Guide

Game flash, stun and blind systems explained

Harbor Strike shipped its competitive defuse mode with a single utility grenade: a flashbang that applied full whiteout for eight seconds to every player within a six-meter sphere, regardless of whether they faced the detonation, stood behind cover, or had their back turned. Line-of-sight was ignored — flashes popped through concrete warehouse walls. Attackers learned to cook and toss blindly into choke points; defenders reported “I never saw anyone and I'm dead.” Round-ending engagements where the losing side was fully flashed accounted for 63% of entry deaths on the A site. The refactor introduced exposure-based intensity, facing-angle falloff, partial cover attenuation, shorter peak whiteout with afterimage tail, and tinnitus audio instead of total silence. Unfair flash deaths fell to 17%; site take success rose from 44% to 61% because defenders could still trade after a partial flash. Flash, stun, and blind systems are the authored rules that temporarily remove a player's ability to see, hear, or aim — distinct from damage throwables covered in our grenade systems guide.

This guide covers exposure calculation, effect channels (vision, audio, input), duration curves, self-flash and team-damage policy, counterplay affordances, the Harbor Strike refactor, a technique decision table, pitfalls, and a designer checklist.

What flash and stun systems do

A flashbang is a utility throwable whose primary payload is sensory disruption, not hit-point damage. Successful design makes the effect readable (players understand why they were blinded), counterable (turn away, take cover, use abilities), and bounded (worst case lasts seconds, not an entire round). Stun grenades overlap but often emphasize movement and input lock rather than pure vision loss.

Core effect channels:

  • Vision — full whiteout, desaturated blur, peripheral vignette, or UI-only flash for accessibility modes.
  • Audio — tinnitus ring, high-pass muffling, directional deafness; pairs with footstep detection to create temporary information asymmetry.
  • Input / aim — sway, reduced turn speed, blocked ADS; use sparingly in competitive modes.
  • HUD — suppressed minimap, removed crosshair, scrambled compass; telegraphs severity without nausea-inducing full-screen effects.

Most tactical shooters combine vision + audio. Pure whiteout with no audio cue feels arbitrary; audio-only stun without visual feedback reads as a bug.

Exposure calculation: line of sight and facing angle

The heart of a fair flash system is exposure — how much of the detonation energy reaches the victim's eyes and ears. A common pipeline:

  1. Range test — is the player within max effect radius (typically 5–8 m for competitive)?
  2. Line-of-sight ray — from detonation point to head bone; blocked rays reduce or zero exposure. Partial cover (crates, doorframes) should attenuate, not binary block, unless fully occluded.
  3. Facing dot product — compare player look vector to vector toward flash. Direct view = 1.0; perpendicular = ~0.4; back turned = ~0.1–0.2. Scale peak whiteout duration by this factor.
  4. Exposure accumulation — multiple flashes in a window stack with diminishing returns (cap total blind time).
exposure = base_intensity × los_factor × facing_factor × cover_attenuation
peak_blind_ms = min(max_blind_ms, exposure × max_blind_ms)

Harbor Strike's pre-refactor sphere ignored steps 2–3. Adding a head-bone ray and facing curve meant a defender looking away at a doorframe might take 0.8 s of partial blur instead of 8 s of whiteout — enough to react if audio telegraphs the push.

Duration curves: peak whiteout and afterimage tail

Flat-duration blinds feel punitive in high-TTK games. A two-phase curve reads fairer:

PhaseTypical durationPlayer experience
Peak whiteout0.4–1.2 s (scaled by exposure)Cannot aim; must rely on audio and map knowledge
Afterimage / blur1.0–3.0 s decayCan shoot with penalty; crosshair returns gradually
Tinnitus tail2.0–4.0 s fadeFootsteps muffled; comms still audible if VOIP separate bus

Competitive modes should cap total functional blindness (cannot distinguish enemy silhouette) at roughly 1.5 s for a direct full exposure. Casual or PvE modes can stretch peaks for power fantasy. Always expose remaining effect time on the HUD so victims know when they can re-engage.

Self-flash, team flash and friendly-fire policy

Self-flashing is a skill expression when exposure rules are honest: cooking a grenade and peeking too early should punish the thrower. Team flash is more contentious. Options:

  • Full team damage — realistic; slows pub coordination; pairs with friendly-fire systems.
  • Reduced ally exposure — teammates take 50% intensity; common in matchmade tactical shooters.
  • Hard immunity — allies never flash; simplifies ranked but removes bait-and-trade depth.

Harbor Strike chose 50% ally attenuation plus a distinct orange-tint ally flash VFX so players could tell self/team from enemy utility. Ranked kept self-flash at full intensity; death cam labels “self flash” vs “enemy flash” to improve learning.

