletics): F_{(3, 173)} = 67.49, p < .001, R²_adj = 0.53

Individualisation 16–19 years 0.54 <.001

Readjustment 16–19 years 0.18 .005

Athlete codetermination 16–19 years 0.18 .002

Dependent variable: training appropriateness at 20–25 years

Model 3: F_{(1, 128)} = 170.85, p < .001, R²_adj = 0.57

Individualisation 20–25 years 0.76 <.001

Model 4: (only athletics): F_{(2, 35)} = 23.70, p < .001, R²_adj = 0.55

Individualisation 20–25 years 0.58 <.001

Athlete codetermination 20–25 years 0.27 .047

When adding the athletes’ age and/or performance level as predictor variables to each model (individually or combined), they were non-significant (p > .10) and were excluded from the model. Thus, the effects displayed in Models 1 to 4 applied regardless of athletes’ age and performance level. Furthermore, when adding the participation in the various performance analytics procedures as additional predictor variables to each of the models in Table 4, they were, again, non-significant (p > .10) and were excluded from the model. That is, when considering individualisation, readjustment and athlete codetermination, participation in performance analytics had no additional, specific explanatory effect on individual differences in training appropriateness.

Table 5 shows the comparison between athletes with multi-year relative performance improvement and those with performance stagnation/decrease (championship level and placing), both during youth and adulthood. Athletes with multi-year performance improvement reported greater extents of individualisation, readjustment and athlete codetermination of training control. Effect sizes were generally small to medium for junior age and large for individualisation and readjustment in adulthood. The effect size of athlete codetermination in adulthood was similar to that during youth age, but only approached significance (p = .075).

Table 5.
Comparison of athletes with multi-year relative performance improvement and stagnation/decrease (championship level and placing) from age 15 to 19 and 19 to 25 years^a with regard to the extent of individualisation, readjustment and athlete codetermination of training control.^b

Improvement Stagnation/decrease Comparison

M SD M SD t p d

Development age 15 to 19 years

Individualisation 2.79 0.72 2.29 0.82 4.64 <.001 0.64

Readjustment 3.16 0.57 2.83 0.81 2.77 .006 0.48

Athlete codetermination 2.78 0.85 2.48 0.92 2.44 .016 0.34

Development age 19 to 25 years

Individualisation 2.92 0.74 2.16 0.79 4.13 <.001 0.99

Readjustment 3.22 0.53 2.53 1.00 2.45 .020 0.85

Athlete codetermination 2.79 0.76 2.47 0.69 1.81 .075 0.43

^a
Age 15 to 19 years: improvement n = 100, stagnation/decrease n = 109; age 19 to 25 years: improvement n = 32, stagnation/decrease n = 39. Other cases not included because multi-year performance development fluctuated across age categories.

^b
5-point scale: 0 = ‘never/almost never’ … 4 = ‘always/almost always’.

Discussion

Athletes’ training varies in the ratio between commonality across athletes versus individualisation (i.e., individual differences) across athletes, adherence to a training plan versus readjustment to the athlete's present constitution, and between coach determination versus athlete codetermination of the training design and control. We explored the extent to which the training of a select sample of elite performers from athletics and volleyball was characterised by individualisation, readjustment and athlete codetermination. Four central findings emerged: (1) Large parts of the training of these high-performance athletes were characterised by high degrees of individualisation, readjustment and athlete codetermination. (2) These characteristics were more pronounced in athletics than in volleyball. (3) Individualisation, readjustment and athlete codetermination of training control facilitated training appropriateness, irrespective of athletes’ age and performance level. Consistently, higher levels of individualisation, readjustment and athlete codetermination of training control were associated with better multi-year performance development. (4) Participation in performance analytics had no additional, specific effect on training appropriateness.

