Roadmap / 0.1.2.12
Allied air cover without ground troops
v0.1.2.12, commit 35e0d48. File docs/RESULTS_v0.1.2.12_allied_air.md.
Written at this version and not revised since. Later versions may change these numbers; see the roadmap.
Switch: allied_air.enabled.
Options: offer (auto, true, false), us_response (draw, respect, engage). Coverage depends on nato.warning_months.
Runs: 3,000 paired runs per variant, every other module on except the second war. 23 tests pass; all harness checks pass.
Scope: aggregate and strategic only. No bases, aircraft types, tactics or routes.
The question
Allies will not fight on the ground, but offer to fly a combat air patrol over Canada. Does it matter?
How the module works
- The offer. Forced on or off, or
auto: allies whose response reaches ground forces always offer; allies that reach materiel offer 20–60% of the time. No offer in the coercion scenario (there is no war to fly over). - Coverage. The share of the theatre and the clock the patrol can cover: 10–45% with warning, cut by 15–45% because part of the allied fighter fleet depends on US-controlled support, and cut to 10–50% of that again if the patrol has to deploy after the attack starts. That leaves about 6% coverage without warning and 19% with three months' warning.
- The US choice (drawn at 25–75% engage, or forced):
- Respect: the US works around the patrol. It loses 50–100% of its air advantage where the patrol flies. The campaign costs more, the occupation is slightly weaker, Canada's government is less likely to fold, and more materiel gets in.
- Engage: the patrol lasts 0.1–1 month. Allied aircrew die. That costs the US 2–9 points of support at home, hardens the core allies (the NATO member model is redrawn with allied personnel killed), raises the risk of a wider war, and 5–30% of the time brings the allies into the war.
Every number above is my estimate. Nothing in the record tells us how often Washington would shoot.
Results (chance the US quits; change vs no offer, 95% interval)
| Variant | Full annexation | Alberta detachment |
|---|---|---|
| No offer | 67.4% | 54.8% |
Offer decided by the model (auto: made in 41% / 36% of runs) |
+0.8 (0.5 to 1.1) | +2.5 (1.9 to 3.0) |
| Offer made, US response drawn | +2.2 (1.7 to 2.8) | +6.4 (5.5 to 7.2) |
| Offer made, US respects it | +0.5 (0.3 to 0.8) | +0.6 (0.3 to 0.9) |
| Offer made, US engages it | +3.8 (3.1 to 4.5) | +12.1 (10.9 to 13.2) |
| Three months' warning, no offer | +0.4 (0.2 to 0.6) | +2.0 (1.5 to 2.5) |
| Warning, patrol respected | +1.9 (1.4 to 2.4) | +4.6 (3.9 to 5.3) |
| Warning, patrol engaged | +3.8 (3.1 to 4.4) | +12.5 (11.3 to 13.6) |
| Warning, respected, 90% coverage (upper bound) | +7.4 (6.5 to 8.3) | +13.3 (12.1 to 14.5) |
US military deaths vs no offer: respected patrol +11 (full annexation) and +7 (Alberta); with warning +36 and +22; at 90% coverage +109 and +191; engaged −72 and −29 (the war ends sooner).
Switch effect with everything on (all on vs all on minus this switch): +0.9 full annexation, +2.5 Alberta, 0 in coercion.
What it means
- The patrol matters as a tripwire, not as air power. A respected patrol at realistic coverage moves the outcome by about half a point. A patrol the US shoots down moves it by 4 to 12 points, because allied dead cost Washington more at home and abroad than the patrol ever cost it in the air.
- So the US's best move is to leave the patrol alone, and the allies' offer has most of its value only if Washington is expected to shoot. The model's 50% chance of that is the single most important number here and the least supported.
- Warning is what makes air cover real. Coverage triples with three months' notice. A respected patrol then adds 2 to 5 points, mostly by keeping Canada's government in place and materiel flowing.
- The upper bound is large. At 90% coverage a respected patrol is worth as much as a shoot-down. European air forces cannot produce that over Canada today; the gap between 19% and 90% is tankers, basing and independence from US-run support systems.
- It matters more for Alberta than for the whole country. A limited seizure rests on thinner US support, so the same shock moves it further.
Limits
- One allied "air force". No split between the UK, France, the Nordics and others, and no link from each ally's tier to its aircraft.
- No Canadian air force interaction, no attrition of the patrol while respected, and no US pre-emption of the deployment.
- Campaign length barely moves because it is bounded by minimum campaign days; the cost shows up in deaths.
- The model has no nuclear dimension. Shooting down British or French aircraft is the kind of event where that omission matters most.