How fast-and-frugal trees can inform diagnostic and intervention decisions for enhancing elite athlete performance

The key to fostering the individual potential of an elite athlete lies in deciding what to prioritize in training. Heuristic decision tools such as fast-and-frugal trees (FFTrees) have proven to be effective and suitable for identifying promising determinants in this context. FFTrees are binary decision trees that can make decisions based on only one reason. The objective of this study was to examine the applicability of FFTrees to inform intervention decisions in elite athletes. We aimed to create FFTrees and evaluate their ability to determine individually beneficial interventions. We collected cognitive, psychosocial, and motor-performance diagnostic data from 466 German elite athletes in different sports disciplines. First, we used principal component analysis to identify components representing types of interventions across sports. These served as cues for the FFTrees. As a result, the PCA identified six cues. Two sport-specific FFTrees were created using these six cues. One FFTree was created for trampoline with four cues (relative grip strength, motor cost, motor inhibition, visual selective attention) and 90% correct predictions. The other FFTree was created for volleyball with four cues (motor inhibition, motor cost, countermovement jump, Y-Balance Test) and 75% correct predictions. To conclude, the high accuracy confirms that FFTrees enable data-based decisions for interventions based on sport-specific demands and the preferences of coaches. We argue that FFTrees are beneficial in projects collecting multidisciplinary variables for personalizing interventions in elite athletes. Coaching practice benefits from using FFTrees by providing reference values when an intervention could enhance performance. In the future, we advocate that team sports develop position-specific FFTrees. In conclusion, FFTrees empower decision-makers by efficiently identifying athletes' adaptation potentials.