FTL Analysis: Missile Launchers

Missiles all completely ignore enemy Shields... except for Zoltan Supershields, which block them but take full damage from them. They also all burn missile ammunition each time they fire; in player hands, this means you need to ask yourself whether a given fight is even worth burning missiles on. In enemy hands, it... technically means they'll eventually run out of ammo for their missiles, but generally if an enemy ship runs out of ammo the fight has cost you horribly. This is not something you want to be relying on.

Anyway, aside ignoring Shields and burning ammo, missiles are fairly straightforward or boring, in that they use standard damage mechanics and so on. In fact, missiles is possibly the single most boring overall weapon category, as most missiles are really just a stat block, with only one missile variant that is more than a stat block... which is exclusive to advanced edition... and which isn't particularly unusual anyway. I'm not sure why missiles in particular trend so boring; they're even more boring than lasers, even though lasers are the Default Boring Weapon Category!

As an aside, I've seen multiple places on the internet claim that missiles are unique -aside another category of weapon that has overlap with missiles- for being possible to be shot down by other weapons fire on the way in. This is flatly incorrect: if two laser shots happen to line up, they'll cancel each other out, same as if it was instead a laser hitting a missile. This behavior is much more pertinent to missiles, since they have limited ammo reserves, and is much more likely to be seen with missiles because they have relatively slow-moving projectiles, as well as being a lot more obvious since most missiles are not volley-based where most lasers are, so it's a lot easier to not notice a laser vanished because it got caught on an enemy laser shot rather than because it impacted their Shield bubble... but it's not actually a missile-specific mechanic.

Light Missiles


Leto
20 (Only matters for sell purposes)
: 1
Charge: 9 seconds.
Does 1 Normal damage, with a 10% chance to breach, set a fire, or stun the crew in the room. (This is three separate 10% rolls that can all succeed simultaneously)

The Leto can never be bought or looted. Either your ship starts with it, or it will never have it. At least one wiki also claims the AI will never use it, but this is clearly incorrect, as AI missile launchers with this graphic are, in fact, capable of doing 1 damage, which would be impossible if they weren't Leto launchers.

The Leto itself is junk. It's not strictly outclassed by any particular missile launcher, but the Hermes provides much better damage and secondary effect chances for the missile count while only taking about 50% longer to fire and, admittedly, requiring 3 Power, but if you're serious about missile usage then ammo efficiency is a much bigger deal than fire rate or Power needs past the very early game where you probably haven't found or bought a Hermes anyway. The Leto is additionally badly outclassed by Swarm Missiles, which charge faster per shot, can achieve superior ammo efficiency by stockpiling charges, and their need for 2 Power instead of 1 is generally easy to shoulder. (And okay yes Swarm Missiles have no stun chance, but who cares?)

The Leto does, admittedly, have nothing beating it out for a combination of firing speed and Power needs, but this stops being useful fairly quickly. In the early game it can let you knock out Shields just in time for some other weapon to fire that needs the Shield down or prevent an enemy weapon from firing by smashing them before they're charged, but the further you get in the game more likely it is that a given enemy ship will shrug off 1 System damage, such as by having an odd Shield level so 1 damage doesn't affect bubble count. (Or by the AI cheating on Shield System count, but that's for later) As such, the Leto is generally something you should transition away from fairly quickly, and use it only sparingly in the early game to boot, especially if you have any interest in using missiles later. And honestly, even if you don't, it's still the case that it's not uncommon to get an event that lets you convert excess missiles into Scrap or other useful stuff; wasting missiles on Leto shots is really just a bad deal all-around.

The Federation Cruiser A and Zoltan Cruiser B should move to be rid of it as fast as they reasonably can. The Federation Cruiser A in particular doesn't actually benefit all that much from having a Shield-piercing weapon, since its Artillery Beam already ignores Shields, can't miss, and has no ammo limits, with its only particular tradeoffs being that you don't get to control it and it has a long delay between shots even when fully upgraded. Those are not really serious enough flaws to give the Leto a meaningful niche.

In AI hands it's largely inferior to the following launcher, but this isn't very helpful when you can't tell them apart unless you get hit, or have super-high Sensors or Hack Weapons to be able to count Power bars, or have a weapon whose charge rate lets you ID them by timing (eg if it fires simultaneously to your Leto, it's a Leto), or are just really good at counting seconds or otherwise correctly IDing the exact charge time.

It's... not ideal.


Artemis
38 (Only matters for sell purposes)
: 1 (2)
Charge: 11 (10) seconds.
Does 2 Normal damage, with a 10% chance to breach, set a fire, or stun the crew in the room. (This is three separate 10% rolls that can all succeed simultaneously)

Like the Leto, the Artemis can never be bought or looted. Either your ship starts with it, or it will never have it.

