The scaling and Brownian approximation conjecture for gambler's ruin probabilities

About 6 years old · traced to

For positive integers A,B,CA,B,C, let PA,B,C(σ)P_{A,B,C}(\sigma) denote the first hitting probability of the permutation σ∈S3\sigma\in S_3 in the gambler's ruin walk, and let PA,B,CBM(σ)P_{A,B,C}^{\rm BM}(\sigma) denote the corresponding Brownian motion extinction probability. For n≥2n\ge2, write nA,nB,nCnA,nB,nC for the scaled initial state. Scaling and Brownian approximation conjecture. (a) For each A,B,C≥1A,B,C\ge1, σ∈S3\sigma\in S_3, and n≥2n\ge2,

∣PnA,nB,nC(σ)−PA,B,C(σ)∣<0.0004.\left|P_{nA,nB,nC}(\sigma)-P_{A,B,C}(\sigma)\right|<0.0004.

(b) For each A,B,C≥1A,B,C\ge1, with N:=A+B+CN:=A+B+C, and each σ∈S3\sigma\in S_3,

∣PnA,nB,nC(σ)−PA,B,CBM(σ)∣=O(1(nN)4)as n→∞.\left|P_{nA,nB,nC}(\sigma)-P_{A,B,C}^{\rm BM}(\sigma)\right|=O\left(\frac{1}{(nN)^4}\right)\quad\text{as }n\to\infty.

These assertions are intended to explain the rapid convergence of the rescaled probabilities and provide a more precise version of the scaling conjecture. The stated bounds are conjectural, and the supplied context gives no resolution status.

References

Primary source

Persi Diaconis and Stewart N. Ethier, “Gambler's Ruin and the ICM”, arXiv:2011.07610 (2021).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.