The observed consistent predictor effects on the athlete-assessed training appropriateness and on the factual development of athletes’ competition level – both during youth and adult age – corroborate the robustness and nomological validity of the findings. Furthermore, differences in multi-year performance development were not explained by differences in athletes’ training volume: Participants with multi-year performance improvement and with performance stagnation/decrease did not significantly differ in their cumulative sport-specific training amounts. The data signify that greater extents of individualisation, readjustment and athlete codetermination were associated with both greater training effectiveness and greater training efficiency (performance improvement per training amount).³⁵

The present findings extend and partly challenge some traditional views of the planning and control of the elite training process in four – presumably interrelated – regards. (1) Among these elite athletes, the coach–athlete power relationship was obviously not as asymmetric as some have postulated.^36–38 For example, participants reported that they codetermined most of their training. (2) Besides the coach's observation of their athletes and performance analytics,^1,2,17 the athlete's communication of their experiences, perceptions and preferences is apparently an important source to inform the appropriateness, effectiveness and efficiency of the training process. (3) Unlike the assumed premise of traditional training theory that planned and executed training are identical,^12,13,39 readjustment of the training design on a day-to-day basis was common. (4) Participation in performance analytics had no specific effect on training appropriateness, which questions the traditional view of the ‘cybernetic loop’ of performance analytics and training control.^12,21 In sum, these observations together with extant research^8–10 suggest that coaches and athletes form a framework concept for a defined period of training that has to be understood as a flexible, somewhat ‘fluid’ hypothesis. This hypothesis is recurrently tested and readjusted based on day-to-day experiences and observations. In this context, the athlete's experiences, perceptions and preferences provide relevant information to inform training appropriateness, which the other sources – coach observation and performance analytics – cannot equivalently provide. The athlete's experiences, perceptions, and preferences are partly subjective, but are highly ‘fine-grained’ and holistic, in that they integrate perspectives of physiology, medicine, kinesiology, psychology, and pedagogy (e.g. Saw et al.,¹⁹ Lazarus,⁴⁰ McNair et al.,⁴¹ Marsh et al.⁴²)

In former literature, the coach–athlete interaction and athlete codetermination have mainly been addressed with respect to the athlete's basic psychological needs satisfaction, well-being and motivation.^23,24,43 The present observations suggest that the effect of athlete codetermination on performance development is not contradictory to the effects on the athlete's psychological well-being and motivation. The effects are rather concordant: greater extents of athlete codetermination may facilitate both, the athlete's psychological well-being and motivation as well as performance progress. However, whether these are unrelated or related effects is still widely unstudied – for example, whether the effect on psychological well-being and motivation is moderated by the effect on performance progress or vice versa.

As expected, all three dimensions – individualisation, readjustment and athlete codetermination of training control – were more pronounced in athletics than in volleyball. Unlike typical individual sports, team sports like volleyball require joint performance. Each player's actions rely on and are aligned with teammates’ actions. To achieve this, playing systems, tactics and plays require joint practice of common tasks.²⁷ This requirement presumably leaves less scope for individualisation, readjustment and athlete codetermination of training control. Furthermore, some reports suggesting asymmetrical coach–athlete power relationships – which may limit athlete codetermination – were from team game sports.^37,44,45

Finally, the effect size of athlete codetermination of training control on adult performance development was similar to the effect during youth age, but was non-significant. It may be that the effect is in fact less reliable in adulthood than during youth age. It should be considered, however, that the adult subsample was comparatively small (n = 71) and below the minimal sample size suggested by the a-priori power analysis (n ≥ 191).

Practical implications

The present findings suggest some clear implications for practitioners. To facilitate the athlete's performance progress, both during youth and adulthood, coaches should seek to tailor large parts of the training to athletes’ individual characteristics, adjust the training to the athlete's current constitution, and involve the athlete in decision-making on the training design. This implies to encourage athletes to share their experiences, perceptions and preferences as well as to develop their confidence and competencies to do so (e.g., by facilitating and deliberately training self-perception). Furthermore, these topics may be integrated in coach education.

In all this, it should be noted, however, that individualisation, readjustment and athlete codetermination did not apply to athletes’ entire training. Even athletes with better performance development reported that considerable parts of their training were determined by the coach alone, uniform across athletes and adhered to the original training plan.