Bizarrely, Artemis Missiles found on enemy ships are inexplicably different from yours, charging slightly faster but requiring 2 Power to function, hence the second number in parentheses for each of those. In practice this tends to amount to being just plain better than Artemis launchers in your hands, since the AI gets ridiculously generous ship parameters and will basically never want for Power outside Ion Storms. One second might not sound like much, but it can be the difference between one of your weapons firing vs it not firing because Weapons got damaged, with each of those potentially cascading into radically different outcomes. By which I mean it can literally be the difference between you defeating a ship with only minor Hull damage vs dying outright. (I wish I was exaggerating)

This is one of the design decisions I resent most in FTL, especially since missiles are already functionally better in enemy hands than in player hands. Strictly speaking AI ships are dealing with a missile limit just like the player, but enemy ships aren't interacting with the strategic component of the game, making their missile limit much more limited in its importance. Indeed, if you drain an enemy ship of its entire missile supply, flee, and then return to the same beacon, their missile supply will have magically restored to full, as the game doesn't actually track individual ship states once you've exited combat and just treats them as fresh if you re-encounter them. The Rebel Flagship is a partial exception in that it will remember lost crew going from one stage to the next... but it doesn't remember anything else, and if you Jump out of a stage and come back it resets to whatever it was when you first fought it (Including crew resurrecting), hence why I say it's a partial exception.

All of that means that it's really, really frustrating that the game furthermore lies to you and has the AI have a Missile launcher that's totally-the-same-thing as yours only actually it's better. Missiles are literally the last category of weapon this is a comprehensible decision on: I would get it if the game had made one or two of the beam weapons have an enemy-only version with a better charge rate, as I could buy that was less about making it better and more about compensating for the enemy making no attempt to time weapons fire for maximum effect. Giving the computer a cheating-better-than-yours version of missile launchers is just baffling.

An additional layer of obnoxiousness to this is that the enemy loves to equip the Artemis. As far as I can tell, it's the single most popular weapon to appear on enemy ships throughout the game. This means that even if you somehow manage to aggressively build your Shield defenses ahead of typical offensive tools -which is very unlikely to happen- you're still going to be constantly getting into fights where the AI can unavoidably do hull damage, costing you long-term, and the potential to directly do System damage means they directly open the way for other weapons to potentially start wreaking havoc as well by virtue of smashing your Shields and if the enemy ship tries to flee they can go lol and smash your Weapons to interfere with your ability to stop them from escaping.

That's all the fairly universal ways widespread Artemis usage is incredibly frustrating design. There's literally dozens of less common yet still awful ways for this to screw over the player, getting crew killed, potentially delaying the player's escape from hostile environments and so creating a compounding problem that rapidly results in an unavoidable game over due to one unlucky hit, ad nauseam.

Meanwhile, for the player... the Artemis is a not-very-good tool running on a limited resource the player must husband 99% of the time.

I love the idea that underlies FTL's design, but the Artemis is representative of why I have a lot of trouble liking FTL's actual design. This is incredibly badly-designed, and the utterly bizarre decision to provide the computer a better Artemis than the player when missiles are already obviously intrinsically better for the AI than for the player just makes the whole thing feel gratuitously hostile to the player. Among other points, the random-targeting nature of enemy weaponry, while it means the AI won't ruthlessly wreck essential Systems or Subsystems over and over, doesn't do anything to mitigate the Hull damage that is far more important an issue for the player than for enemy ships. Every point of Hull you strip from an enemy ship except for the very last is worthless, whereas every point of Hull you lose meaningfully costs you for the rest of the run. This is very obviously not at all equal.

It's especially frustrating since they made the advanced edition update, with rather substantial overhauls to a lot of mechanics... and not only didn't address this at all, but in fact the stun chance on Missiles was specifically added in by advanced edition and provides another way for widespread Artemis use to arbitrarily RNG-screw the player. Like yes technically this benefits the player as well, but the player is using actual plans where disrupting them is a serious problem. Randomly stunning the enemy crew can, strictly speaking, help the player, but the inability to make any sort of plan around it means it's 'well, sometimes the RNG will conspire to create a situation in which stun helped the player'. So basically one step removed from being completely worthless.

This isn't even getting into the problems created from it sharing a graphic with the Leto while they're both allowed on enemy ships.

FTL is largely a fairly transparent game, but then it thoughtlessly throws in lying-to-the-player moments like this, and it's awful.

