🔢 mathematics

The Dynamical Landscape of Beggar-My-Neighbour: Ultra-long Matches, Loops, and Infinite Matches

This paper provides a rigorous mathematical and computational analysis of the card game Beggar-My-Neighbour, characterizing its statistical dynamics, identifying non-terminating cycles through an automated "Infinite Loop Factory," and confirming the existence of infinite matches in both standard and generalized settings.

Original authors: Nicolas Andorno, Giulio Cernoia, Simone Duiz, Alessandro Michelangeli

Published 2026-03-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Nicolas Andorno, Giulio Cernoia, Simone Duiz, Alessandro Michelangeli

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a card game called Beggar-My-Neighbour. It sounds like a game of chance, but it's actually a game of pure, unbreakable logic. Once you shuffle the deck and deal the cards, the entire future of the game is written in stone. There are no choices, no strategies, and no "what ifs." The cards just fall where they must, one by one, until one player wins everything or... the game goes on forever.

This paper is like a deep-dive detective story into the hidden universe of this simple game. The authors, a team of mathematicians and computer scientists, asked three big questions:

How long do games usually last?
Why do some games last so long they seem to go on forever?
Do games actually exist that never end?

Here is the breakdown of their findings, translated into everyday language.

1. The "Rolling Dice" Illusion

You might think that because the game is 100% deterministic (no randomness once the deal is made), predicting the length of a game should be easy. But the authors found something weird.

If you play millions of games, the length of the matches follows a pattern that looks exactly like rolling dice. Most games end quickly. Some last a bit longer. A very few last a very long time.

It's like a "memory-less" system. Imagine you are walking through a foggy forest. Every step you take, you have a tiny, fixed chance of stumbling out of the fog and finishing your walk. It doesn't matter if you've been walking for 10 minutes or 10 hours; your chance of finishing right now is the same. The game behaves this way. Even though the cards are moving in a strict, predictable pattern, the duration of the game feels random.

2. The "Ultra-Long" Marathons

The authors simulated hundreds of millions of games. They found "marathon" matches that lasted hundreds of tricks.

Think of these long games as a tug-of-war between two players.

The Oscillation: For most of the game, the players are evenly matched. One player gains a few cards, then the other gains a few. It's like a pendulum swinging back and forth.
The Near-Death Experience: Sometimes, one player gets down to just 2 or 3 cards. It looks like they are about to lose. But then, a lucky (or unlucky) card appears, they win a huge pile, and suddenly they are back in the game.
The "Stock Market" of Cards: The number of cards each player holds fluctuates wildly, looking like a chaotic stock market chart, even though the rules are rigid.

The authors discovered that these long games have a hidden rhythm. They aren't just random noise; they have "macro-cycles" (big swings) and "micro-cycles" (small jitters) that repeat in a complex dance.

3. The "Infinite Loop" Factory

The biggest mystery of Beggar-My-Neighbour has been: Do infinite games exist?

For a long time, people guessed "no" because the deck is finite. If you run out of new card combinations, you must eventually repeat a state. If you repeat a state, you are in a loop. But finding a loop is like finding a specific grain of sand on a beach.

The authors built a "Loop Factory."
Imagine you are a sculptor trying to build a bridge that leads to a specific spot. Instead of guessing, they worked backwards.

They started with a small, simple loop (a tiny circle of cards that just keeps spinning).
Then, they asked: "What cards could have been added before this loop to create it?"
They kept adding cards backwards, like rewinding a movie, until they built a full deck of 52 cards that would naturally flow into that loop.

The Result: They proved that infinite games do exist. They found specific starting decks where the players will play forever, passing the cards back and forth in a perfect, unending cycle.

4. The "Unbalanced" vs. "Balanced" Secret

Here is a fascinating twist.

