← ~/visualizations

zero-sum-games

Interactive zero-sum game view: a 2×2 payoff matrix A (Row’s payoff; Column’s payoff is −A), draggable mixed strategies p and q, and the resulting expected payoff p^T A q. The bottom band visualizes minimax security levels (min_q p^T A q and max_p p^T A q) and how they align at the game value v near equilibrium.

canvasclick to interact
t=0s

practical uses

  • 01.Designing robust strategies in adversarial settings (security, pricing, auctions)
  • 02.Understanding Nash equilibrium in competitive games via mixed strategies
  • 03.Evaluating guarantees: worst-case performance under an optimal opponent

technical notes

Pure Canvas2D, green-on-black blocky UI. Pointer-driven bars adjust p(R1) and q(C1) with a damped spring for smoothness; when not dragging, strategies auto-oscillate to demonstrate changing expected value. Uses a fixed 2×2 zero-sum matrix so minimax value is clearly shown (v=0) and responsive scaling via scale=min(w,h)/baseSize with grid snapping.

← relationskkt-conditions →