G: I’ll separate this into two questions: SPECIAL-to-weapon interactions, and weapon behaviors that aren’t obvious from the displayed damage number. I’ll favor mechanics that are actually documented or reproducible, and flag anything that seems to be community inference rather than a confirmed rule.

G: [Non-text content: thoughts; content_type=thoughts]

G: The first important correction is already fairly clear: there does not appear to be a hidden “this weapon scales with Strength/Perception/etc.” system. In quest combat, the weapon-independent SPECIAL effects are basically Agility (attack rate), Luck (critical charge rate), and Perception (slower/easier crit reticle); Strength does not make melee weapons hit harder.

The interesting layer is instead the weapon’s attack behavior—especially target restrictions, damage splitting, and whether excess damage can spill into another enemy.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: I found a much cleaner answer than I expected. Fallout Shelter has remarkably little weapon/SPECIAL coupling. Most of the interesting hidden mechanics belong to a handful of weapon behavior classes, rather than individual guns.

First, on SPECIAL: for quest combat and the explorable locations encountered in the wasteland, Agility increases attack frequency, Luck makes the critical meter become available faster, and Perception slows the moving critical-hit reticle so ×5 is easier to land. Endurance matters indirectly because it determined the dweller's accumulated HP while leveling. I could not find credible evidence that Strength boosts melee damage, Perception boosts gun accuracy/damage, Intelligence improves energy weapons, etc. In fact, the current Fallout Shelter SPECIAL documentation explicitly gives Strength no combat function. citeturn362326search0turn362326search1turn269726reddit26

Also, combat inside your Vault works differently. There, SPECIAL does not affect combat at all: the important variables are the dweller's HP, weapon damage, and relevant pet effects. The firing animations don't represent individual attacks; the room effectively accumulates the dwellers' weapon damage and applies it as damage over time. So Agility doesn't make somebody with a minigun more effective against a Deathclaw invading your Vault. citeturn269726search1turn269726search2

One trap worth calling out: when a weapon says it uses, say, Strength or Perception for crafting, that's only the SPECIAL used to determine crafting speed in the Weapon Workshop. It does not mean that SPECIAL scales the weapon after you equip it. citeturn362326search0turn594000search1

Here is the useful exception list you were asking for:

:::writing{variant="document" id="48317" title="Fallout Shelter — Weapons With Hidden Combat Properties"}

• Fat Man family — including MIRV

  • Area-of-effect attack.
  • The displayed damage is NOT inflicted on every enemy separately.
  • Instead, the weapon rolls its normal total damage and divides that damage among every enemy in the room.
  • Example: if an attack produces 30 damage against three enemies, they receive roughly 10 each, not 30 each.
  • You cannot concentrate that damage onto one target.
  • Major disadvantage on quests: enemies tend to remain alive together, so all of them continue attacking you instead of one enemy being eliminated early.
  • Much less of a disadvantage inside the Vault, because Vault combat is resolved as aggregate damage-over-time rather than the quest targeting system.

• Missile Launcher family — including Guided Missile Launcher and Miss Launcher

  • Same area-of-effect / divided-damage mechanic as the Fat Man family.
  • The displayed 20–25 damage of something like the Miss Launcher is its total attack damage, not 20–25 against each enemy.
  • Thus a launcher that looks considerably stronger numerically than a rifle can actually kill an important single enemy more slowly when several enemies are present.
  • Same quest disadvantage as Fat Man weapons: poor focus fire and enemies tend to survive simultaneously.

• Minigun family — especially Lead Belcher

  • Multi-shot/burst attack.
  • The displayed weapon damage is the total damage of the burst, not the damage of every individual bullet.
  • Its major hidden advantage is overkill transfer: if the current target dies partway through the burst, the remaining shots can move to another enemy.
  • Consequently, very little of the weapon's nominal damage is wasted.
  • Allows normal focus fire on dangerous individual enemies.

• Gatling Laser family — especially Vengeance

  • Same important multi-shot / target-switching behavior as the minigun.
  • If the target dies halfway through the firing sequence, remaining shots can hit another enemy.
  • This is the reason Vengeance is often considered better for quests than its raw 22–29 damage number suggests.
  • Dragon's Maw can have better raw damage, but Vengeance wastes much less damage when finishing weakened targets.
  • Particularly useful when rapidly clearing several enemies or focus-firing a boss and then rolling the remainder of the burst onto somebody else.