Limitations and future directions

The study has several strengths, such as exploring a new subject, involving a select elite sample and combining athlete-assessed training appropriateness with participants’ factual performance development. However, the study does have limitations. The major limitation is that the study is correlational and, strictly speaking, does not allow us to draw causal conclusions. Furthermore, the adult subsample was comparatively small and below the size suggested by the power analysis. Also, athletes estimated training appropriateness, individualisation, readjustment, and codetermination over relatively long time periods. In addition, although reliable and valid, the retrospective recording of several variables implied the common potential limitations in power, selection effects (e.g. survivor bias), and possibly recall bias. Finally, the participants competed at a very high level. The scope of the findings may not extend to younger athletes or lower performance levels.

The goal for future work is to extend the present research questions to larger adult samples, to other sports and perhaps to younger and lower performing athletes. The characteristics of training control may be recorded over shorter timescales. This may, for example, include observations during the preparation and competition period within a season. Furthermore, it will be interesting to investigate whether and how the concordant effects of athlete codetermination on psychological well-being, motivation, and performance progress are related: e.g., whether the effect on performance progress moderates the effect on psychological well-being and motivation or vice versa. Finally, and perhaps most importantly, future research may explore two pursuing questions: First, in which way do individualisation, readjustment and athlete codetermination occur in the coach–athlete interaction? And second, which aspects of the training process and what ranges of variation do they concern (e.g., types of exercise and game formats, ways of executing them, volume and intensity of loads, etc.)? To begin to explore these aspects, qualitative coach and athlete interviews may be a promising approach.

Variable	Athletics	Volleyball
Age (years)
Current	16.8	1.3	24.8	5.8	18.0	1.0	26.8	6.1
Start practice^c	8.6	2.7	9.1	4.0	9.6	1.9	10.1	2.9
Start competitions^c	9.4	2.6	9.9	3.6	10.5	1.7	10.9	2.8
Practice amount (hours)^c
Until age 15 years	1791	854	1551	1323	1637	842	1448	1075
Age 16–19 years	2577	916	2405	1655	1906	796	2053	1033
Age 20–25 years			4913	2528			2868	1559

Dimension	Variables, operational definition	Cronbach's α
Performance	Highest championship level – Olympic Games, world, European, national, regional, district championship – and placing within each age category: under 11, 12–13, 14–15, 16–17, 18–19, 20–21, 22–25 and 26 + years. Each athlete's highest performance within each age category was computed by a 3-digit code: 1^st digit = championship level (1 = Olympic Games/world championship, 2 = European, 3 = national, 4 = regional, 5 = state, 6 = district championships, 7 = below), 2^nd and 3^rd digit = placing.	n.a.
Relative performance development	Improvement (plus > 3 ranks) versus stagnation/decrease of the athlete's highest championship level and placing. For two age periods: 15 to 19 years and 19 to 25 years.	n.a.
Appropriateness of the training	‘What and how we have trained was exactly right for me personally’. ‘What and how we have trained was not quite the right thing for me personally’. ‘The types of exercises/training forms we did were exactly right for me personally’. ‘Certain types of exercises/training forms we did were not quite the right thing for me personally’. ‘The way we did the exercises/training forms was exactly right for me personally’. ‘The way we did certain exercises/training forms was not quite the right thing for me personally’. ‘The strength of the training load was exactly right for me personally’. ‘The strength of the training load was not quite the right thing for me personally’. In each case for two age periods: 16 to 19 years and 20 to 25 years. 5-point scale: 0 = never/almost never … 4 = always/almost always.	16–19 years:α = 0.9020–25 y.:α = 0.91
Individualisation of the training	‘The training was … exactly tailored to my individual wishes and interests; … exactly tailored to my individual strengths and weak points; … mainly aimed at expanding my individual strengths’. In each case for two age periods: 16 to 19 years and 20 to 25 years. 5-point scale: 0 = never/almost never … 4 = always/almost always.	16–19 years:α = 0.8520—25 y.:α = 0.86
Readjustment of the training design (only for athletics)	‘The training load … was exactly adjusted to my current physical constitution/shape-on-the-day; … was exactly adjusted to my current health-related constitution’. ‘I shared with my coach … how my current physical constitution was; … how my current health-related constitution was’. ‘My coach and I spoke about my current constitution’. In each case for two age periods: 16 to 19 years and 20 to 25 years. 5-point scale: 0 = never/almost never … 4 = always/almost always.	16–19 years:α = 0.8120—25 Y.:α = 0.81
Athlete codetermination of the training control (adopted from the Sport Climate Questionnaire, SCQ³¹⁾	‘I feel that my coach provides me choices and options’. ‘My coach encouraged me to ask questions’. ‘My coach listens to how I would like to do things’. ‘My coach tries to understand how I see things before suggesting a new way to do things’. In each case for two age periods: 16 to 19 years and 20 to 25 years. 5-point scale: 0 = never/almost never … 4 = always/almost always.	16–19 years:α = 0.8420—25 y.:α = 0.81
Participation in performance analytics	Whether or not an athlete participated in the following procedures and, if yes, annual frequency of participation: sports-medical health assessment, video analysis, biomechanics, kinesiology, physiology and exercise science. In each case for two age periods: 16 to 19 years and 20 to 25 years.	n.a.

