A live stochastic-SIR betting game built on a real extinction/endemic threshold theorem.
arXiv:2510.21371.
The paper extends the classic demographic SIR model by (1) replacing the fixed transmission coefficient
ω with ω + φ(z(t)), a bounded random perturbation built from an
Ornstein–Uhlenbeck process passed through φ(z) = (2d/π)·arctan(z) so the noise always stays inside
[−d, d]; and (2) letting the vaccination rate p(t) ramp over time instead
of staying constant. Their Theorem 5.1 proves the epidemic dies out for good whenever
ω < (ε+μ)(μ+p̄)/μ − d, and Theorem 5.2 proves it becomes permanently endemic whenever
ω ≥ (ε+μ)(μ+max(ε,p̄))/μ + d. Between those two lines lies a band of width 2d around
the critical point ϱ₀ = (ε+μ)(μ+p̄)/μ where no theorem resolves the outcome — that's exactly where
this game gets interesting. This demo runs the actual random ODE live in-browser (no canned animation) and scores your bet
against both the Monte Carlo result and the paper's "sharp bettor" (the theoretical threshold).
ω is the mean transmission rate. d is how wide the bounded noise band [ω−d, ω+d] is (the paper's actual noise construction, not a plain Gaussian).
How hard you vaccinate susceptibles over time: the cap p̄ and how fast it ramps up (k).
Click the line to jump ω there. Green = eradication proven. Red = endemic proven. Amber = neither theorem resolves it — only simulation can tell.
Pick a side, set your stake, then run the live Monte Carlo simulation.
| # | ω | d | p̄ | your bet | stake | empirical split | theory zone | actual | payout | bankroll |
|---|---|---|---|---|---|---|---|---|---|---|
| No rounds played yet. | ||||||||||