Anyway... the Artemis is the Leto, but slightly less junk. It takes a little longer to fire, but in exchange it does twice the damage and sells for nearly twice as much, making it kind of competitive with other missile launchers. It's actually worth considering holding on all the way into the end of the game, since it can be slipped into your Power requirements fairly easily and just having missile access provides some important utility, and you loot missiles for free intermittently. The Kestrel Cruiser A, Rock Cruiser A, and Slug Cruiser B are thus the best platforms for giving a missile-focused run a try.

Well, not really the Rock Cruiser A because it's awful, but that's for another post.


Pegasus
60
: 3
Charge: 20 seconds.
Fires 2 missiles while only costing 1 missile per volley, with each missile doing 2 Normal damage and each having a 30% chance to set the room on fire, 20% chance to breach, and 10% chance to stun the crew in the room.

Enemy ships never use the Pegasus. Thankfully. Bad enough there's two enemy missile launchers with identical graphics, we don't need three.

Sadly, while a launcher that literally fires two missiles for the price of one sounds great, in practice the Pegasus is frustratingly poor. It has a shockingly long charge time, is too Power-intensive to readily fit into a loadout, and the only capacity in which it's uniquely impressive is that it can singlehandedly and ammo-efficiently get missiles past Defense Drones... and this is only true if you play with advanced edition disabled, as Swarm Missiles have the same utility for less Power. Swarm Missiles do worse damage in general conditions, but when trying to get Missiles past Defense Drones realistically they work out to about the same damage.

It doesn't help that Drone-using enemy ships are fairly rare, and that there's a wide variety of better tools for bypassing them.

The Pegasus is the best missile launcher at inflicting fires per-volley and per unit of ammo, as two 30% chances is more than a 50% chance to start a fire, and it actually is possible to start two fires with one volley, but this isn't a particularly amazing utility. If you want to inflict fires using missile ammunition, the Fire Bomb exists and is far superior at that duty and especially at getting crewkills instead of vaporizing a ship, which is one of the main utilities of setting enemy ships on fire. You also have to consider the fires-per-second aspect -the Hermes fires in slightly over 2/3rds the time with the same fire infliction chance, so it's actually not much worse at inflicting fires per second while being less vulnerable to Weapons being crippled before you get a chance to fire, among other implications.

If the Pegasus only used 2 Power, or if it charged in a noticeably more reasonable timeframe, it would probably be a solid enough launcher. As-is, it's too Power-intensive to easily fit in, and too slow to be good for much of anything, with its most obvious niche -overcoming Defense Drones- being inadequately helpful at giving it a niche.

It is slightly more resistant to Rock Plating triggering than a Breach launcher, I suppose?...

Heavy Missiles


Hermes
45
: 3
Charge: 14 seconds.
Does 3 Normal damage, with a 30% chance to set the room on fire, 20% chance to breach, and 10% chance to stun the crew in the room.

Your 'generic' heavy Missile, the Hermes is a solid enough option and, thanks to ammo mechanics, you'll ideally switch to it over eg Leto/Artemis launchers before you reach the Rebel Flagship so you get more value out of your missile ammo. (Assuming you don't drop missiles from your arsenal entirely, which is a pretty questionable assumption actually)

Straightforward enough.


Hull
65
: 2
Charge: 17 seconds.
Does 2 Normal damage, 4 if the room has no System or Subsystem, with a 10% chance to set the room on fire, a 30% chance to breach, and a 10% chance to stun the crew in the room.

Note that enemy ships will never have Hull Missile launchers. This is probably for the best, as their random-targeting behavior would have very different implications with Hull Missile launchers than with most weapons, as I touched on with the Hull Smasher lasers.

Hull Missiles themselves aren't terribly appealing, unfortunately. It's actually surprisingly common for enemy ships to not even have empty spaces to trigger the double-damage effect, in which case it's basically just an Artemis launcher that takes longer to charge and has a higher Breach chance. Hooray?

More importantly, missile utility in player hands is oriented toward crippling Systems through Shield bubbles, such as damaging the Shields System to make it easier/possible for laser weaponry to contribute real damage. Spending missile ammo for the purpose of inflicting Hull damage is much less helpful than damaging key Systems so your unlimited-ammo tools can take the enemy ship apart.

You'll occasionally make use of them simply because they're one of the least Power-intensive launchers you can find in Stores/as loot -if you play with advanced edition content disabled, they're the only launcher that costs less than 3 Power while being possible to find during a run! Also because the Rock Cruiser A comes with one equipped.

But if you can get get an alternative online, almost any other launcher is more appealing. Even Pegasus launchers are arguably better if you can bear the Power load, which is just... sad.


Breach
65
: 3
Charge: 22 seconds.
Does 4 Normal damage, with 30% chance to set the room on fire, an 80% chance to breach, and a 10% chance to stun the crew in the room.

