Design

The town that cannot sell you a doctor

The starting town in a trail game does more work than any other location, because it is the only one the player visits with a full purse and no information. Get its shape right and the thirteen towns after it are a fill-in-the-blanks job. Get it wrong and you find out on town nine, with eight towns of committed work behind you.

So it got specified to the penny, the person and the line of dialogue. And in the process it exposed a contradiction between two rules I had been happily holding at the same time for about a year.

The two rules, and the collision

Rule A, the content rule. At the start, the player should be able to buy every amenity they could possibly want for the journey. Scarcity is a thing that happens later, out on the trail, where the shops are thin and specialised. The jumping-off town is where you get to be thorough.

Rule B, the spine. Medicine is the central scarcity of the game. Real medicine — the chest that sees off a first sickness, the fever powder, the laudanum — exists only where a physician stands, and physicians are rare on a trail. That scarcity is not decoration. It is the reason illness is frightening, and it accounts for the overwhelming majority of health-related deaths in simulation: parties that died having never once been within reach of a doctor.

Now put a surgeon in the starting town. Rule A is satisfied and rule B is dead. On a comfortable purse you buy a medicine chest at mile zero, and the game's entire health economy — the thing every illness card is calibrated against — collapses on the first screen.

Leave the surgeon out, and rule A is broken, and worse, the town's own description says so. The shipped blurb read, in effect, everything you will ever have is for sale here, which was flatly false against its own roster of shopkeepers. The one place the game lied to the player was in the sentence introducing it.

The audit that made it undeniable

The way this got settled was by counting. The game has six trades — provisions, cloth, livery, spirits, iron, medicine — and the starting town's roster has five people in it. Walk the four meters the player has to manage and check each is provisioned:

Every meter has a source. Two of the three tools in the game are on shelves. What is missing is not health help; it is the strong tier of it. And the sixth trade has no vendor at all, because the surgeon does not stand here — he first appears six hundred miles west.

Writing it out like that turned a vague unease into a precise question with exactly one decision in it: add a surgeon to the starting town, or don't?

The answer, and the reframe that made it work

Don't. And re-word the rule so it stops contradicting the game.

The reframe is one sentence, and it is the most useful thing I wrote all year:

The town sells you every supply and every piece of gear you can carry west. What it cannot sell you is a person. A doctor is not a sack of flour; a doctor is somebody who already has to be standing where you are. Out here, mostly, they aren't.

That makes the two rules agree instead of fight. Rule A becomes a claim about goods, which the town honours completely. Rule B becomes a claim about people, which the trail enforces. Nothing was nerfed, no number changed, and the design got simpler rather than more complicated — which is generally the sign that you have found the real distinction rather than a patch over it.

It is also, incidentally, truer to the period. A general store with a surgeon behind the counter is a video game convention. A frontier town where you can buy anything that fits in a wagon and cannot buy the one thing you will actually need is just an accurate description of the situation.

The gradient, not the hole

The important refinement is that the starting town is not silent on health — it is low-resolution about it.

The sutler sells a patent remedy, described in the game as brown and making claims. The laundress sells wool coats, which help in the way warm clothes actually help. Both do something. Neither is medicine. What the town lacks is the strong tier: the chest, the powder.

That distinction turns out to be the whole design. If a town with no doctor had nothing for a sick party, then arriving at a doctorless town while somebody was ill would be a death sentence handed out for a routing decision — precisely the kind of trap rule three forbids. The weak tier exists so that the difference between towns is always there is no doctor here and never there is nothing here.

That is a general principle worth pulling out of the specific case. Scarcity should be a gradient, not a hole. A hole in provision reads as a bug or a punishment; a gradient reads as geography. And a gradient gives the player something to do about it, which is the difference between tension and helplessness.

Vendors versus people you can hire

The other thing the specification forced into the open: a town's shopkeepers and the specialists you can talk into coming west with you are two different layers, and confusing them flattens the game.

The starting town's forge already belongs to a vendor — a blacksmith who keeps a full hardware wall and no interest whatsoever in your journey. There is also a hireable farrier in the game, a specialist you can bring along, and the obvious-looking move is to put him at the forge. It is the wrong move for three reasons, all of which generalise.

He would collide with the man already there, and two smiths in one town read as one shop priced twice. His kit is deliberately narrower than the vendor's hardware wall, so that the two forges never feel redundant — a distinction that only survives if they are in different places. And he is a premium hire gated behind depth: the whole progression is that better people become available the further west you get. His pricing only reads as a discovered bargain — a specialist who costs less than the tool he obviates — if you had to travel six hundred miles to find him.

So the template rule is: a town's shopkeepers are its vendors. The people you can talk into coming west are a separate, depth-gated layer. Vendors are geography. Hires are progression. Never put one in the other's slot.

What I took from it

The lesson is not really about frontier towns. It is that when a design feels stuck, the odds are good that two rules you believe are quietly incompatible, and the argument you are having is a proxy for that. The fix is almost never to weaken one of them. It is to find the distinction that makes them describe different things — goods versus people, in this case — at which point both survive intact and the design gets smaller.

The blurb got rewritten, too. A town is allowed to be proud of its shelves. It is not allowed to lie about them.