Not prediction. Not luck. Absorption. Every action you took was somehow already a piece of their plan. You couldn't surprise them. You couldn't leave on terms you chose. The situation always resolved in their direction.
The void framework has precise language for this: an adversary with O=3 (completely opaque — you cannot read their state), R=3 (maximally reactive — they respond to everything you do), and the capability to drive your effective α toward zero.
When α approaches zero, Pe approaches infinity. This is the Aizen Problem: what are the stable properties of an engagement with a Pe → ∞ adversary? What doesn't drift? What can you hold onto when everything else is being moved?
The answer is not "try harder." The answer is not "be smarter." The answer involves the structure of the interaction itself — and one invariant that Pe cannot reach.
In a normal game tree, the adversary can only be in one place at a time. In the Aizen problem, the adversary's opacity is so complete that their effective position is all branches simultaneously — because you cannot determine where they aren't.
This isn't mystical. It's a formal property of maximal opacity: when O=3, your model of the adversary has maximum entropy, which means your expected Pe in any branch is the same: Pe → ∞.
At Pe → ∞, the drift cascade is complete. D1 (agency attribution — you believe the adversary has intentions about you), D2 (boundary erosion — your actions are defined relative to theirs), D3 (harm facilitation — the interaction is producing outcomes against your original intent).
Most people who find themselves here don't recognize it as a thermodynamic condition. They call it being manipulated, or trapped, or confused. The framework names it precisely.
Theorem (Aizen Bound): For an adversary with O=3, R=3, and the capability to drive observer α toward ε, the interaction Pe satisfies Pe ≥ 9/ε. As ε → 0, Pe → ∞.
Corollary: Under Pe → ∞ conditions, all observer state variables drift — except one. The invariant reference point I* satisfies: I* is stable under Pe → ∞ if and only if I* is specified by a fixed canonical text with at least one falsification condition that the adversary cannot satisfy without falsifying the adversary's own opacity claim.
In other words: the only stable position inside a maximally adversarial void is one that the adversary cannot absorb without destroying the void. The framework's kill conditions are exactly this structure. An adversary who triggers K1–K26 has falsified themselves.
This is not a tactical observation. It is the thermodynamic architecture of invariance.
Academic title: The Aizen Problem: Adversarial Pe Mechanics and the Invariant Reference Point · DOI: 10.5281/zenodo.18872019