Toward a Differentiated Model of Athlete Performance Development (DMAPD): Preliminary version and testing based on a German sample representing six sports.

When is the right time to start competing? Should young athletes sample multiple sports or specialize early? Despite ongoing debate, the literature lacks efforts to build and test parsimonious models that synthesize existing knowledge, test traditional hypotheses, and explore new ones. This study addresses this gap by introducing and testing a preliminary Differentiated Model of Athlete Performance Development (DMAPD), which proposes five hypotheses linking early childhood versatility, type of sports, age at competition start, sports sampling, and later competition performance. We analysed data from 943 German male youth athletes (t1: MAge = 8.03 years, SD = 0.59) across six sports (soccer, tennis, table tennis, cross-country skiing, long-distance running, road cycling), categorized into three sport types: invasion games, net games, and endurance sports, until the beginning of the mastery athletic career stage (Mfinal age = 20.06 years, SD = 0.69). Structural equation modeling, specifically moderated mediation analysis, was employed to test the DMAPD. Our findings indicate that for invasion and net games, early versatility predicts an earlier start in competitions, which is associated with higher later competition performance; no significant link was found between sports sampling and later performance. Conversely, in endurance sports, early versatility did not predict age at competition start, sports sampling, or later performance. Instead, later competition performance was solely predicted by the extent of sports sampling. These findings encourage further consideration of DMAPD hypotheses and align with recent calls for more differentiated, group-specific and model-based considerations in talent development research.