The fact that this shares a graphic with the Hermes is unnecessarily confusing when it comes to enemy ship loadouts, as while they're both threatening in enemy hands the Breach missile launcher is unequivocally much worse to face, with more reliable damage and a near-guaranteed chance to breach on impact in exchange for a not-that-much-longer charge time; if you're low on Drone Parts and have a Defense Drone, it might be okay to hold off on deploying it against a Hermes, but probably not against a Breach launcher, but it's very difficult to know which launcher you're facing before you've been hit, where it might be too late!

Anyway, the Breach launcher is the 'ultimate' launcher, in terms of doing a lot of damage but with a long charge time and so on, but in practice is primarily notable for being extremely likely to inflict a breach as it does its damage. This makes it unusually effective at getting key Systems knocked out, since enemy crew will almost always have to spend time closing the breach before they can start repairing System damage.

A Breach Bomb II is honestly usually better if you can get a hold of it, as while it misses out on Hull damage, it can't be shot down and always generates a breach while charging faster and requiring less Power. That said, the Breach Bomb II does do less System damage, and the Breach Missile launcher doing Hull damage makes it appealing when dealing with Autoships and the Rebel Flagship, since you have to knock their Hull to 0. If you can fit it into your Power load, there's decent arguments for at least holding onto it to swap in for the Rebel Flagship. The modest chance to inflict fires also can't be discounted -among other points, crew tries to put out fires and then close breaches, so a fire triggering alongside a breach is actually especially effective at delaying System repairs even though you might expect breach chance and fire chance to be anti-synergistic. (Which, to be fair, they are if crew refuses to move in)

In enemy hands, the Breach launcher is worryingly powerful, but counter-intuitively actually tends to be less of a threat than Leto or Artemis launchers, due to the combination of its very lengthy charge time and higher Power demands. 22 seconds is more than enough time to use any number of weapons to do damage to the enemy's Weapon System, and enemy ships are bizarrely likely to shut down high-Power weapons first; doing damage to the Weapon System of a ship with a Leto often leads to the Leto only being disabled once Weapons is wiped out. Doing the same to a ship with a Breach Missile launcher will often knock it out with one point of damage.

Some setups -such as some boarding-focused ships- have decent odds of winning the fight before the Breach Missile launcher fires! Among other points, the player can get a Weapon Pre-Igniter, while enemies cannot.

The big thing is that the massive charge time lets you get ahead of the curve and snowball things in your favor instead of things snowballing against you. A Leto is almost impossible to prevent from firing; if you don't have any of a very narrow list of tools (Defense Drone, Zoltan Supershield, Weapon Pre-Igniter and a lot of firepower, Hacking that you aim at Weapons... that's about it) it will get a chance to damage Systems, start a fire, open a breach, and harm crew. This can then open the way for further damage; maybe the Leto hits your Weapons System, preventing you from knocking out their Weapons, and before you can get that fixed another missile has hit your Shield System, and suddenly the enemy ship's other weapons that couldn't get through your Shield bubbles start doing damage, leading to yet more snowballing.

Whereas there are almost no weapons that take longer than a Breach Missile launcher to fire... and notably, enemy crew always have no experience, and enemies can't have Automated Reloaders or a Weapon Pre-Igniter. To give a concrete example: the Glaive Beam takes a horrific 25 seconds to get around to firing at base. If your Weapons System is manned by a master of Weapons, that drops down to 20 seconds; a Breach Missile manned by an unskilled crew results in 19.8 seconds of charge time. The Glaive Beam won't be able to prevent the Breach Missile from firing off just a more experienced crew, but it certainly won't have its initial charge interrupted by the Breach Missile, and one Automated Reloader will bring it ahead, potentially letting it hit Weapons to interrupt the Breach Missile launcher's charge.

So while the Breach Missile launcher is very lethal -enemy ships will normally have enough Missile ammo to take you from full Hull to dead off just a Breach Missile launcher- it honestly tends to be less worrying a sight than a Leto or, to a lesser extent, an Artemis. (Not that you can tell those apart on sight...)

It should also be pointed out this this only partially applies to Hermes launchers. They share the same high Power demand as a Breach launcher, and their charge time is higher than Leto and Artemis launchers, but not by enough to be so widely vulnerable to interruption/winning before it can fire.

Still, this is one of FTL's more painful errors, that the basic Missile launchers are overall more worthy of dread than the advanced ones...

Bizarrely, the Breach Missile provides a blue option, above and beyond the three blue options all missile weapons can be used for. Even more bizarre is that it's nearly impossible to have cause to use it: it only occurs in the Hidden Crystal Worlds, and the effect of the Breach Missile blue option is directly inferior to using a Crystal crewmember-provided blue option for the event. You literally must have a Crystal crewmember to even enter the Hidden Crystal Worlds, so the only way you'd actually want to use this blue option is if your Crystal died after you got in but before you reached the event.

