Quantum State Betting Games

• Quantum state betting games are probabilistic decision tasks that leverage quantum states and measurements to structure bets and define risk-adjusted outcomes.
• They extend classic betting models by incorporating aspects like isoelastic utility and quantum ergotropy to maximize long-term capital growth.
• Operational criteria such as Arimoto α-mutual information and multivariate Rényi divergences quantify measurement informativeness and resource advantages in these games.

Quantum state betting games constitute a diverse and technically rich class of probabilistic decision tasks in which quantum states, measurements, or operations define the structure of bets, payouts, and optimization criteria. These games not only generalize classical betting and lottery paradigms to the quantum domain but also tightly interlink quantum information, utility theory, and the mathematical structure of quantum resource theories.

1. Operational Foundations and Mathematical Structure

Quantum state betting games typically involve a scenario where an agent attempts to assign bets, portfolios, or guesses based on quantum state preparation and measurement, with payoffs and risk preferences formally specified. A general template consists of:

• Preparation: An adversary (or nature) prepares one of several quantum states $\rho_x$ (or, in lotteries, several outcome distributions) according to a known prior $p(x)$.
• Measurement/Information Gathering: The agent selects a measurement (POVM) $\mathcal{M} = \{M_g\}$, obtaining outcome $g$, with $p(g|x) = \mathrm{Tr}[M_g \rho_x]$.
• Betting Decision: Based on $g$, the agent assigns conditional bets $b(x|g)$, subject to normalization, or partitioning of capital across possible outcomes.
• Odds and Payoff: Payouts are set by odds $o(x)$ (possibly reflecting bookmaker fairness). For multiple lotteries, $k$th bookmaker’s odds are $o^{(k)}_X(x)$.
• Utility: Nonlinear risk-modelling is incorporated via isoelastic utility $u_R(w)$ for constant relative risk aversion $R$ (with multi-lottery generalization $u_{\underline R}(\vec{w})$).
• Objective: The principal figure of merit is the isoelastic certainty equivalent (ICE), defined as the unique $w^\mathrm{ICE}_R$ solving $u_R(w^\mathrm{ICE}_R) = \mathbb{E}[u_R(W)]$, quantifying the risk-adjusted value of the betting outcome (Ducuara et al., 2021, Ducuara et al., 25 Jan 2026).

Formally, for single-lottery quantum state betting (QSB): $$w^\mathrm{ICE}_R(b,\mathcal{M}; o,\mathcal{E}) = \Big[ \sum_{x,g} (b(x|g) o(x))^{1-R} p(x) p(g|x) \Big]^{1/(1-R)}$$ for an ensemble $\mathcal{E} = \{p(x), \rho_x\}$. Bob’s strategy is to optimize the pair $(\mathcal{M}, b)$ to maximize $w^\mathrm{ICE}_R$ (Ducuara et al., 2021).

2. Quantum Kelly Betting and Continuous-Variable Models

Quantum generalizations of the Kelly betting criterion—maximal long-term growth of capital—replace classical wealth by quantum ergotropy (extractable work) of a single mode. In this framework (Tirone et al., 2020):

• The gambler’s capital is stored as the ergotropy $\mathcal{E}(\rho)$ of a bosonic mode in initial state $\rho_0$.
• Bets are placed by splitting mode $A$ into $J$ branches through a Bosonic Gaussian channel (BGC) with splitting ratios $\eta_j$ (satisfying $\sum_j \eta_j^2 = 1$).
• If outcome $j$ occurs, branch $j$ is amplified by gain $k_j$, all others discarded.
• The net process is an iterated sequence of lossy and amplifier BGCs acting on $\rho_0$, with capital growth determined by:

$$G(\vec{\eta}, \vec{k}, \vec{p}) = \sum_j p_j [\log_2 k_j + \log_2 \eta_j]$$

• The quantum Kelly criterion is achieved at $\eta_j = \sqrt{p_j}$, paralleling the classical result.
• For Gaussian input states (displacement $\alpha$, squeezing $r$, thermal noise $\bar n$), coherent states ($\bar n = r = 0$) maximize long-term growth for fixed ergotropy.

This approach enables quantum generalization of Kelly-type optimal gambling in systems where both the capital representation and noise/information acquisition are intrinsically quantum (Tirone et al., 2020).

3. Quantum Betting under Risk Aversion and Information-Theoretic Characterization

A major advance is the operational characterization of quantum betting games for arbitrary risk attitudes, connecting the agent’s maximal attainable utility to Rényi-type mutual information and divergences (Ducuara et al., 2021, Ducuara et al., 25 Jan 2026):

• For isoelastic utility with risk $R=1/\alpha$, the maximal ICE ratio between an informative measurement $\mathcal{M}$ and an uninformative measurement is given by the Arimoto $\alpha$–mutual information:

$$I_\alpha(X; G) = \frac{\alpha}{\alpha - 1} \log \left( \sum_g \left[ \sum_x p(x) p(g|x)^\alpha \right]^{1/\alpha} \right)$$

• The maximum risk-adjusted value is operationally:

$$I_\alpha(X;G) = \mathrm{sgn}(\alpha)\,\log\left( \frac{\max_b w^\mathrm{ICE}_{1/\alpha}(b,\mathcal{M}; o^c, \mathcal{E})} {\max_{N\in UI, b} w^\mathrm{ICE}_{1/\alpha}(b, N; o^c, \mathcal{E})} \right)$$

• In the limits $\alpha \to +\infty$ (risk-neutral), one recovers minimum-error quantum state discrimination; $\alpha \to -\infty$ yields quantum state exclusion.