Variable	Overall	Athletics	Volleyball	Sports comparison
Individualisation
Age 16–19 years	2.49	0.86	2.71	0.80	2.17	0.84	5.93	<.001	0.67
20–25 years	2.41	0.92	2.75	0.86	2.24	0.90	3.21	.002	0.58
Readjustment
Age 16–19 years			2.98	0.77
20–25 years			2.95	0.78
Codetermination
Age 16–19 years	2.60	0.94	2.87	0.89	2.23	0.87	6.57	<.001	0.73
20–25 years	2.49	0.77	2.89	0.64	2.29	0.76	4.60	<.001	0.83

	β	p
Dependent variable: training appropriateness at 16–19 years
Model 1: F_{(2, 314)} = 198.28, p < .001, R²_adj = 0.56
Individualisation 16–19 years	0.66	<.001
Athlete codetermination 16–19 years	0.18	<.001
Model 2 (only athletics): F_{(3, 173)} = 67.49, p < .001, R²_adj = 0.53
Individualisation 16–19 years	0.54	<.001
Readjustment 16–19 years	0.18	.005
Athlete codetermination 16–19 years	0.18	.002
Dependent variable: training appropriateness at 20–25 years
Model 3: F_{(1, 128)} = 170.85, p < .001, R²_adj = 0.57
Individualisation 20–25 years	0.76	<.001
Model 4: (only athletics): F_{(2, 35)} = 23.70, p < .001, R²_adj = 0.55
Individualisation 20–25 years	0.58	<.001
Athlete codetermination 20–25 years	0.27	.047

	Improvement	Stagnation/decrease	Comparison
Development age 15 to 19 years
Individualisation	2.79	0.72	2.29	0.82	4.64	<.001	0.64
Readjustment	3.16	0.57	2.83	0.81	2.77	.006	0.48
Athlete codetermination	2.78	0.85	2.48	0.92	2.44	.016	0.34
Development age 19 to 25 years
Individualisation	2.92	0.74	2.16	0.79	4.13	<.001	0.99
Readjustment	3.22	0.53	2.53	1.00	2.45	.020	0.85
Athlete codetermination	2.79	0.76	2.47	0.69	1.81	.075	0.43

Footnotes

Acknowledgement

The authors would like to express their sincere thanks to the sport directors and national coaches of the German Athletics Federation and the German Volleyball Federation for their support of the project. As well, the authors would like to thank the reviewers for valuable feedback on an earlier draft of this paper.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Arne Güllich

References

Rynne

Mallet

Rabjohns

MWO

. High performance coaching: demands and development. In: Thelwell

Harwood

Greenlees

(eds) The psychology of sports coaching: research and practice. London, New York: Routledge, 2017, pp.114–126.

Cushion

Nelson

Armour

, et al. Coach learning and development: a review of the literature. Leeds: Sports Coach UK, 2010.

Werthner

Trudel

. A new theoretical perspective for understanding how coaches learn to coach. Sport Psychol 2006; 20: 198–212.