Why does this blue option even exist?

Other


Swarm Missiles
65
: 2
Charge: 7 seconds. (21 seconds for a full volley)
Can stockpile up to three charges, and fires missiles equal to the stockpiled charges. Regardless of number of missiles fired in a volley, always consumes one missile per volley. Missiles aim in a randomized area centered on the targeted room, and will automatically miss the ship entirely if their destination isn't inside a room. The missiles themselves do 1 Normal damage apiece and have a 10% chance to set the room on fire or breach. (This is two separate 10% rolls that can both succeed simultaneously)

Swarm Missiles are advanced edition content and will not appear if you have it disabled. They also will never show up in enemy hands, not even on ships added by advanced edition.

Swarm Missiles are one of the few charge-up weapons that isn't horrible junk. The ability to translate charge time into ammo efficiency is very good, and in fact they're very straightforwardly better than a Pegasus launcher in this regard, taking 5% longer to launch 50% more missiles; yes, you have a lower chance of breaching or setting the room on fire and can potentially miss via shots not centering on a room, but if you're aiming at a large room or in a dense spot the latter doesn't matter while the former is too random to be particularly influential, especially since missiles are something you use sparingly.

Note that the randomized targeting is the same as with Flak weaponry, but is much more a flaw than with Flak. Ammo efficiency is a concern for Swarm Missiles, where it's potentially a bad idea to even try to use Swarm Missiles against a ship because there's nowhere you can guarantee room-targeting results, and Swarm Missiles don't hit Shield bubbles, so where Flak is still useful for popping Shields even if the ship is made entirely of 1x2 rooms a Swarm Missile lacks equivalent utility.

Indeed, Swarm Missiles are... pretty bad in practice, simply because missiles as a whole are fairly underwhelming for the player. When I said it's a charge-up weapon that isn't junk, that wasn't an understated way of saying that it's good; I merely meant that it's one of the least awful of a largely-awful category.

Still, they're one of the more useful missile launchers, and the ability to launch 1-2 missiles before full charge can be actually be useful if you're trying to prevent enemy weaponry from getting to fire or are very close to finishing the ship anyway. Their low side effect chances are also deceptive, since of course that's per missile; Swarm Launchers actually have something in the vicinity of a 30% chance to trigger each given side effect if you volley a full charge of three. (Making them unusually effective per-volley at inflicting stun, actually) Taken altogether, a given volley of three actually expects a side effect infliction if every missile hits and the room has crew to stun in it. You can't predict or count on any particular effect, but still, they're more reliable than the numbers might suggest. The low Power demand also makes them unusually easy to slip in, and of course they're very effective at overwhelming Defense Drones -which the Rebel Flagship uses one in its second stage, so this is reliably relevant to every run.

---------------------------------------------

All missile weapons will enable blue options in three different events. (Well, except there's a bug that means Hull Missiles don't work for one of the events...) Technically.

In actually, one of the blue events is essentially fake -you offer to use your missiles to help, and get told that the sentiment is appreciated but no thanks. This blue option doesn't actually do anything.

Another of the three is annoyingly bad; you spend a unit of missile ammunition to trigger a fight with a Mantis ship... when trying to leave will often result in the exact same fight, but without spending a missile. Why does this blue option exist, exactly?

The only actually good blue option is limited to Rock Sectors, and involves a mine attaching itself to your ship. If you have no blue options, either nothing will happen, or you'll lose a crewmember and take 6 Hull damage. The missile launcher blue option will spend a missile to reduce the Hull damage to 4, avoid a crew death, and actually provide you a little Scrap -this is the worst of the blue options for the event, and in fact is actually worse than if you gamble on not using a blue option and the RNG likes you, but it does at least ensure no crew dies; crew losses are sufficiently impactful this is actually worth considering, assuming you don't have a Beam Drone or level 5 Engines to do better.

So if you're going into a Rock Sector and don't have either of those, a missile launcher is worth considering holding onto just in case.

... yeah, I dunno why the devs made missiles so thoroughly awful. It's... conspicuous consistent.

------------------------------------------------

I've largely already been over how missiles are far more useful for the AI than the player, but part of what's perhaps not so clearly communicated is that this difference in quality is a somewhat-superficial consequence of how the change in context means ostensibly-identical weapons are effectively completely different tools in AI hands than in player hands. As in, it would be more honest of the game if the missile category was replaced with a weapon category only used by the player plus a weapon category not available to the player; this is how fundamentally divergent their true functionality is.

