Roadmap / 0.1.3.1
Population and geography by province
v0.1.3.1, commit 1b04e2b. File docs/RESULTS_v0.1.3.1_geography.md.
Written at this version and not revised since. Later versions may change these numbers; see the roadmap.
Switch: spatial.enabled (now the geography version by default; spatial.uniform: true keeps the v0.1.3.0 skeleton for the aggregation check).
Options, each measurable on its own: population (real / scenario), dispersion, presence, weather (regional / national), local_shocks.
Runs: 3,000 paired runs per variant, every other module on. 26 tests pass; all harness checks pass, including the aggregation check.
Scope: provinces as aggregates. No sites, routes or installations.
Headline
Geography moves where the troops are, not who wins. With all five attributes on, the chance the US quits changes by +0.1 points in full annexation (−0.1 to +0.3) and 0.0 in the Alberta scenario (−0.2 to +0.2).
I said this would be the first version in which provinces diverge. They diverge in troops, a lot. They barely diverge in control (one point), and the reason is structural, explained below.
What was added
| Attribute | What it does | Data and parameters |
|---|---|---|
| Real population | The provinces' 2026 populations define the theatre. Alberta is 12.2% of Canada, not the 12% the scenario assumed. Canada's population in the actor file is updated from 41.6 to 41.8 million | Statistics Canada, 1 July 2026 |
| Dispersion | A province with a more rural population needs more troops per person: requirement × [1 + k × (rural share − national rural share)], k = 0 to 1 | Rural shares from the 2021 Census (17.8% nationally; 12.7% in BC to 54.0% in PEI). k is my estimate |
| Presence floor | Troops needed just to be present across the land, whatever the population: 100 to 800 per 100,000 km². These troops are not among the population, so they are not exposed to attack | Land areas from the 2021 Census. The rate is my estimate; there is no historical anchor for holding a nearly empty territory this size |
| Regional weather | Each province gets the weather of the regions its people live in, in place of the national average | The five weather regions from v0.1.2.6. The Atlantic provinces borrow the corridor's weather: there is no Atlantic region yet |
| Local shocks | 30 to 70% of the month-to-month variation in US casualties is local to a province | My estimate |
Whether the insurgency ends is still one national event, judged on national control and resistance, because that hazard is calibrated on national insurgencies.
Results: chance the US quits (change vs the single-area model, 95% interval)
| Variant | Full annexation | Alberta detachment |
|---|---|---|
| Single-area model | 70.3% | 59.2% |
| Population data update alone (41.6 → 41.8 million) | +0.1 (0.0 to +0.2) | 0.0 |
| Only real population | 0.0 | +0.1 (0.0 to +0.3) |
| Only dispersion | exactly 0 | −0.1 (−0.2 to 0.0) |
| Only presence floor | +0.1 (−0.1 to +0.3) | 0.0 (−0.1 to 0.0) |
| Only regional weather | 0.0 | 0.0 (−0.1 to +0.1) |
| Only local shocks | 0.0 (0.0 to +0.1) | 0.0 (−0.1 to +0.1) |
| All five | +0.1 (−0.1 to +0.3) | 0.0 (−0.2 to +0.2) |
| All five, presence floor at its high end (800) | +0.2 (0.0 to +0.5) | 0.0 |
Economic coercion: exactly 0 (no occupation, so no path).
What does move, in full annexation: - Control falls from 52% to 49%. The floor adds about 35,000 troops to what the US needs, and the US is already at its ceiling. - US deaths fall by 31 (1.9%), from 1,678 to 1,647, because about a tenth of the force is now spread over empty land and away from the population. - In the Alberta scenario the US needs 5% more troops (51,700 to 54,500) and has them.
Where the troops go (full annexation, average over the occupation)
| Province | Share of population | Rural share | Land (thousand km²) | Troops, uniform | Troops, with geography | Troops per 1,000 people |
|---|---|---|---|---|---|---|
| Ontario | 38.9% | 13.3% | 892 | 143,200 | 134,900 | 8.3 |
| Quebec | 21.7% | 19.0% | 1,299 | 79,800 | 79,000 | 8.7 |
| British Columbia | 13.7% | 12.7% | 921 | 50,300 | 48,500 | 8.5 |
| Alberta | 12.2% | 15.2% | 635 | 44,900 | 43,500 | 8.5 |
| Manitoba | 3.6% | 25.3% | 540 | 13,400 | 14,300 | 9.4 |
| Saskatchewan | 3.1% | 31.7% | 577 | 11,300 | 12,700 | 9.9 |
| Nova Scotia | 2.6% | 41.1% | 53 | 9,700 | 10,400 | 9.4 |
| New Brunswick | 2.1% | 49.1% | 71 | 7,800 | 8,700 | 9.8 |
| Newfoundland and Labrador | 1.3% | 40.0% | 358 | 4,900 | 5,900 | 10.7 |
| Prince Edward Island | 0.4% | 54.0% | 6 | 1,600 | 1,800 | 9.9 |
| The North | 0.3% | 41.6% | 3,437 | 1,200 | 8,600 | 61.6 |
The North goes from 1,200 troops to 8,600, seven times what its population implies, and Ontario gives up about 8,000. Under uniform settings every province had 8.8 troops per 1,000 people.
Why control barely differs between provinces
Troops are spread in proportion to need. So when the US is short, every province gets the same fraction of what it needs, and the shortfall is shared equally. Control is 48.5 to 48.7% everywhere except the North (47.6%, from colder weather). Resistance responds to control, so it stays equal too.
Provinces will only pull apart when one of two things differs: - How much they resist to begin with (political geography, v0.1.3.2). - Where the US chooses to put its troops (allocation, v0.1.3.4).
Those two are where the provincial split will matter, if it matters. This version shows that population, land and weather are not.
Also learned
- Dispersion cannot change the national total in full annexation. It is measured against the national rural share, so it only moves troops between provinces. That is deliberate: the national requirement is calibrated on whole countries.
- Local shocks do nothing visible. Spread of US deaths across runs falls from 1,555 to 1,553.
- Regional weather does nothing visible (+0.3 US deaths). The national average was already population-weighted.
Limits
- The presence rate and the dispersion coefficient have no empirical anchor. The presence floor uses total land area, which overstates the need where land is uninhabited; the range's low end covers that.
- Terrain (mountains, forest, cover) is not in. The research literature ties rough terrain to insurgency, so it belongs with political geography or straight after.
- No Atlantic weather region; interior British Columbia and northern Ontario and Quebec are not separated.
- The opening campaign is still national.
Sources
- Statistics Canada, population estimates for 1 July 2026 (https://www150.statcan.gc.ca/n1/daily-quotidien/260923/dq260923b-eng.htm).
- Statistics Canada, Census in Brief, "Population growth in Canada's rural areas, 2016 to 2021" (https://www12.statcan.gc.ca/census-recensement/2021/as-sa/98-200-x/2021002/98-200-x2021002-eng.cfm): rural population and rural share by province.
- Statistics Canada, 2021 Census, Table 98-10-0001-01: land area. The 13 figures used sum to the published national land area of 8,788,702.80 km²; Ontario, Quebec and Nunavut were checked individually, the other ten from my recall of the same table, validated only by that sum.