C.A.T.S.Crunchy Automated Tactical System

← All documents

Coverage Checklist and Gap Analysis

What a military science-fiction combat system must handle, derived from a survey of openly licensed and freely available systems. Functional coverage only — what situations need a rule and what a mechanic structurally does. No rules text, tables, or values from any surveyed system, per the originality policy.

Naming a source in design discussion is permitted. Discussion is not publication.


The headline finding

The open corpus is rich on personal combat and rich on spacecraft combat, and almost entirely empty in the middle.

That middle — squad morale, suppression as a tactical grammar, communications, logistics, CBRN, casualty evacuation, and electronic warfare against the command layer — is exactly what makes a game feel military rather than merely armed. It is unclaimed.

C.A.T.S. should be built there.

The gap list, ranked by opportunity

Situations where no open system serves, ordered by how badly the corpus fails and how valuable the space is. This is the design opportunity list.

#GapNote
1CBRN and protective postureContamination persistence, detection and alarm, dressing time, degraded capability while suited, decontamination, tempo cost. Zero open coverage. Total greenfield
2EW against the command layerEvery open system treats jamming as a to-hit modifier. Nobody makes it degrade activation, orders, and coordination. Highest-leverage unclaimed idea found
3Drones, remote weapons, operator-link severanceDefines contemporary war. Modelled nowhere
4Logistics as a campaign layerWargames assume ammunition away because scenarios are short. That assumption breaks in an RPG
5Medical as a tactical costNobody models the consequence of having wounded — CASEVAC consuming a fireteam's combat power, triage under fire
6Squad-scale moraleNothing sits between individual fear and army-scale mass battle. "Does this fireteam break?" is unanswered
7Indirect-fire sophisticationCounter-battery, time on target, pre-registered fires, danger close — all absent
8Multi-spectral signatureSeparate visual / thermal / radar / emissions profiles defeated by different sensors. C.A.T.S. is already building this from physics — see the space and to-hit models
9Newtonian space movement and vacuum stealthBarely attempted, and emissions-based detection is untouched. Also already ours
10CQB, room clearing, breaching as a disciplineThe entire close-quarters domain is a hole
11Fortification assault, trench fighting, bunker reductionAbsent as subsystems
12Hasty vs deliberate ambush; counter-ambush; counter-sniperAbsent as procedures
13Fire and movement / bounding overwatch as a named, rewarded actionEmerges structurally in one wargame; never explicit
14Communication failure, delayed or garbled orders, delegationAbstracted as activation friction, never modelled as comms
15Boarding and hull breaching, especially at capital scaleNearly untouched
16Ground-vehicle crew stationsOpen systems assign crew roles for starships but never for tanks
17Reactive armour, spall, top-attack, ammunition cook-offThin to absent
18Injury that is both mechanical and narratively invokableEveryone does one or the other, never both
19Non-lethal capture as a real tactical tradeNobody mechanises the decision
20Unit-scale mass combat, mercenary contracts as working subsystemsAbsent

Three of these — multi-spectral signature, vacuum stealth, and Newtonian movement — C.A.T.S. already covers by having started from physics rather than from genre. That is not a coincidence: they are gaps precisely because they need real physics to model honestly.

Suppression is the most important military mechanic

The survey is unambiguous about this, and it bears directly on OQ-26, which flagged that C.A.T.S. currently has no volume-of-fire representation at all.

Three distinct structural solutions exist:

ApproachHow it works
Zone controlFiring consumes ammunition to establish a threatened area, converting firepower into area denial rather than damage
A status conditionSuppression is a state penalising attack and movement, applied by weapons with an appropriate trait
The primary output of shootingThe main effect of fire is suppression; casualties are the side effect

The third inverts the usual assumption and is the most military-feeling of the three.

The pattern worth adopting is two-axis: a short-term suppression flag plus a persistent confidence tier that steps down through willingness states — advance, hold, withdraw, break — as a unit takes casualties or loses its leader. Better training absorbs more suppression before breaking, which is what makes asymmetric fights work without fudging lethality. Rally is then a leader action tested against troop quality, so both quality and leadership pay off in recovery.

This also answers gap 6 (squad morale) with the same mechanism.