The Old Way: Previous researchers found infinite loops, but they were "unbalanced." One player was a giant with a massive deck, and the other was a tiny deck that just barely survived. It was like a bully and a victim in a dance that never ended.
The New Way: The authors' "Loop Factory" found loops where the players are perfectly balanced. They start with 26 cards each, and they stay perfectly balanced forever. It's a true, endless stalemate.

5. The "Backwards" Problem

The paper also explains why it's so hard to predict the game.

Forward: If you know the current state, you know exactly what happens next. (Easy).
Backward: If you see the current state, you often cannot know what happened before. Many different past situations can lead to the exact same present.

It's like seeing a finished cake. You know exactly how it tastes, but you can't be 100% sure if the baker used 2 eggs or 3, or if they added the sugar before the flour. The game "forgets" its history. This lack of "backwards determinism" is why finding those infinite loops was so hard—you couldn't just reverse-engineer them easily; you had to build them piece by piece.

Summary

The authors turned a simple children's card game into a complex mathematical landscape. They showed that:

Most games are short and follow a predictable "exponential" pattern (like a dying battery).
Long games are chaotic marathons with hidden rhythms.
Infinite games are real, and they can be constructed by working backwards from a simple repeating loop.

They didn't just solve a card game; they mapped out the "dynamical landscape" of a system where order and chaos dance together, proving that even in a game with no choices, the universe of possibilities is vast, strange, and occasionally, endless.

1. Problem Statement

The paper investigates the deterministic card game Beggar-My-Neighbour (also known as Cavacamisa or Pataiola), played with a standard 52-card deck (or generalized $(N, R)$ settings where $N$ is total cards and $R$ is the number of special card ranks).

Core Question: Does the game always terminate, or can it enter an infinite loop?
Context: While the game is fully deterministic (no player choice), its state space is vast ( $C(52,4) \approx 10^{21}$ ). Previous literature and amateur attempts have struggled to definitively prove the existence of infinite games in the standard $(52, 4)$ setting, with conflicting claims regarding loop frequencies.
Theoretical Gap: The paper addresses the lack of a rigorous mathematical framework for the game's state space, the non-injectivity of the transition function (preventing unique backward reconstruction), and the statistical nature of match durations.

2. Methodology

The authors employ a hybrid approach combining rigorous mathematical formalism, large-scale numerical simulation, and algorithmic search strategies.

Formal Framework (Sections 5–7):
- Defined the state space $S$ as ordered sequences of cards split between two players.
- Formalized the trick rule function $F: S \times \{A, B\} \to S \times \{A, B\}$ .
- Proved that $F$ is not injective even when restricted to reachable states. This means multiple distinct previous states can lead to the same current state, making "backward determinism" (reconstructing a unique history) impossible.
- Established that if any infinite match exists, the expected game duration $E(T)$ is formally infinite.
Numerical Simulation (Sections 4 & 8):
- Conducted brute-force simulations of $10^7$ to $10^8$ random initial decks for various $(N, R)$ settings.
- Analyzed the distribution of match lengths, ultra-long matches (hundreds of tricks), and statistical indicators like card separation and position entropy.
Algorithmic Construction (Sections 10–12):
- Developed an automated "Infinite Loop Factory" algorithm.
- Unlike previous heuristic approaches, this algorithm uses adaptive backtracking and diverse insertion strategies (e.g., "Low-rank-first," "Cluster specials," "Rank-1 blocks") to construct non-terminating decks.
- It searches for unbalanced loops and then performs backward reconstruction to find balanced initial configurations ( $|Deck_A| = |Deck_B| = N/2$ ) that evolve into these loops.

3. Key Contributions

A. Resolution of the Infinite Game Existence

The paper provides a definitive proof of existence for infinite games in the standard $(52, 4)$ setting and generalized settings.

The authors explicitly construct initial balanced decks that enter a loop (e.g., a 62-trick loop and a 72-trick loop).
They refute previous literature claims that infinite games in the standard setting were found with high frequency ( $1/150,000$ ), clarifying that earlier reports likely referred to modified, smaller deck sizes.

