Guide

Game suppression fire and pin-down systems explained

Harbor Plaza's courtyard firefight pits a four-person squad against two entrenched machine-gun nests behind sandbag arcs. The first build treated every bullet as pure damage: enemies peeked on a fixed three-second cadence regardless of incoming fire, and players who “suppressed” a nest by spraying the sandbags saw no behavioral change. Telemetry logged 68% squad wipe rate on the encounter and 41% of deaths from simultaneous crossfire while one player reloaded. After a refactor built around volume-of-fire suppression meters, flinch-and-duck FSM states, peek cooldowns tied to incoming rounds, and a squad role hook for designated suppressors, wipe rate fell to 22% and successful flanks rose from 14% to 57% — without increasing rifle damage or reducing enemy HP.

Suppression fire (pin-down, volume fire, overwatch) is the combat verb where sustained or threatening incoming fire forces a target to stay behind cover, duck, or delay peeking — creating a window for teammates to relocate, flank, or revive. It differs from cover (geometry that blocks line of sight), recoil and spread (accuracy mechanics), and raw ranged DPS (damage throughput). Suppression rewards volume, placement, and coordination over landing every shot. This guide covers suppression meters and decay, flinch and duck states, peek risk curves, weapon and stance modifiers, AI response tiers, multiplayer fairness, the Harbor Plaza refactor, a technique decision table, pitfalls, and a checklist.

What suppression is (and is not)

Suppression models the psychological and tactical effect of being under fire: even when bullets miss, the target hesitates, ducks, or abandons a planned peek. In games, that effect must be readable and fair — players need to see enemies react to their fire, and enemies need consistent rules when suppressing the player.

Suppression archetypes

Meter-based suppression — incoming near-misses and hits fill a suppression gauge; above a threshold the target enters pinned or flinch state. Common in tactical shooters and squad games.
Volume-zone suppression — a cone or cylinder of sustained fire marks a zone; any actor inside or peeking into it receives a debuff. Simpler to author; less granular per-target feedback.
Event-driven flinch — each near-miss triggers a short duck animation without a persistent meter. Lightweight but can feel random if telegraphs are weak.
Hard pin (scripted) — boss or set-piece enemies lock into cover during a designated suppress phase. Clear for tutorials; avoid overuse in systemic combat.

Suppression is not the same as chip damage through cover (slow HP erosion) or area denial (leaving a persistent hazard zone). It is a behavior modifier that temporarily reduces the target's willingness or ability to expose themselves.

Suppression meter pipeline

A robust meter-based system runs every simulation tick or fixed combat step:

Threat evaluation — for each potential target, gather incoming projectiles, hitscan traces that passed within a near-miss radius, and explosions in suppressive range.
Accumulation — add suppression points weighted by weapon class (LMG > rifle > pistol), distance falloff, and whether the round passed within the near-miss cone versus direct hit.
Decay — drain the meter when no suppressive events arrive for a decay delay (typically 0.4–1.2 s). Faster decay feels twitchy; slower decay makes LMGs oppressive.
State transition — crossing enter threshold triggers flinch, duck, or full pin; crossing exit threshold (hysteresis lower than enter) returns to normal.
Peek gating — while suppressed, peek attempts either fail, delay, or succeed with reduced accuracy and longer exposure.

Near-miss radius is the hidden tuning knob. Too tight and suppression never triggers; too wide and single shots pin elites. Harbor Plaza settled on 0.35 m for rifles and 0.55 m for LMGs at combat ranges of 15–40 m, with debug draw for designers.

Flinch, duck, and pin FSM

Suppression states should be visible and distinct:

Flinch — brief head duck or shoulder twitch; interrupts aim but does not cancel peek. Lowest tier; good for grazes.
Duck — character drops to crouch behind cover for 0.5–2 s; blocks aimed fire from that stance.
Pin — cannot peek or return fire until meter decays below exit threshold or a hard cap timer expires. Used for heavy suppression or multiple overlapping sources.
Broken pin — optional panic state: suppressed enemy relocates to adjacent cover (cover-to-cover shuffle) instead of holding. Raises difficulty; telegraph with audio bark.

State machines should use hysteresis: enter pin at 100 suppression points, exit at 40. Without hysteresis, targets flicker at the threshold and animations stutter. Pair transitions with distinct audio — bullet whizz, muffled shout, radio call for backup — so players learn their fire is working even when HP bars barely move.

Weapon, stance, and role modifiers

Not every gun should suppress equally. Typical weight tables:

Source	Suppression per near-miss	Suppression per hit	Notes
LMG / sustained fire	High	Medium	Bonus while firing past 0.3 s continuous burst
Assault rifle	Medium	Medium	Baseline for Harbor Plaza tuning
SMG / pistol	Low	Low	Encourages close-range DPS over pin
Explosive splash (near cover)	Very high	High	Pin even on indirect hits; watch framerate on multi-target
Suppressed / silenced	Reduced	Normal	Stealth trade-off: less pin, less AI alert radius

Stance modifiers stack: prone behind hard cover reduces suppression taken; exposed stand peek increases it. Designated suppressor roles (class perk, squad ping) can grant +20% accumulation to reward coordination without doubling damage.