Structural ideas worth deriving

IdeaWhy it matters for C.A.T.S.
One progress-track primitive, reskinned everywhereChases, influence, research, infiltration, hacking, repair and hexploration all run on one accumulator. This validates our Margin-and-progress model — we already have the primitive; the lesson is to reuse it aggressively rather than inventing subsystems
Sense taxonomy: precise / imprecise / vagueLayered over an observed / hidden / undetected state ladder. Makes "thermal imaging is imprecise" expressible without new rules — directly useful for our sensor model
Detection as a distinct stepWhether you see a unit depends on range, whether it moved, whether it fired, and its cover. Staying still is a mechanic, not flavour
Cover and concealment kept separateCover is protection; concealment is not being seen. Most systems collapse them, which is wrong for a military game — they are different resources with different failure modes
Phased combat by crew stationCombat resolves in named phases and each crew member acts in theirs. The cleanest solution to "only the pilot has anything to do"
Minor / major / push action tiersGuarantees an understaffed or unskilled seat still contributes without a check
Mobility kill / firepower kill / catastrophic killThree distinct vehicle outcomes rather than one hit-point bar. Maps onto our depth-tier model
Scan as an information ladderA better check reveals progressively more — specs, then defences, then weapons. Information as a combat resource
Two-directional time modifierPricing rushing and preparing symmetrically. Cheap and expressive
Success at a cost priced into the diceRe-rolling a failure costs attribute damage, gear degradation or stress, rather than being GM-adjudicated
Degree of success as a menu of effectsOne roll produces both magnitude and the character of the outcome. Our Margin already does the first half
The technological / analog tag pairTagging gear as one or the other, with a glitch condition on the former, makes EMP, hacking and ion weapons real without bespoke rules
Half-health as an inert trigger hookA threshold that carries no penalty itself but that other rules can key off

Situations C.A.T.S. must not forget

Compressed from the full survey. Anything unchecked here is a hole in the system.

Engagement: initiative structure · action economy · multiple attacks · delay and hold · overwatch and reaction fire · surprise and ambush · detection before engagement.

Position: cover by degree · concealment separately · flanking · elevation · prone and stance · engagement and disengagement · reach · gang-up · forced movement · hull-down.

Ranged: range bands · point-blank and extreme · aiming as spent time · fire modes and autofire · suppression · firing into melee · firing while moving · area and sweep · called shots · reload and ready · jams and malfunctions · optics.

Close: active defences · shields · defensive stance · grappling · disarm, trip, shove · knockback · stun and subdue · breaking objects · helpless targets.

Environment: light levels · night-vision tradeoffs · smoke, including multi-spectral · camouflage · weather · vacuum and decompression · atmosphere types · radiation · zero-g · fire, falling, drowning, poison, disease, deprivation.

Explosives: area templates · grenades and scatter · indirect fire with call-to-impact delay · minefields · demolitions · anti-armour penetration vs protection by facing.

Psychology: morale and rout · suppression recovery and rally · fear and panic · stress as a parallel track · long-term attrition · unit cohesion and veterancy.

Scene and campaign: chases · boarding · mass combat · faction layer · contracts · sieges · orbital assault.

Electronic: electronic defence as its own attack surface · hacking as a combat action · lock-on and target designation · ECM and jamming · drones.

Sustainment: ammunition · encumbrance · resupply · maintenance · fatigue over days · wound states and first aid · bleed-out clocks · CASEVAC.

Licensing conclusion

The survey reached the same conclusion the originality policy already sets, with more force behind it.

Take no license. Write it all ourselves.

Practices already in AGENTS.md and the originality policy — write from understanding with the source closed, derive our own numbers, name everything ourselves, keep dated design notes — are exactly the right ones. Independent creation is a complete defence, but only if it can be shown.

Verification flags

The survey could not settle several licensing details and flagged them explicitly. None affect C.A.T.S. while we incorporate nothing, but they matter if that ever changes: several publisher SRD licensing pages are dead or unreachable, at least one product carries open-license boilerplate while declaring none of its content open, and one widely cited system's books and SRD sit under different licenses. Read the actual declaration in the actual document, every time.

Source: docs/09-coverage-checklist.md — this page is generated from it.