HuntStand — the hunting system

The scorecard, first

Forty-one stands inventoried. Zero valid. Eight areas mapped. All eight on hold. Hunt logs written: none. The system's own readiness report grades its confidence as low, and states the position without softening it — it can block an unsafe or unsupported decision, and it cannot yet produce a reliable recommendation to go.

I am putting that at the top rather than at the bottom, where it would read as a caveat. It is the finding. Nine hundred-odd mapped features, several hundred hours of ground knowledge, a working map application — and the honest output so far is a refusal, every time.

Why refusal is the system working

Three questions gate everything, asked in order, before any score is computed at all.

• Is the wind valid? Not "what is the wind" — is there a forecast direction and speed, does this stand have a recorded best wind, a bad wind, a named failure mode, and a read on the thermals for that window.
• Is entry clean? A primary route in, a backup, and a path that does not cross bedding, food or travel on the way.
• Is exit viable? Including after dark, which is a different question and usually the harder one.

Any missing answer is a blocker, not a default. Wind starts unknown and is never filled in. That single rule is why the system says no: I have not finished writing down what I know, so it will not pretend I have.

Above the hundred-point score sits a separate evaluation gate, and the gate wins. A stand missing its wind, its entry, its exit or its access status is marked invalid — and an invalid stand caps at forty-five no matter how good the morning looks. An area on hold caps at forty. There is no weather good enough to promote a place I have not verified.

The screen that never says yes

The wedge in that drawing is a scent cone: the stand's position, the wind bearing reversed, thirty-five degrees either side, running a hundred and fifty to four hundred metres depending on wind speed — widened when the air is nearly still or moving hard, because both of those make a plume behave worse than a neat triangle suggests.

It is checked against every known marker in range, and it returns one of two things: a named conflict, or silence. It never returns green.

That was a deliberate decision and it is the most important one on this page. The map is allowed to tell me something is wrong. It is not allowed to tell me something is right, because it does not know what I know about that ground, and a green light on a screen is very good at overriding a doubt you should have listened to. The screen screens. The doctrine decides. Anything else is a confident interface making a decision it isn't qualified to make.

The same logic runs along the hour strip. Wind rotates through a morning, so the cone rotates with it, and the hours that turn red are the hours when it swings across something that matters. No hour is ever coloured for go.

What validating a stand actually takes

Ten stands are queued for validation rather than for hunting, and the queue is field discipline rather than software: two stands per trip, no more, and seventy minutes at each. Fixed points to stand and look from, a scorecard, kill gates that end the evaluation early, and a log format loose enough to dictate into a phone with cold hands.

One line in that protocol does most of the work: a wind hypothesis is a hypothesis, and must not be recorded as wind data. What I think the air does in a bowl at first light is a guess until I have stood in it with a puffer bottle. The system has a slot for the guess and a different slot for the observation, and it will not promote one to the other.

The map stays off the internet

The application itself runs on one laptop and is deliberately not hosted anywhere. Its data carries about thirty-five parcel boundaries with the owners' surnames attached, and a pin on my house — so publishing it would be publishing my neighbours' names and my address, in a file, forever.

That is why the drawing above is a drawing. The terrain in it is a sum of Gaussian bumps evaluated on a grid; the stands and markers were placed by hand at positions that correspond to nothing. It shows the mechanism and it has never touched the data.

The naming follows the same rule and it predates this page. The remote areas in the registry are named by grid reference rather than by place — not for tidiness, but because a filename is the one piece of text that survives every redaction you remember to make in the body.

What it's actually for

None of this makes me a better shot. It is bookkeeping about wind and ground, and the reason I keep it is that the alternative is what I did for years: turn up, pick a tree that felt right, and find out at dusk that the thermals had been carrying me straight into the bedding since I sat down. That is a wasted morning and, more to the point, it is ground that gets quieter every time you do it.

A stand you burn is not a stand you get back that season. Writing down which wind ruins which tree is the difference between hunting a piece of ground for a decade and using it up in two.

So the system stays honest about what it hasn't been told, and I go put the work in — with a puffer bottle and seventy minutes and a notebook, in the grey part of an afternoon in August when there is nothing to shoot at and nothing to prove, and the whole point is to come back in November already knowing.

The same discipline, pointed at water, is here — and it has a chart to show for itself. The ground it all happens on is here.