This is a pretty notable flaw for a game where the enemy and player are meant to be roughly equivalent -FTL doesn't have multiplayer, but the framework of its combat system could support a VS mode readily, unlike a lot of 'player vs enemy' combat systems that are constructed from the ground up so that the player and their tools are completely different from the enemies they fight. That multiplayer potential nicely illustrates how deeply baked-in the assumption of player/enemy equivalence is to the design.

The ammunition component is more straightforward to talk about, because this piece honestly has a straightforward fix: simply make ammo a per-fight missile limit, not an ammo count tracked across battles. (That is, say the Kestrel Cruiser A started with 4 missile ammo; in that case, it would fire four missiles in its first fight and now be out of ammo, and then in its second fight would get to fire four more missiles, and so on, with any unused missiles 'wasted') This would instantly make ammo meaningfully identical between the player and the AI; the AI already only cares about ammo as a per-fight limit, and now the player would too. It wouldn't completely fix the problems with missile design, but even the stuff not directly rooted in ammo would become much less egregious simply because the AI would stop getting missile ammo 'for free' over the player.

Hull is unfortunately another key part of this, and much harder to fix while keeping missiles recognizable, for reasons that require talking about FTL's 'defensive anchor': Shields.

See, FTL is clearly designed so that Shields are the only defense that the game is actually assuming will go up as a run progresses and balancing itself around that. You have evasion, but this is unreliable by its very nature, so the game can't actually balance itself as 'at 35% evasion the player will evade every third attack', because it's random and won't reliably do that. You can get Drone-based defenses, but the game clearly doesn't want to assume Drone usage. Similarly, there's Zoltan Supershields, but the game intends those to be almost exclusive to Zoltan ships, not a standard part of any ship's kit in the late game. Cloaking is meant to be wholly optional for most ships. And so on.

Missiles are then the one and only tool for directly, reliably tearing off Hull points without caring about Shields. (That a player ship has standard access to, I guess I should specify) Only a handful of esoteric defenses provide more-or-less reliable protection against them, and they all fall inside 'the game isn't willing to balance itself on the assumption you have these defenses at all'.

In player hands, this uniquely powerful capability is bounded by the consideration of ammo: the player can't just skip the combat game against 90% of ships by smashing missiles into them until they explode, because they'll run out of ammo in fights it's unnecessary for and regret it when they run into a fight they really do need the missiles for.

In AI hands, this uniquely powerful capability is bounded by -well. Nothing. Nada. Zip. There is no counterbalancing factor. Yes, AI ships have ammo, but as I just emphasized, their ammo count is per-fight; the AI can't waste a bunch of missiles trying to kill you with Autoships in early Sectors and so have the Rebel Flagship unable to fire missiles at you when you reach it, or anything of the sort. They just get a bunch of free, unavoidable-by-default damage on you.

Now, if you made the ammo change I laid out a second ago, the formulation of this problem would change, but the problem itself wouldn't really go away, because right now I'm talking about how inequal Hull is. For the player, every Hull point is a precious, limited resource, where even a series of solid victories can still be chewing away at their resource potential because they keep getting chipped here and there in individual fights. Unless the player gets improbably lucky with events that provide free Hull repairs (These exist, but are one of the rarer types of events; you can easily go multiple runs in a row without seeing one), they're going to be burning Scrap on the repairs, which means less Scrap going to upgrades or other purchases, which means less combat ability -but of course you can't just not spend Scrap on repairs, because Hull running out terminates your run.

Now, in a superficial sense Hull is identical between the player and the AI, in that in both cases Hull hitting 0 means the ship explodes and is done doing things. However, FTL is not tuned around a 1v1 between the player ship and an enemy ship, but rather is the player vs the collective forces of the AI, where the player has to fight through potentially dozens of ships to even get to the end of a run at all. An enemy ship hitting 0 Hull is just one step on the road; the player ship hitting 0 Hull is the end of the road. They're simply not equivalent. It might help to think of this as if FTL was a board game between two players, one playing the role of the Federation hero ship and the other playing the role of controlling all the enemies the Federation ship fights; for the Federation player, one ship going down loses them the game, while for the All Hostile Forces player they can literally throw away a couple dozen ships to take out that one ship because they instantly win if it dies while most of their ships don't matter.

Returning to actual FTL, though, let's go to the concrete example of the death clock: the phenomenon where a ship has enough missile ammo and a strong enough missile launcher that, if they are allowed to fire all their ammo, the other ship is going to die, and so the other ship has to win or disable Weapons or escape or something before that point is reached. This death clock concept applies in both directions, but the significance of it is completely different: for the player, death clocking a ship is an expensive investment of a limited resource, something to be done carefully if at all. For the AI, death clocking the player's ship is just this thing that happens intermittently once a run gets far enough, completely for free.

