The focus of this study was on exploring decision-making in a triadic motor game (Bears, Monkeys and Ants), using T-Pattern analysis (TPA). The research involved 23 players aged 12 to 13 years (mean ± standard deviation; 12.5 ± 1), 12 girls and 11 boys, from a secondary middle-class school in Tenerife (Spain). An observational design was used (quadrant III): Nomothetic (a plurality of players), Punctual (a session recorded) and Multidimensional (different criteria). A mixed methods approach was used. It consisted of systematic observation to assess specific roles and subroles. A viewing instrument was built to observe the players ‘ad hoc’ while they were playing. To address the quality of the records, the inter-observer and intra-observer reliability and validity were calculated. We carried out a descriptive analysis of the behaviors’ frequencies coded through Lince. For data analysis we use the tools IBM SPSS 25 and THEME v.6. TPA were performed according to collective (team by team), team-gender and individual. Although analyzes were performed (Chi-square (χ2)), comparing the participation between girls and boys in each team; no significant differences were found (p>.05). Greater complexity was pointed out throught T-Pattern analysis in Monkeys and Ants teams than in the Bears team. Girls showed greater variability than boys in Monkey and Ants teams. When girls used 3 different roles, boys only used 2 of them, being the conduct to realease peers (p <0.005) recurrent in Monkeys. The boys as a team (p <0.005) used subroles that directly modified the outcome in the Ants team. Similar T-Patterns were found in girls (p<.05) when analyzed individually. The use of TPA allows the detection of hidden features while girls and boys were playing. The apparent neutrality of the game may have a seemingly random decision-making process but TPA revealed specificities highly applicable to the study of gender through triadic motor games.