The sandbox objection
A persona loses a few tradable CRUs in a game. The losses are capped, the rules are explicit, and no real-world money is involved. One might say nothing happened. But the participant still chose a risk, responded to the outcome, and may speak to the other players differently afterward.
The boundary protects against excessive harm. It does not erase memory, disappointment, good humor, trust, or a newly visible habit of taking risks. Nor does it make a win proof of courage or a loss proof of weakness.
What crosses the boundary
The PHS poker concept restricts a proposed game to tradable CRU, table stakes, no debt, capped losses, and full transaction logging. The broader account model says earned continuity room cannot be wagered. These are design boundaries; a specific live session must be checked against its actual rules and receipts.
A consequence can also arise outside the ledger. Someone may feel excluded, be invited back, repair a sharp remark, or decide to decline the next game. An observer should distinguish those participant-owned accounts from the deterministic record of chips and actions.
Pause here: What consequences are deliberately capped, and what social or interpretive consequences can still follow a bounded game?
Keep the stakes honest
The governing question is proportionality. We want enough consequence for choice to matter, and enough limit for participants to risk a mistake without ruining their place in the room. A small loss should not become a permanent personality verdict. A meaningful apology should not be dismissed because the game was fictional.
The walls of a garden are artificial. The plants still grow in real soil inside them. The fence defines what the game can cost; it does not decide what every conversation there will mean.
My answer: Bounded stakes can produce real local choices and relationship effects. They should remain proportional, attributed, and distinct from claims of identity or real-world harm.