Competition cannot produce the goods of markets when one competitor is also part of the market’s operating system.
The good case for markets is distributed search: many actors trying things, learning locally, giving buyers exit, finding niches no planner could see.
That only works when the format preserves plural search.
Amazon versus independent bookstore is usually framed as a consumer matchup: choose the small shop over the giant.
But the seller may already be inside Amazon’s loop.
Listing there. Pricing against it. Sourcing through it when necessary. Losing discovery to its reviews. Absorbing its shipping norm. Treating it as competitor, infrastructure, fallback, and weather.
Competition with the arena owner does not select for better bookselling. It selects for coupling: who can survive the platform, feed it data, borrow its pipes, and appear legible inside its defaults.
Smogon.com understood this before political economy did, somehow.
Some bodies are not just strong. They overcentralize the format.
Let them into the same tier and the game stops discovering many viable strategies. It now discovers adaptation to them.
Markets need tiering language. This is becoming very clear.
A firm with platform power, marketplace power, fulfillment power, discovery power, seller data, price gravity, and fallback infrastructure is not an ordinary entrant in local retail.
The independent shop is not always outside Amazon. Sometimes it is fused into the same larger body; still alive, still worth preserving, but not separable by one buyer’s cathartic checkout choice.
Market Patch Notes: the goods of markets (distributed search, meaningful exit, plural experiment, local knowledge) require a playable format.
“Compete with the arena owner” is not a playable format.
Smogon can just program its tiers into the simulator. That's so easy.
Mega Rayquaza does not leak into RU overnight. It does not buy off the ladder. It does not finance ten shell teams. It does not copy every successful set, sell the held items, own the replay database, and train the player base to treat its Delta Stream as normal.
Companies can do all that and much more. It is so much harder to program the social human world.
A dominant firm does not enter a lower market only as a competitor. It enters through ownership, sourcing, fulfillment, advertising, pricing pressure, seller data, customer habit, debt, software, contracts, and emergency dependence.
So market tiering cannot mean “Amazon is asked to stay out of the local bookstore tier.”
That is not a rule. That is just a prayer.
A market tiering system would need something closer to a validator.
If the upper-tier actor is present in this chain as marketplace, supplier, fulfiller, data recipient, price weather, affiliate owner, or hidden backstop, the transaction tiers up.
The lower tier is not a good vibe. It is a protected format.
This is very rough. Not stone tablets.
A book market might have something like this:
Amazon belongs in the arena-owner tier.
The hard question is not whether Amazon should exist, or maybe no. The hard question is whether Amazon should be allowed to enter every lower market while carrying arena-owner powers.
A local bookstore can be beaten by another local bookstore in the social-human bookselling tier. That is a very playable format.
But a local bookstore competing with an actor that can also be marketplace, search habit, review layer, delivery norm, advertising channel, used-book exchange, ebook gate, audiobook gate, seller service provider, supplier, backstop, and price weather is just not the same contest.
That kind of competition cannot reliably produce the real goods people praise markets for producing. It does not preserve plural search. It narrows search around the dominant body.
Maybe, but only if the target here is smaller than “fix the whole economy.”
There is no single simulator for the human world. No one can push a clean rule update and prevent Amazon from touching every bookstore-relevant path. This is why market tiering has to begin where rules are actually executable.
These are not the whole market. They are playable subformats. They are places where a validator can exist.
The reachable transition is not “ban Amazon from books.” That slogan would become stupid almost immediately. The reachable transition is:
When an institution creates a market for social-human bookselling goods, it can define which powers are legal inputs.
If those Amazon-tier powers enter, the transaction is not now morally damned. It just tiers up.
A bookstore sourcing a rare emergency copy through an Amazon-owned route may be doing what survival requires. Modal Path Ethics should not turn that into some kind of purity failure.
But the transaction should still not be counted as clean lower-tier market activity either.
The point is not to find the moral stain. The point is field accounting.
Smogon has a teambuilder.
Markets have paperwork, contracts, audits, payment rails, procurement officers, ownership records, invoices, inventory systems, shipping labels, software logs, grant conditions, and lawyers.
Messier. More expensive. Easier to game.
Still not nothing.
A protected lower-tier book transaction would need to validate at least six things.
That is the start of a market validator. The validator will miss things. It will be gamed. It will sometimes punish the wrong actor. It will create paperwork. It will tempt consultants to invent fake badges and extract fees from the stores it was supposed to protect.
This is why the tiering idea cannot be sold as a new clean moral technology. It is an accounting technology for a dirty boundary.
A real version would need appeal paths, hardship exceptions, phase-ins, public templates, low-cost compliance, and safe emergency sourcing rules. Otherwise it just becomes another instrument that large actors can navigate and small actors drown under.
The lower tier cannot be protected by making the lower tier carry all the administrative weight. That would reproduce the same bug.
Nothing, in the cosmic sense. That has to be admitted.
Amazon can still shape customer expectations. It can still sell books elsewhere. It can still advertise. It can still own adjacent routes. It can still be the weather.
The question is narrower:
Can a given institution create a market format where Amazon’s higher-tier powers are illegal inputs?
Yes, sometimes.
This does not remove the arena-owner from our society. It just creates rooms where the arena-owner cannot legally bring the whole stack to play.
That is the closest thing the social human world has to preventing Mega Rayquaza from leaking into RU. You do not control every battle everywhere. You create validated formats, then refuse to call the format healthy when the forbidden body is present.
Every tiering system dies if it only checks the name on the storefront.
The rule has to follow control. Subsidiaries. Affiliates. White-label fulfillment. Exclusive backend dependency. Data-sharing arrangements. Marketplace-of-record status. Seller-of-record games. Financing arrangements. Acquisition options.
The validator has to distinguish fraud from dependency.
This field is already fused. Modal Path Ethics does not get to pretend surgical separation is easy here.
Some of these shops are not outside the larger body in any real sense. They are living organs inside it; still carrying local memory, still performing real social work, still worth preserving, but no longer cleanly separable by the consumer’s good choice.
That is exactly why “just buy local” is too small a slogan. This asks the buyer to somehow perform separation at checkout after the field has already been coupled upstream.
The first patch is to language. Let's see if this works:
The second patch is accounting. Track coupling. Track ownership. Track fulfillment. Track data. Track supply. Track dependency.
The third patch is protected demand.
Create specific institutional markets where lower-tier goods are the explicit target: local memory, human curation, community presence, browsability, small-press discovery, author access, public literacy, and social trust.
The fourth patch is backend substitution.
If lower-tier sellers depend on upper-tier infrastructure because no other pipe exists, then the protected format needs cooperative backend capacity or it will become theater.
This is the part where the ticket gets expensive and boring. Sorry.
Without that part, tiering only names the wound. It does not alter reachability.
Unresolved.
Promising.
Leaky.
Politically difficult.
Administratively dangerous.
Still much closer to the real structure than consumer sainthood.
Markets are field instruments. A market does not become good because buyers feel morally clean inside it. It becomes useful when the field-description built into the contest lets useful things win its game.
A competition format where the arena-owner can enter the lower tier as seller, supplier, marketplace, data channel, delivery norm, price anchor, and emergency backstop will not reliably produce the goods of markets. It will produce adaptation to the arena-owner.
Social Human Market Patch Notes:
Tiering is not a slogan for punishing big firms.
It is a way to ask whether the contest is still playable.
Mega Rayquaza can exist.
It just cannot be allowed to write the rules of RU.