Assessing the practicality of using freely available AI-based GPT tools for coach learning and athlete development

Area of Science:

• Sports Science
• Artificial Intelligence
• Human-Computer Interaction

Background:

• Coach-athlete dialogue is crucial for performance and development.
• Analyzing these conversations traditionally requires significant time and resources.
• Advancements in AI offer potential for more efficient analysis methods.

Purpose of the Study:

• To evaluate the efficacy of freely available AI tools (ChatGPT v4, DeepSeek v3) for analyzing coach-athlete dialogue.
• To assess the utility of a prompt-based analysis framework (R²-PIASS) with these AI tools.
• To explore the implications of AI-driven dialogue analysis for coaching practice and athlete development.

Main Methods:

• Utilized two AI models: ChatGPT v4 and DeepSeek v3.
• Applied a pre-determined, context-specific, prompt-based analysis framework (R²-PIASS).
• Analyzed a selected coach-athlete dialogue dataset using the specified AI tools and framework.

Main Results:

• Both ChatGPT v4 and DeepSeek v3 successfully extracted quantitative and qualitative conversational data using R²-PIASS prompts.
• AI platforms demonstrated efficiencies in cost, usability, accessibility, and convenience for coaches.
• Identified risks included the robustness of statistical outcomes and data privacy concerns.

Conclusions:

• Freely available AI tools show promise for analyzing coach-athlete dialogue, offering practical benefits for coaches.
• Further research is necessary to address limitations concerning statistical validity and data privacy.
• AI-driven analysis has the potential to enhance coach learning and athlete development if implemented carefully.