---
id: unit-rules
title: A force's battle value is ours to compute, and the same on both sides
status: blocked
repos: [helm, headquarters]
dependsOn: [unit-library]
exitCriterion: >
  A force's battle value is computed the same way in the browser and on the
  server, agreeing with MegaMek on every unit in the library except where a
  deviation is registered as their bug — including the crews' skill, which
  changes it.
---

# unit-rules

Battle value is computed when MegaMek loads a design, and helm borrows it from
`bridge/`. That is enough while a force is a list of stock units and nothing
else, and stops being enough the moment anything about a unit changes.

## What this epic is not, any more

It was written assuming a force builder where players design units, which
needed the whole construction rules layer. That is not what is being built:
players pick from the library, then vary pilots, models and ammunition; fully
custom designs are a much later question, and porting MegaMekLab is
[declined](declined.md).

So **construction validation is out of scope here** until custom design is on
the table. What is left is smaller and sharper, and most of it is arithmetic
over figures helm already has.

- [ ] **Battle value for a varied unit.** Swapping ammunition or an omni
      configuration changes a design, so its battle value is no longer the one
      the bridge computed for the stock unit. This is the first real rules
      work, and it is narrower than a full implementation: the delta from a
      known design rather than a number from nothing.

- [ ] **Double verification across the wire.** The same crate compiled twice —
      wasm in the browser while a player edits a force, native on the server as
      the authority. The failure mode to design against is version skew: a
      cached stale wasm disagreeing with the server looks like a rules bug and
      is not, so every result carries a `rules_version` the server acts on.

- [ ] **Pin the rounding.** Battle value is full of half-point rounding steps,
      and "client and server round differently" is a miserable bug to chase.

- [ ] **The oracle harness still earns its place, with a smaller target.**
      10,988 units each carrying MegaMek's own battle value turns "does our
      arithmetic agree" into a number rather than a hope. The first use is not
      replacing the bridge but checking the skill multiplier and the force
      totals against forces MegaMek scores itself.

## Deferred, deliberately

- [ ] **Construction validation.** Needed when a design can be authored here
      and not before. MegaMek flags 83 units `invalid`, which is a partial
      oracle waiting for it.
- [ ] **Battle value from nothing.** A full conformant implementation, for
      designs the library has never seen. This is the long pole and it stays
      parked until custom design is real.

## Where the campaign layer picks this up, and an open question

Salvage, repair, refit and damage carried between matches came up as wanted.
Some of that is already provided for: a `.mul` records armour points, internal
structure, destroyed locations and remaining shots per unit, so **the
interchange format already carries a damaged unit** and
[campaign](campaign.md) does not need a new one.

The open question is where the line sits. `campaign`'s exit criterion is "a
pilot carries damage and experience from one match into the next", and
[declined](declined.md) puts MekHQ and StratCon outside the project on the
grounds that campaign is "what a pilot carries out of a match, not a company's
books". Salvage and refit read like the company's books. Either the line moves
or the feature is smaller than it sounds, and that is worth settling before
either epic plans around it — refits in particular are what would drag full
construction validation back into scope.

## Done

- [x] **Pilot skill changes battle value, and a force carries pilots.**
      `helm-force` ports MegaMek's 9x9 `bvSkillMultiplier` table and rounds per
      unit the way MegaMek does rather than on the total. The same four Meks
      are 4,809 BV with regular crews and 6,348 with veterans, which is what
      made "a 5,000 BV force" ambiguous until the crews were named.
- [x] **Force-level totals.** A roster totals, with the base and the skill
      adjustment reported separately — a force over budget because of its units
      is a different problem from one over budget because of its pilots.