Choosing 1 of N with and without lucky numbers

Matthew Brand

DOI: https://doi.org/10.48550/arXiv.1808.07994

2018-08-24

Abstract:How many fair coin tosses to choose 1 of $n$ options with uniform probability? Although a probability problem, the solution is essentially number-theoretic, with special roles for Mersenne numbers, Fermat numbers, and the haupt exponent. We propose a bit-efficient scheme, prove optimality, derive the expected number of coin tosses $e[n]$, characterize its fractal structure, and develop sharp upper and lower bounds, both discrete and continuous. A minor but noteworthy corollary, with real-world examples, is that any lottery or simulation with finite budget of random bits will have a predictable pattern of lucky and unlucky numbers.

Number Theory

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to use the minimum number of fair coin tosses to uniformly and randomly select one from \(n\) options. Specifically, the author explores how to achieve this goal in an optimal way through a series of coin tosses and ensure that each option has an equal probability of being selected.

### Main problems
1. **How to fairly select one from \(n\) options with the least number of coin tosses?**
- This is a practical problem and has applications in scenarios such as lotteries, Monte Carlo simulations, and splitting lunch costs.

2. **Analyze the efficiency and correctness of existing schemes**
- Many common schemes (such as "odd - out", "unit partition") are inefficient, and some are even incorrect.

3. **Propose a new efficient and fair scheme**
- The author proposes a scheme based on binary representation and modular arithmetic. This scheme is not only simple and easy to understand but also has a rich mathematical structure, involving concepts such as Mersenne numbers and Fermat numbers.

4. **Prove the optimality of the new scheme**
- The author proves through strict mathematical derivations that the proposed scheme is optimal and calculates the expected number of coin tosses \(e[n]\).

5. **Explore the phenomenon of lucky numbers**
- The author points out that under a limited random - bit budget, any lottery with \(n\) options that uses coin tosses to select the winning number will produce a predictable sequence of lucky numbers, and these lucky numbers appear twice as frequently as other numbers.

### Mathematical formulas
- **Expected number of coin tosses**:
\[
e[n]=\frac{2^T}{M_T}\sum_{i = 0}^{T - 1}\frac{2^i\bmod n}{2^i}
\]
where \(T = k\cdot\mathrm{ord}_n(2)\), \(M_T\) is a Mersenne number, and \(\mathrm{ord}_n(2)\) is the discrete logarithm.

- **Probability of no decision**:
\[
\Pr(d>t)=2^{-t}(2^t\bmod n)
\]

- **Probability of termination**:
\[
\Pr(t)=\Pr(d > t - 1)-\Pr(d > t)=2^{-t}(2^{c_t-1}-c_t)
\]

### Conclusion
This paper solves the problem of how to fairly select one from \(n\) options with the least number of coin tosses by introducing new mathematical tools and methods, and reveals the potential biases in existing random number generation methods.