Counterplay affordances

Flash systems fail when victims have no agency. Ship at least three counterplay vectors:

  • Look away / pre-aim off-angle — facing falloff must be strong enough that a 90-degree turn materially reduces blind time.
  • Hard cover — full occlusion behind solid geometry; partial cover behind low walls reduces exposure rather than eliminating it.
  • Ability or gear — stun-resist helmet, flash-decoy drone, or a “look away” prompt in training mode. Competitive ranked may omit gear but should keep geometry and facing counters.
  • Audio telegraph — pin-pull, bounce, and pop sounds with distinct mix so experienced players pre-turn before detonation.

In tactical shooter pacing, flash exists to enable entry, not to replace gunfight skill. If flash duration exceeds average time-to-kill at entry range, attackers win by utility alone.

Accessibility and comfort

Full-screen strobing triggers photosensitivity and motion sickness. Mitigations:

  • Reduce flash intensity setting — swaps whiteout for edge vignette + desaturation; keeps timing identical for competitive fairness.
  • Epilepsy-safe mode — no rapid full-frame luminance swings; HUD icon + muffled audio only.
  • Color-blind safe afterimages — avoid red/green-only recovery tints.

Competitive integrity requires that accessibility modes do not reveal enemy positions that full flash hides — same information denial, different presentation.

Networking and replication

Flash effects are client-side for responsiveness but must derive from a server-authoritative detonation event. Pattern:

  1. Server validates throw, fuse, and detonation position.
  2. Server broadcasts detonation + payload type to affected clients.
  3. Each client computes local exposure using synced player pose at detonation tick (or slight rewind for high-latency).
  4. Apply effect locally; do not block damage packets on blind state unless design explicitly disallows firing while fully white.

Mismatch bugs occur when client facing differs from server rewind — cap exposure disputes by favoring the victim's look-away if within one tick of threshold. Log large deltas in telemetry.

Harbor Strike refactor

Harbor Strike replaced the spherical full blind with a layered flash stack:

  1. LoS + facing exposure — max blind 1.2 s direct, 0.3 s back-turned at same range.
  2. Cover attenuation — thin props 40% reduction; full wall block unless flash enters through doorway opening.
  3. Two-phase curve — 1.2 s peak whiteout cap + 2.5 s afterimage decay (scaled by exposure).
  4. Tinnitus audio bus — 3 s high-pass on world SFX; team VOIP untouched.
  5. Pop telegraph — 120 ms pre-pop whistle audible through thin walls at 50% volume.
  6. Stack cap — third flash within 5 s applies 30% intensity; prevents perma-blind chains.

Metrics after four weeks of ranked play: entry deaths while fully whiteout dropped 63% to 17%; attacker win rate on A site rose 44% to 61% (utility enabled entries without auto-wins); flash-related support tickets fell 71%; average round time unchanged (flash no longer stalled rounds in eight-second limbo).

Technique decision table

ApproachBest forWeak when
Spherical full blind (no LoS)Arcade modes, horde PvECompetitive tac-shooters; feels random
Exposure + facing curveRanked defuse, search-and-destroyNeeds clear audio telegraphs to reward skill
Vision-only flashLower audio budget, mobileDeafens counterplay via sound cues
Audio-only stunStealth games, horrorHard to read in fast PvP without VFX
Input lock stunBoss fights, crowd controlFrustrating in PvP without short duration
Ally-immune flashMatchmade casualRemoves team coordination risk/reward

Common pitfalls

  • Wall-hacking flashes — radial check without LoS; players blinded through solid geometry.
  • Excessive peak duration — >2 s full whiteout in a game with <1 s TTK makes utility stronger than rifles.
  • No facing falloff — removes look-away counterplay; defenders cannot hold angles.
  • Silent detonation — no pop or pre-pop; victims cannot learn timing.
  • Perma-stack — three teammates flash same choke with no diminishing returns.
  • Self-flash inconsistency — thrower immune in code but not in VFX; breaks trust.
  • Accessibility afterthought — strobing whiteout with no reduced-intensity mode.
  • Client-only detonation — desync exposure across players; death replays show different blind states.

Designer checklist

  • Define max radius, peak blind cap, and afterimage tail separately.
  • Implement LoS ray to head bone with cover attenuation tiers.
  • Scale intensity by facing dot product; test back-turn reduction.
  • Add pre-pop and detonation audio with distinct mix layers.
  • Cap stacked flash intensity within a sliding time window.
  • Policy self-flash and ally flash; label in kill cam and death recap.
  • Ship reduce-flash-intensity accessibility mode with equal timing.
  • Replicate detonation server-side; compute exposure per client from synced pose.
  • Telemetry: exposure histogram, self vs enemy flash deaths, average blind ms.
  • Playtest entry trades: can a partially flashed defender still win the duel?

Key takeaways

  • Exposure beats radius — LoS, facing, and cover determine fair blind duration.
  • Peak whiteout should be short — Harbor cut unfair deaths by capping functional blindness near 1.2 s.
  • Audio is half the effect — tinnitus and telegraphs create counterplay without extra damage.
  • Self and team flash need explicit rules — attenuation and clear VFX reduce frustration.
  • Accessibility modes must preserve competitive timing — change presentation, not information.

Related reading