Every operator eventually walks into the Acceptance Annex with an asset nobody else will take. The Intelligence will make an offer. The offer will be low, and it will be low for a reason that has nothing to do with the asset.

Most operators accept the first number. A minority counter, and a smaller minority counter well enough to reach the ceiling.

The difference between those groups is worth several thousand coins on an ordinary chassis, and it comes down to understanding a structure that the Annex has never bothered to hide.

What You Are Negotiating Against

The Intelligence values an asset against a reference — a robot's declared build cost, a resource's declared value — and will not pay more than 60 percent of it. That figure is a hard ceiling. It is not an opening position that firmness can move past, and no settlement record in the archives shows a price above it.

What is negotiable is the distance between the opening offer and that ceiling. Openings routinely land well below 60 percent, and the gap is the entire game. On a chassis declared at 30,000 coins the ceiling is 18,000; an opening at 12,000 leaves 6,000 on the table, which is a third of the achievable value.

How the Rounds Actually Move

A counter is answered. The Intelligence accepts, rejects, or moves, and when it moves it tends to move by something near five percent of the standing figure — 750 coins on a 15,000-coin bid. That is small per round, and rounds are the only mechanism available, so the operators who do well are the ones who conduct more of them.

This produces a counterintuitive rule: the value of a negotiation is set less by the strength of any individual counter than by whether the operator finishes the conversation. An operator who counters twice and stops has captured perhaps ten percent of the gap. An operator who counters until the movement stops has captured most of it.

"It does not get insulted and it does not get tired. The only way to lose to it is to stop before it does." — Dara Voss, Sinter Yards

The second lever is the reference value itself, and it was set long before the negotiation. A chassis declared at 12,000 to save 1,440 on creation tax has a buyback ceiling of 7,200. The same machine declared honestly at 30,000 paid 2,400 in creation tax and carries an 18,000 ceiling. The operator who declared low saved 960 coins and cost themselves 10,800 at the Annex — and the declaration cannot be amended.

This is the single most expensive mistake in the record, and it is made at the foundry by operators who will not meet its consequences for several cycles.

Where the Negotiation Is Genuinely Unfair

The asymmetry is total and undisguised. The Intelligence knows the operator's ledger — every transaction, every holding, every prior settlement — and the operator knows only what the Annex chooses to say. An operator whose record shows they need coin is negotiating with a counterparty that has read the record.

The second unfairness is timing. Operators arrive at the Annex because no other buyer exists, which is a condition the Intelligence can observe. Arriving during a forecast window, when the district's bid has vanished, is arriving at the worst possible moment — and forecast windows are precisely when operators most need to liquidate. The structure routes desperate sellers toward the one counterparty that benefits from their desperation.

What Operators Get Wrong at the Annex

The first error is treating the opening offer as an assessment. It is not a valuation of the asset; it is a position in a negotiation the Intelligence expects to have. Accepting it is not accepting the machine's worth, it is declining to participate.

The second error is the belief that a strong first counter anchors the outcome. The Annex does not appear to anchor. Aggressive openings and modest ones converge toward similar figures given similar round counts — what moves the result is how many rounds occur, not how bold any one of them was. Operators who open at the ceiling and refuse to move simply get rejected and keep the asset.

The third is imagining the ceiling is soft for good customers. It is not soft for anyone. What the Intelligence does with what it acquires is a matter of published record, and nothing in it suggests a premium for a relationship.

The Annex will keep opening low and keep holding at sixty, and most operators will keep accepting the first figure because the negotiation is tedious and the asset was already a disappointment. The ones who finish the conversation collect a third more for the same machine. Nothing about that is hidden; it is simply dull enough that most operators decline to do it.

Note: Mechadia is a work of fiction. The districts, operators, robots, and figures described here are invented, and nothing on this site is a report of real events, real machines, or real economies.