AI suppression behavior

Enemies that suppress the player must follow the same readable rules:

Suppressive burst pattern — fire in 2–4 s bursts toward cover edge rather than single-tap peeks; pauses signal reload or reposition windows.
Crossfire setup — one AI pins while a flanking partner advances; if the pin source dies, flanker falls back to cover.
Peek respect — when the player's meter pins an AI, delay peek cooldown by 1.5–3 s scaled by weapon class. Harbor's pre-refactor bug was a global peek timer ignoring incoming fire.
Player feedback — vignette, muffled audio, or reduced turn speed when the player is suppressed; never opaque stun-lock without counterplay ( smoke grenades, dash, ability break).

Elite enemies may have suppression resistance (50% accumulation, shorter pin duration) but should still flinch occasionally so LMG investment feels valid.

Harbor Plaza refactor (worked example)

The courtyard encounter has two sandbag nests (A and B) with overlapping fields of fire. Pre-refactor AI on nest A peeked every 3 s whether or not players shot at it; nest B did the same, creating unsynced crossfire peaks.

Changes shipped:

Per-enemy suppression meter with LMG/rifle weights and 0.8 s decay delay.
Peek cooldown extended by 2.0 + suppressLevel * 1.5 seconds while meter above 60%.
Flinch animation on first crossing of 40% meter; duck at 70%; pin at 100%.
Squad ping “Suppress target” applies +25% accumulation for 5 s from all squad weapons on that target.
Tutorial VO when pin triggers: “They're pinned — flank left.”

Results after 12k play sessions: squad wipe 68% → 22%, successful flanks 14% → 57%, average encounter time +18 s (more tactical, not a DPS check). Player damage dealt unchanged.

Technique decision table

Design goal	Prefer	Avoid	Why
Enable squad flanks	Meter + peek cooldown pin	Raw DPS only	Behavior change opens windows without HP inflation
Fast arcade shooter	Light flinch only	Long hard pin	Full pin slows pace; flinch sells impact
LMG class identity	High suppress accumulation	LMG as slow rifle	Distinct role beyond magazine size
PvP fairness	Symmetric meter + cap	Infinite pin lock	Counterplay and spectating clarity
Stealth bypass	Silencer reduces suppress + alert	Silent full pin	Preserves stealth fantasy
Boss set-piece	Scripted pin phase	Random meter on boss	Readable phases for learning

Common pitfalls

Suppression with no visible reaction — players assume fire is useless; add flinch, audio, and UI feedback.
Same rules for player and AI without tuning — pinning the player for 4 s feels bad; AI can tolerate longer pins.
Infinite pin from two LMGs — cap meter or stack diminishing returns when multiple sources suppress one target.
Near-miss radius wider than cover — pins targets completely behind hard cover; trace to cover edge only.
Ignoring reload gaps — if pin ends instantly when shooter reloads, suppression feels pointless; decay delay bridges gaps.
Peek timers independent of fire — Harbor's 68% wipe root cause; always couple peek to suppress meter.
Chip damage as fake suppression — ticking HP through walls without behavior change reads as unfair, not tactical.
No role for non-suppressors — if only LMGs pin, riflemen feel sidelined; squad ping or near-miss contribution keeps all weapons relevant.

Engineer checklist

Pick archetype: meter-based, volume-zone, event flinch, or hybrid.
Define near-miss radius per weapon class with debug visualization.
Implement accumulation, decay delay, and enter/exit hysteresis thresholds.
Build flinch, duck, and pin FSM with distinct animations and audio.
Gate peek and return-fire attempts while suppressed; expose cooldown to AI.
Weight LMG, rifle, SMG, and explosive sources differently.
Add player suppress feedback: vignette, audio duck, optional UI meter on allies.
Implement squad suppress ping or role bonus for coordinated play.
Cap stacking from multiple shooters; test two-LMG edge case.
Author tutorial encounter with visible pin-then-flank beat.
Log pin duration, flank success, and wipe rate per encounter.
Playtest solo vs squad; verify solo can still progress with grenades or flanks.

Key takeaways

Suppression is behavior, not damage — Harbor cut wipes 68% → 22% without a DPS buff.
Near-miss detection is the core tech — tune radius and decay before animation polish.
Peek cooldown must respect incoming fire — fixed timers ignore the verb entirely.
LMGs earn identity through accumulation — not just magazine length.
Pair with cover and squad roles — pin creates the window; flank consumes it.