The neuroscience of play is remarkably strong—precise mechanisms show how play shapes brain architecture through BDNF upregulation and synaptic pruning. Yet behavioral evidence reveals modest effect sizes (g ≈ 0.3-0.4) and limited transfer across domains. This episode examines the paradox: strong neurobiological foundations meet surprisingly modest measurable gains. We explore guided play as the optimal approach (outperforming both free play and direct instruction), the deprivation-enrichment asymmetry, the minimum effective dose (35+ minutes), and why at-risk children benefit most. The evidence favors "equifinality"—play as one of multiple valid developmental routes—challenging claims that play is uniquely necessary. Full research report: https://research.yuda.me/podcast/episodes/kindergarten-first-principles/ep2-play-pedagogy/report.md