Crucially, these results establish a four-way correspondence between (1) operational quantum betting tasks, (2) Arimoto-type information measures, (3) (quantum) Rényi divergences, and (4) resource monotones for measurement informativeness (Ducuara et al., 2021).

4. Multiple Lotteries and Multivariate Rényi Divergences

Extending betting games to multiple concurrent lotteries, the value assigned by a rational, risk-averse agent to a whole portfolio is determined by the multivariate Rényi divergence $D_{\underline\alpha}$ of the involved distributions (Ducuara et al., 25 Jan 2026):

• For $d$ lotteries with odds $o^{(k)}_X(x) \propto 1/p_X^{(k)}(x)$, and risk parameters $R_k$:

$$D_{\underline\alpha}(\vec{P}_X) = \frac{1}{\alpha_\star - 1} \log \left[ \sum_{x\in \mathcal X} \prod_{k=0}^d (p_X^{(k)}(x))^{\alpha_k} \right]$$

with $\sum \alpha_k = 1$, $\alpha_0 = (1+\sum_{k=1}^d (R_k - 1))^{-1}$.

• The maximal isoelastic certainty-equivalent is:

$$w^\mathrm{ICE}_{\text{opt}} = \exp( D_{\underline\alpha}(\vec{P}_X) )$$

• Conditional multivariate Rényi divergence quantifies the value increment from side information and satisfies a strong data processing inequality, reflecting the operational principle that side information can only increase an agent’s risk-adjusted expected value.

Quantum generalization is achieved via replacement of distributions $\vec{P}_X$ by quantum state ensembles $\vec{\rho}$, using Petz-type quantum Rényi divergences. This yields a decision-theoretic foundation for quantum resource monotones in measurement informativeness (Ducuara et al., 25 Jan 2026).

5. Quantum Lottery and Practical Implementations

Quantum lotteries, as realized in several protocols, are a class of quantum state betting games emphasizing unconditional fairness and unpredictability through quantum random number generation, entanglement, and cryptographic protocols (Mishra et al., 2022):

• Protocols employ single-photon (BB84), entanglement-based, and semi-quantum strategies, with ticket distribution, authentication, and cryptographically protected draw.
• Unconditional fairness is enforced mathematically via the uniform XOR of independent QRNG outputs; outcome entropy is maximal, and no coalition of $n-1$ participants or external adversaries can bias or predict the outcome.
• Comparative analyses show resource trade-offs and implementation complexities across protocols, with entanglement-based schemes offering device independence but at higher experimental cost.

Quantum lotteries demonstrate practical applications of quantum betting games in cryptographic and distributed settings, leveraging the foundational information-theoretic principles established above (Mishra et al., 2022).

6. Connections to Quantum Parrondo Effects and Game Theoretic Generalizations

Quantum state betting games interface with broader quantum game theory and dynamical phenomena, such as Parrondo’s paradox for quantum walks (Chandrashekar et al., 2010), "quantum Chinos games" (Centeno et al., 2021), and quantum board games with hedging phenomena (Ganz et al., 2017):

• Quantum Parrondo games exhibit the reversal of losing biases through alternation or composition of non-commuting "coin" operators, underlining the role of quantum interference and nonclassical correlations in betting/strategic contexts.
• Quantum Chinos and quantum board games formalize competitive multi-round protocols with omega-strategies and resource advantages conferred by entanglement or measurement informativeness.
• Hedging in quantum games highlights scenarios where strategies leveraging quantum resources can break classical product rules for winning probabilities in parallel repetition (but not in sequential play) (Ganz et al., 2017).

These games provide concrete arenas for exploring the operational impact of distinctively quantum phenomena on rational betting and optimization strategies.

7. Resource Theories and Informational Monotones

Resource-theoretic perspectives unify quantum state betting games with quantitative measures of how much a given resource (state, measurement, or channel) enhances the agent’s expected risk-adjusted payoff:

• For measurement informativeness, the family of resource monotones $M_\alpha(\mathcal{M})$ defined via Arimoto $\alpha$-capacities

$$M_\alpha(\mathcal{M}) = \mathrm{sgn}(\alpha) \, [2^{\mathrm{sgn}(\alpha) C_\alpha(\Lambda_{\mathcal{M}})} - \mathrm{sgn}(\alpha)]$$

interpolates between robustness at $\alpha\to\infty$ and weight at $\alpha\to-\infty$, is monotonic under classical post-processing, and fully characterizes the operational advantage in all risk regimes (Ducuara et al., 2021).

• In the multi-lottery scenario, similar arguments yield quantitative monotones for information unlocked by informative measurements or quantum resources (Ducuara et al., 25 Jan 2026).

Resource theory thus formalizes the conversion of abstract quantum properties into calculable advantage in rational quantum decision-making.

---

References:

• "Characterisation of quantum betting tasks in terms of Arimoto mutual information" (Ducuara et al., 2021)
• "Multivariate Rényi divergences characterise betting games with multiple lotteries" (Ducuara et al., 25 Jan 2026)
• "Kelly Betting with Quantum Payoff: a continuous variable approach" (Tirone et al., 2020)
• "Quantum and Semi-Quantum Lottery: Strategies and Advantages" (Mishra et al., 2022)
• "Quantum coin hedging, and a counter measure" (Ganz et al., 2017)
• "Parrondo's game using a discrete-time quantum walk" (Chandrashekar et al., 2010)
• "General quantum Chinos games" (Centeno et al., 2021)