B. Statistical Characterization of Finite Matches

Distribution: The duration of finite matches follows an approximate exponential (geometric) distribution over a wide intermediate range, rather than a Poisson distribution.
Memory-less Property: The game exhibits an emergent "memory-less" dynamic: the probability of a match ending in the next trick is roughly constant regardless of how long it has already lasted. This explains the "stock-market-like" erratic oscillations in card counts observed in ultra-long matches.
Ultra-Long Matches: Matches lasting hundreds of tricks are rare but consistent with the exponential tail of the distribution. They exhibit multi-scale oscillatory patterns (macro, meso, and micro-cycles) in card counts and special card separation.

C. The "Infinite Loop Factory"

The authors introduce a novel algorithmic method to generate infinite loops.

Mechanism: The algorithm starts with small, known cycling seeds (e.g., minimal decks with rank-1 terminators) and systematically inserts cards while maintaining the non-termination property.
Innovation: It successfully identifies balanced initial decks that lead directly into loops, correcting the limitation of previous works (like [2]) which found loops only in unbalanced configurations or required complex, non-reproducible heuristics.

D. Entropy and Separation Analysis

Position Entropy: The paper analyzes the "position entropy" of special cards. In ultra-long matches, entropy fluctuates around a moderate value, indicating that the game dynamics prevent both extreme clustering and perfect uniform dispersion of special cards.
Separation: The average separation between special cards oscillates pseudo-periodically, serving as a marker for the "tactical position" of a player.

4. Key Results

Existence of Loops: Infinite matches exist for standard $(52, 4)$ and various generalized $(N, R)$ settings. The authors provide explicit initial configurations (Table 5) that result in immediate or transient loops.
Non-Injectivity: The trick function $F$ is not injective on the set of reachable states. This implies that the game's history cannot be uniquely reconstructed from a current state, and the state space contains "merging" trajectories.
Statistical Regime:
- Finite matches have a mean duration that decreases as the ratio of special cards ( $4R/N$ ) increases.
- The variance-to-mean ratio indicates over-dispersion (clustering) for low special-card ratios and under-dispersion for high ratios.
Loop Characteristics:
- Some loops exhibit symmetry (players swap dominance periodically, e.g., the $(40, 1)$ loop).
- Others exhibit persistent asymmetry (one player acts as a "reservoir" with a larger deck, while the other oscillates near depletion, e.g., the $(52, 4)$ loop).
Existence Window: Loops appear to exist in an intermediate regime of the special-card ratio ( $4R/N$ ), roughly between 7% and 43%. Extremes (0% or 100% special cards) yield trivial one-trick games.

5. Significance and Open Problems

Theoretical Impact: The paper resolves a long-standing open problem in combinatorial game theory regarding the termination of Beggar-My-Neighbour. It demonstrates that a simple, choice-free deterministic system can possess a complex dynamical landscape including infinite attractors.
Methodological Advance: The "Loop Factory" provides a reproducible, algorithmic framework for finding non-terminating states in deterministic systems, moving beyond ad-hoc heuristic searches.
Broader Implications: The findings bridge combinatorial game theory, dynamical systems, and statistical mechanics. The emergent memory-less behavior in a deterministic system offers a case study for how stochastic-like properties can arise from deterministic rules.

Open Problems Identified:

Taxonomy: Classifying initial decks into equivalence classes based on their asymptotic behavior (terminating vs. specific loops).
Metrics: Defining distance functions on the state space to predict match duration or proximity to a loop.
Backward Determinism: Rigorously characterizing exactly when and why injectivity fails (merging trajectories vs. loop entry).
Landscape Completion: Constructing explicit examples for hypothetical trajectory types, such as merging terminating paths or unique transient paths entering loops.

In conclusion, the paper establishes that Beggar-My-Neighbour is not merely a finite game but a rich dynamical system capable of infinite cycles, with a statistical structure that mimics stochastic processes despite being entirely deterministic.