• Power Fist family — including Ed's Custom Power Fist

  • Unusual melee exception with overkill transfer.
  • It attacks one target with its full damage, but excess damage after killing that target can be transferred to another hostile.
  • That makes it considerably less wasteful than ordinary melee weapons.
  • It remains subject to the peculiarities of melee positioning/engagement on quests.

• Ordinary melee weapons — baseball bats, knives, pickaxes, pool cues, Fire Hydrant Bat, etc.

  • Have an important negative hidden property on quests.
  • Only one melee dweller can engage a particular enemy at a time.
  • Other melee dwellers may literally stand around waiting if there aren't enough enemies available for them to engage.
  • Ranged dwellers can still shoot the enemy being fought by the melee dweller.
  • Consequently, three extremely powerful melee weapons can perform considerably worse as a quest team than their displayed damage suggests.

• Flamer family — including Burnmaster

  • Despite looking like an area weapon, it is single-target, not splash damage.
  • Its attack can be focused normally.
  • However, unlike a minigun/Gatling laser burst, excess damage is wasted when the target dies.
  • So if Burnmaster produces a 20-damage attack against an enemy with 2 HP remaining, the other 18 does not carry over to somebody else.

• Ordinary single-target / single-shot guns and energy weapons — Dragon's Maw is the important high-end example

  • No special bonus beyond their displayed damage, but there is one mechanically important consequence: overkill is wasted.
  • Dragon's Maw's 22–29 gives it exceptional raw damage, which makes it excellent for Vault defense, where weapon behavior doesn't matter.
  • On quests, however, putting 25 damage into an enemy with 3 HP remaining wastes roughly 22 points of that attack.
  • This is why Vengeance can outperform Dragon's Maw in practical quest clearing despite similar/slightly lower nominal damage. :::

The rocket-launcher question you specifically raised therefore has the slightly disappointing answer you suspected might be possible: the listed damage is distributed, not duplicated. A MIRV doing, say, 24–29 damage isn't secretly doing ~26 damage to five targets for ~130 total. It's doing roughly 26 total, divided among them. citeturn269726search0turn594000reddit34

And there's a second hidden number that matters a lot: multi-shot does not multiply the listed damage. If Vengeance rolls, say, 25 damage and visually fires a stream of shots, those projectiles collectively constitute the 25 damage. The advantage is that the game can stop spending pieces of that 25 on an already-dead enemy and redirect the remainder. That's the mechanical origin of its reputation rather than some secretly enormous DPS multiplier. citeturn510823search0turn510823reddit26

So if I reduce the whole system to a practical ranking of mechanics for quests, it's approximately:

multi-shot/target-switching > ordinary focused single-target > AoE/divided damage, with melee carrying an additional targeting handicap.

For Vault defense, much of that hierarchy disappears: since combat there is essentially aggregate weapon damage-over-time, just compare average weapon damage. A 22–29 weapon averages 25.5; a 21–26 averages 23.5, etc. And importantly, the game inventory apparently sorts primarily by maximum damage, which means its ordering can sometimes mislead you about which weapon actually has the higher average. citeturn269726search2

The one area where I would leave a small asterisk is the collection of newer weapons added in recent updates (.50 cal machine guns, Ultracite weapons, alien weapons, etc.). I found their stats and crafting SPECIALs, but I did not find reliable documentation demonstrating some new armor-piercing, energy vulnerability, freezing, poison, stun, accuracy, or other hidden effect. I therefore wouldn't infer one from the weapon's name or animation. As far as the evidence I could find goes, Bethesda mostly treats those names as flavor plus a different damage range. citeturn510823search0turn594000search1

That last point is actually quite useful: “armor piercing,” “incendiary,” “guided,” “focused,” “tactical,” “cryolator,” etc. generally should not be interpreted literally as adding an RPG status/property unless the weapon belongs to one of the behavioral classes above. Fallout Shelter is much less simulationist about its weapons than the main Fallout games.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: Yes. What you saw is almost certainly a real melee/Ripper bug, rather than a creature immunity mechanic. I found several reports that match your description unusually closely—including reports from 2026, so this isn't merely an old launch-era bug.