Wenger

Snyder

. Communities of practice: the organizational frontier. Harv Bus Rev, (January_February), 2000; 56: 139–145.

Button

Seifert

Chow

, et al. Dynamics of skill acquisition. 2nd ed. Champaign, IL: Human Kinetics, 2020.

Hodges

Williams

(eds). Skill acquisition in sport. 2nd ed. London, New York: Routledge, 2012.

Meeusen

Duclos

Foster

, et al. Prevention, diagnosis and treatment of the overtraining syndrom: joint consensus statement of the European College of Sport Science (ECSS) and the American College of Sports Medicine (ACSM). Eur J Sport Sci 2013; 13: 1–24.

Allen

Bell

Lynn

, et al. Identifying excellent coaching practice along the sporting pathway. Leeds: Sports Coach UK, 2012.

Becker

. It's Not what they Do, It's How they Do It: athlete experiences of great coaching. Int J Sports Sci Coach 2009; 4: 93–119.

10.

Nash

Sproule

Horton

. Excellence in coaching: the at and skill of elite practitioners. Res Q Exerc Sport 2011; 82: 229–238.

11.

Fleck

Kraemer

. Periodization breakthrough. New York: Advanced Research Press, 1996.

12.

Issurin

. New horizons for the methodology and physiology of training periodization. Sports Med 2010; 40: 189–206.

13.

Platonov

. The system of athletes’ preparation in Olympic sport. In: Textbook for coaches (in Russian). Kiev: Olympic Literature, 2004.

14.

Bouchard

Rankinen

. Individual differences in response to regular physical activity. Med Sci Sports Exerc 2001; 33: 446–451.

15.

Hardy

Barlow

Evans

, et al. Great British medalists: psychological biographies of super-elite and elite athletes from Olympic sports. Prog Brain Res 2017; 232: 1–119.

16.

Bourdon

Cardinale

Murray

, et al. Monitoring athlete training loads: consensus statement. Int J Sports Physiol Perform 2017; 12: S2161–S2170.

17.

Gomez-Ruano

Ibáñez

Leicht

. Editorial: performance analysis in sport. Front Psychol 2020; 11: 611–634.

18.

Sands

Kavanaugh

Murray

, et al. Modern techniques and technologies applied to training and performance monitoring. Int J Sports Physiol Perform 2017; 12: 63–72.

19.

Saw

Main

Gastin

. Monitoring the athlete training response: subjective self-reported measures trump commonly used objective measures: a systematic review. Br J Sports Med 2016; 50: 281–291.

20.

Güllich

Emrich

. Individualistic and collectivistic approach in athlete support programmes in the German high-performance sport system. Eur J Sport Soc 2012; 9: 243–268.

21.

DeWeese

Hornsby

Stone

, et al. The training process: planning for strength–power training in track and field. Part 1: theoretical aspects. J Sport Health Sci 2015; 4: 308–317.

22.

Hardy

Laing

Barlow

, et al. Comparison of the biographies of GB serial medal and non-medalling Olympic athletes. London: UK Sport, 2013.

23.

Mageau

Vallerand

. The coach–athlete relationship: a motivational model. J Sports Sci 2003; 21: 883–904.

24.

Nichol

Hall

Vickery

, et al. Examining the relationship between coaching practice and athlete “outcomes”: a systematic review and critical realistic critique. Int Sport Coach J 2019; 6: 13–29.

25.

Macpherson

McLaren

Gregson

, et al. Using differential ratings of perceived exertion to assess agreement between coach and player perceptions of soccer training intensity: an exploratory investigation. J Sports Sci 2019; 37: 2783–2788.

26.

Fédération Internationale de Volleyball. Coaches manual. Lausanne: FIVB, 2011.

27.

Thompson

PJL

. Introduction to coaching. In: The official IAAF guide to coaching athletics. Lincolnshire, UK: Warners Midlands plc, 2009.

28.

Güllich

Emrich

. Investment patterns in the careers of elite athletes in East and West Germany. Eur J Sport Sci 2013; 10: 191–214.

29.

Güllich

Emrich

. Considering long-term sustainability in the development of world class success. Eur J Sport Sci 2014; 14: S383–S397.