More sinister is the related consideration of attrition; for the player, every ship with missiles they encounter is a problem because, even if they win pretty handily, they probably took Hull damage. For the AI, attrition simply isn't a thing, up to and including the absurdist element that AI ships will instantly and completely reload and repair the second the player isn't fighting them. This is one of the key inequalities that's why missiles are so much better in AI hands than in player hands; because just having a missile launcher at all is unavoidably inflicting long-term damage to the player where other weapons might not, where for the player missiles are something you don't necessarily want to use at all in a given fight.

All of this is pretty hard to fix while keeping the game vaguely recognizable, because fundamentally it comes back to the player having a strategic resource management game and the AI not, which is actually by itself perfectly valid and reasonable design. You'd have to have missiles simply not be defined around being 'the primary offensive tool for doing Hull damage without regard for Shields' to escape the core issues here.

Also correlated to all this is the problems with missing: broadly speaking, the ability to miss is not an unreasonable idea, being a natural part of roguelike 'feast and famine' design where part of how the player's resource management is tested is that events are not sufficiently clockwork to be able to reliably plan perfectly. I have issues with the exact implementation I'll be coming back to later, but mostly it's fine...

... outside of missiles, where it's deeply, obviously problematic.

In terms of AI ships with missiles, missiles being able to miss basically ends up functioning as a shield for bad design. All this stuff I've said about 'death clocking' and whatnot can technically be pointed to as incorrect statements because missiles can miss; a ship that has 10 ammo and a Breach Missile launcher will expect to kill the player before running out of ammo if the player's evasion is 35% and their Hull is currently 20, and in fact its average damage is 26, a comfortable margin over the Hull value... but hey, the Breach Missile launcher could miss several times more than expected and so run out of ammo without taking out the player ship. You can't actually say it will take out the player.

But that doesn't meaningfully change the actual thrust of what I'm getting at; being saved by RNG doesn't make the core design functional. It just makes it harder for me to word these problems accurately.

In terms of the player's usage of missiles, the problems are a lot more direct and straightforward: the game has designed missiles to be a limited, powerful resource you should only break out when you absolutely need it... and then it can randomly whiff? That's not pressuring the player to be careful in managing their resources or anything, that's just making it so that playing correctly will sometimes fail to work for no real design reason. Mostly, it ends up contributing to missiles being something seriously worth considering skipping entirely, which is clearly very contrary to the intentions of the design.

In such an RNG-riffic game, rare and precious resources like this should almost always be something designed to mitigate RNG. Player missiles should be one of the only tools exempt from accuracy, guaranteed to always hit their mark, and prevented from being a powerful gamebreaker by the limited ammo. Instead, pure RNG is the only defense all ships are expected to have access to of relevance to protecting against missiles, which is basically backwards.

As a bonus problem, this leads into semi-invisible ammo effectiveness issues. The game broadly tunes itself as if your missile ammo needs will rise little, if at all, over the course of a run, clearly operating on the premise that individual enemy ship Hull will rise but you're liable to upgrade to missiles that do more Hull damage per unit of ammo. Leaving aside the question of whether that particular assumption bears up in practice, it's a model that completely leaves out how dramatically enemy ship evasion rises -the earliest enemy ships will have 5-10 evasion (Level 1 Engines, possibly manned), while endgame ships will reliably have 30-ish evasion. Against early ships, even if you never smash Piloting or Engines, it's not that surprising to have every missile hit, even if you literally only use 1-damage missiles to chew through their entire Hull bar.

Against later ships with 20+ Hull and 30~ evasion, even a Breach Missile launcher expects to require 5 or more hits and in turn is probably going to have multiple misses in there. Or alternatively we can say roughly every third ship you fire exactly one missile at will have the missile miss, where firing one missile per enemy ship at the beginning of a run can easily get through an entire Sector without missing even once. Either way, this illustrates that missing rises substantially in prominence, where missile ammunition effectively degrades invisibly -I can compare this to how Store-bought repairs rise in cost as a run progresses, only where that's clearly deliberate and the game's tuning accounts for it, with missiles I genuinely doubt the devs noticed a given amount of missile ammo won't get as much done late in a run as early in one.

One way of looking at this is to look at the price of missile ammo and modify it by looking at evasion values to arrive at the expected Scrap cost of funding a given amount of missile damage. At the beginning of the game, we can take the base price of 6 and just say you're effectively spending on average 6.6 Scrap per missile; an increase, but not much of one. Once enemy ships are routinely at 30 evasion so our ship has to fire roughly 42% more shots, we end up at roughly a Scrap cost of 8.5; this might seem still pretty insignificant, but it's worth pointing out here that Drone Parts cost 8 Scrap per part and do not suffer an equivalent rise in effective cost; the game understands that Drone Parts are better than missile ammo and prices them higher, but not by enough for them to truly stay higher when we take this into account. (Not even getting into the more subjective point that Drone utility compared to missile utility is far, far higher than a 33% increase)

This directly, mathematically illustrates my point: that I don't think the devs genuinely understood how much missiles are harmed (in player hands...) by evasion applying to them.

