Healthy Algorithm Game

Can a regulator’s reward make competing apps choose healthier recommendations?

C = healthy; D = addictive (assumed labels). Two apps choose simultaneously, know all payoffs, and maximize their own payoff. Each C earns reward r.

Try r = 0, 1, and 2. Choose both apps’ actions. Can either app gain by switching alone?

Reward r:

App A: App B:

Hypothetical payoffs (A, B); reward included
A / B	C	D
C
D		2, 2

Scope: Q1 incentive simulation only. Q2 benefit/regret ratings and Q3 informed user choices remain planned; no human data or demonstrated health effects.

SDG 4: students learn incentives through play. Planned evaluation: compare reward-threshold answers before and after play; learning gains are untested.