And I think I identified your weapon: the Surgical Ripper family, added in update 1.18.0. The variants are Rusty Surgical Ripper (12), Surgical Ripper (13–14), Enhanced Surgical Ripper (14–15), and Hardened Surgical Ripper (16–17). citeturn0search0turn0search1turn0search4turn0search7

The bug matches what you saw

A February 2026 Steam report describes melee weapons simply ceasing to inflict damage during quests. The player specifically tested it long enough to eliminate the possibility of merely missing an animation: a solo dweller with a 12–18 melee weapon attacked three mole rats repeatedly, used more than 20 Stimpaks, and none of the mole rats died. Other players in the discussion agreed that this was abnormal rather than the ordinary melee-targeting restriction. citeturn0search6

More strikingly, just this month—August 4, 2026—there was another discussion specifically complaining that melee attacks were visually connecting inconsistently. One player explicitly says that melee weapons used to hit reliably, but now sometimes the enemy's damage animation doesn't occur, while at other times the animation occurs but no damage is actually inflicted. citeturn0reddit37

There's also a May 2026 discussion where someone independently singles out Rippers:

“I've seen rippers miss … but it may be unique to that weapon type”

Crucially, that player had observed it with solo wasteland explorers, so the ordinary “only one melee dweller can engage an enemy” limitation cannot explain it. citeturn0search8

And a March 2026 player specifically noticed the Ripper/Hardened Surgical Ripper apparently doing far less damage than their numbers suggested, although that report isn't conclusive enough by itself to establish the bug. citeturn0reddit40

So your observation actually slots into a pattern remarkably well.

It does not appear to be an immunity/resistance system

I specifically looked for the alternative explanation you proposed. I found no documented weapon-versus-creature immunity system—nothing along the lines of “Rippers cannot hurt mole rats,” “melee is ineffective against robots,” “energy weapons get a bonus against X,” etc.

Bethesda's own explanation of quest combat discusses target assignment and Agility determining attack speed, but gives no indication of weapon/creature compatibility or immunity. citeturn0search10 Likewise, the weapon documentation simply classifies the Ripper as melee and describes the normal one-melee-attacker-per-enemy restriction. citeturn0search2turn0search5

So I'd distinguish three phenomena:

1. Normal melee restriction: Two melee dwellers can't attack the same enemy simultaneously. One can therefore stand around doing nothing. citeturn0search2
2. Visual confusion: When several creatures pile onto a dweller, overlapping health bars can make it look as though their target isn't losing HP when it actually is. This is another explanation players report. citeturn0search6turn0search8
3. Actual melee/Ripper failure: There are current reports in which an eligible melee attacker repeatedly completes attack animations but does no damage at all. Your observation sounds much more like this third category if you followed it through several rooms and saw the same dweller repeatedly swing without any corresponding HP loss. citeturn0search6turn0reddit37

The fact that you specifically remember a Surgical Ripper makes me more, rather than less, inclined toward the bug explanation because I found the independent 2026 observation specifically calling out Rippers as apparently missing attacks. citeturn0search8

One useful correction to the list I gave you before, therefore: I would put an asterisk next to the entire Ripper/Surgical Ripper family for quest use. There isn't enough evidence for me to say every Ripper is permanently broken, but there's enough current evidence that I would not trust its displayed 12–17-ish damage in quests. Swap the same dweller to an ordinary ranged weapon; if the enemy health starts dropping normally, you've essentially reproduced the bug yourself.

And I would not add “certain enemies resist certain weapons” to our mechanics list. I found no evidence for such a system. This appears to be buggy melee resolution, not hidden Fallout-style damage resistance.

面白(おもしろ)いバグですね — an interesting bug, although considerably less interesting when it's eating twenty Stimpaks. memcite

G: Yeah. You'd essentially empirically rediscovered the game's combat meta without having the mechanics underneath it.

“Just bring guns” turns out to bundle several advantages together:

• Ranged dwellers can focus fire on the same enemy, whereas melee dwellers compete for engagement slots.
• You don't risk the current melee/Ripper no-damage weirdness.
• High-end ranged multi-shot weapons such as Vengeance can carry unused burst damage into the next target.
• You retain complete freedom to redirect everybody onto the enemy you actually want dead first.
• And there apparently isn't a compensating system where melee gets armor penetration, higher crits, creature-type bonuses, or Strength scaling.

So melee isn't really a distinct build with advantages and disadvantages. It's mostly a targeting restriction attached to weapons whose larger displayed damage sometimes compensates for it. Once your armory is good enough that you don't need that compensation, ranged weapons become the obvious choice.

That also explains why your earlier playthroughs could teach you “guns work better” without ever making the reason apparent. The interface gives you weapon damage numbers but doesn't expose damage waste, target switching, engagement limits, or AoE splitting—which are precisely the properties that make two nominally equivalent weapons behave very differently.

In other words, your old rule of thumb was sound. Now we can make it slightly more precise: quests → ranged weapons, preferably high-damage multi-shot; Vault defense → highest average damage, largely irrespective of weapon type. memcite