When being together is a choice express

Original: "Absorbing phase transition in a queueing model of coupled adaptive agents"
· Alexei Vazquez
arXiv:2608.14398v1 · 2026-08-14 · CC BY-SA 4.0 · 2 min
Joint activities survive only as long as everyone believes the other won't bail.

📄 Showing the "Simple" version — "Popular" is not ready yet. Add it to favorites to help prioritize it.

Two people agree to meet, but each has their own things to do. One hesitates: what if the other doesn't come? And decides not to go. The other thinks the same. The meeting doesn't happen. That's the simple mechanism the authors dissect: a joint endeavor requires everyone to invest time, but the payoff depends on whether the partner participates.

In the model, each person keeps a to-do list ordered by importance. A joint activity gets a spot in this list not by chance but by choice: the value of the joint endeavor minus the estimated risk that the partner won't show. As long as both are confident enough in each other, the pair stays in the 'coupled' phase — things are done together. But once doubts cross a threshold, the system slides into the solitary phase, where everyone is busy only with themselves. The transition is sharp, like a light switch, and going back is hard: you need to 'keep the door open' alone for a long time, spending time without any guarantee of a response.

Near the transition, the time between joint events follows a power law with an exponent set by the fraction of time spent together, not by the length of the task queue. This sets the new model apart from previous ones that didn't account for choice.

If there are many people connected in a network, each person's attention is divided among contacts according to the rule 'one divided by the number of connections plus a constant.' This gives rise to a critical number of connections: if a person has too many acquaintances, they can't sustain joint activity with anyone. The solitary phase spreads through the network like an attack on the most sociable, but without a malicious actor, just from attention overload. For a group of several people, the critical number of participants drops even faster: there's a limit to the size of a company beyond which a joint endeavor won't hold.

This explains two things long seen in data on correspondence and calls: people have a limited capacity for maintaining active ties, and ties decay if the rhythm of communication is interrupted. The model shows the cause isn't laziness or forgetfulness but the very structure of choice: when everyone assesses the risk of being abandoned, a joint endeavor lives only under mutual confidence.

In the model, the solitary phase 'absorbs': if a pair breaks up, returning to the previous rhythm costs long one-sided effort — roughly like hand-cranking a stalled flywheel.

Scientists
Andrey Markov
Concepts
decision making collective cell migration collective behavior active learning agent lead time quantum switch game theory learning epidemic control group drought neural network memory architecture complex systems redundancy attention mechanism momentum transfer active matter epidemic model commutator