30.

Hornig

Aust

Güllich

. Practice and play in the development of German top-level professional football players. Eur J Sport Sci 2016; 16: 96–105.

31.

Deci

. The sport climate questionnaire: www.selfdeterminationtheory.org/questionnaires/10-questionnaires/84, retrieved May 20, 2008, 2001.

32.

Kassing

Anderson

. Contradicting coach or grumbling to teammates: exploring dissent expression in the coach-athlete relationship. Commun Sport 2013; 2: 172–185.

33.

Partington

Cushion

Harvey

. An investigation of the effect of athletes’ age on the coaching behaviours of professional top-level youth soccer coaches. J Sports Sci 2014; 32: 403–414.

34.

Turman

. Athletes’ perception of coach power use and the association between playing status and sport satisfaction. Commun Res 2006; 24: 273–282.

35.

Barth

Güllich

. Non-linear association of efficiency of practice of adult elite athletes with their youth multi-sport practice. J Sports Sci 2020; 39: 915–925.

36.

Cruickshank

Collins

. Culture change in elite sport performance teams: examining and advancing effectiveness in the New Era. J Appl Sport Psychol 2012; 24: 338–355.

37.

Cushion

Jones

. Power, discourse, and symbolic violence in professional youth soccer: the case of Albion football club. Sociol Sport J 2006; 23: 142–161.

38.

Potrac

Jones

. Power, conflict, and cooperation: toward a micropolitics of coaching. Quest 2009; 61: 223–236.

39.

Bompa

(ed). Periodization training for sports. Champaign IL: Human Kinetics, 1999.

40.

Lazarus

. How emotions influence performance in competitive sports. Sport Psychol 2000; 14: 229–252.

41.

McNair

Lorr

Droppleman

. Profile mood states manual. San Diego, CA: Educational and Industrial Testing Service, 1971.

42.

Marsh

Redmayne

RSA

. Multidimensional physical self-concept and its relations to multiple components of physical fitness. J Sport Exerc Psychol 1994; 16: 43–55.

43.

Cronin

Allen

. Developmental experiences and well-being in sport: the importance of the coaching climate. Sport Psychol 2015; 29: 62–71.

44.

Groom

Cushion

Nelson

. The delivery of video-based performance analysis by England youth soccer coaches: towards a grounded theory. J Appl Sport Psychol 2011; 23: 16–32.

45.

Nelson

Potrac

Groom

. Receiving video-based feedback in elite ice-hockey players: a player's perspective. Sport Educ Soc 2014; 19: 19–40.

Variable	Athletics				Volleyball
	Juniors		Adults		Juniors		Adults
	n = 147 (50 m, 97 f)^b		n = 47 (18 m, 29 f)^b		n = 47 (15 m, 32 f)^b		n = 95 (42 m, 53 f)^b
	M	SD	M	SD	M	SD	M	SD
	Age (years)
Current	16.8	1.3	24.8	5.8	18.0	1.0	26.8	6.1
Start practice^c	8.6	2.7	9.1	4.0	9.6	1.9	10.1	2.9
Start competitions^c	9.4	2.6	9.9	3.6	10.5	1.7	10.9	2.8
Practice amount (hours)^c
Until age 15 years	1791	854	1551	1323	1637	842	1448	1075
Age 16–19 years	2577	916	2405	1655	1906	796	2053	1033
Age 20–25 years			4913	2528			2868	1559

Variable	Overall		Athletics		Volleyball		Sports comparison
Variable	M	SD	M	SD	M	SD	t	p	d ^b
Individualisation
Age 16–19 years	2.49	0.86	2.71	0.80	2.17	0.84	5.93	<.001	0.67
20–25 years	2.41	0.92	2.75	0.86	2.24	0.90	3.21	.002	0.58
Readjustment
Age 16–19 years			2.98	0.77
20–25 years			2.95	0.78
Codetermination
Age 16–19 years	2.60	0.94	2.87	0.89	2.23	0.87	6.57	<.001	0.73
20–25 years	2.49	0.77	2.89	0.64	2.29	0.76	4.60	<.001	0.83