To be honest, I don't really get why missiles were so mis-designed. There's bits and pieces that sort of make sense, but overall the situation is so badly constructed it's just intuitively awful; the math and whatnot I'm laying out isn't really necessary to see the problem when just playing the game makes it rapidly obvious that missiles are underwhelming in player hands and unpleasantly oppressive in AI hands. Nor is there some sound-in-theory set of design principles underlying all this that just happen to have failed to line up with the coded reality, which is a common and totally understandable design failing.

Of FTL's many confusing design decisions, the collage that is missiles confuses me the most.

This isn't even touching on the Rebel Flagship, which adds new layers of bizarre to the whole thing...

----------------------------------------------

Next time, we move on to Ion weaponry.

See you then.

Comments

  1. I have to say, even this early in the nalysis, you came down on the game a lot harsher than expected!

    I'll definitely agree with you that missiles are kinda wonky in players hands. A big thing that bothered me was that missile launchers were wildly different in how "efficiently" they use missiles, as you point out. So "bad" launchers were even worse than they appeared, since using them was depriving you of uses of a much better weapon down the line. I would have preferred a system in which the ammo-damage conversion was roughly equal across launchers, and it was mostly the volley size / firing rate / side effect chance that changed from weapon to weapon. The point about missile ammo being effectively worse late in a run than earlier due to evasiveness of enemy ships is also an astute point that I hadn't noticed.

    I kind of disagree with your point about the asymmetry between the player and the AI. It's absolutely true that the AI ships are effectively suicidal in that they throw away hull and ammo (in a way the player can't) because they don't have to worry about a "strategy layer". With a better AI, maybe we could have most factions be a lot more careful about these and reduce this asymmetry. But on the other hand, the player's ship is frequently stronger and definitely much more durable than the AI ships, even before we take into account the AI's random targetting : a better-coded AI would just avoid the fight altogether!

    Of course, a solution could be to rebalance all enemy ships to be similar in power to the player's, but have them be much more squeamish about damage, i.e. trying to run away or surrender after a couple of hits. It could be workable, although it would make for a very different experience, and would make to-the-death engagements considerably more costly even if you win.

    The other thing on which I feel differently than you is about the effect of missiles in AI hands on the game balance. I fully agree that you will come out of almost all engagements in which the AI has missiles with some hull damage. It's almost inevitable. But I don't think that's a bad thing. The game is clearly balanced around the notion that you will be taking attrition damage as you go, and some of your resources will be going towards fixing that. In every fight, you have to ask yourself whether that damage will be worth more or less than the resources you'll get from winning. You can (almost) always run away with minimal damage if the answer doesn't look good and/or there's no shop nearby. Yes, sometimes the RNG conspires into straight up killing you with nothing you could do about it, and missiles are frequently involved in those cases; that is absolutely a bad thing, but it's *really* rare, and each run is short enough that I don't really hold it against the game. And furthermore, while the game doesn't expect you to have any one *specific* answer to missiles (stealth cloak, defense drone, hacking being the ones you'll reliably come across), it does expect you to have at least one sometime in the second half of the run. The cloak in particular is a huge help against the rebel flagship "surge" attack, so I would argue it's effectively required (although the rebel flagship has its own set of problems as a final boss) : it won't protect you against all missiles on your way there, but it will against their first missile volley, and there's a good chance you won't give AI ships enough time for a second one.

    ReplyDelete
    Replies
    1. To be honest, I waffled as to whether to do the FTL series at all, precisely because it's one of the most widely-praised games I've ever seen; I've had prior experiences with people assuming I'm being dishonestly contrary when I say 'X popular thing is bad in Y ways', often even if I say I like the thing overall, let alone cases like FTL where I like the idea behind the game but am far more negative on the execution. Missiles is probably the thing I'm harshest on, admittedly, but it's not like the rest of this series is going to be me gushing about how perfect the game is, so...

      Asymmetry-wise, it's worth pointing out here that for their following game -Into The Breach- the devs switched very firmly to a PvE sort of model, where the fundamental assumptions of the game are that your enemies are not equivalent to your forces. (Among other points: enemies move and then telegraph their attack, which only happens at the start of their next turn. The player just gets to responsively attack within a turn, and in fact is free to mix up the order on whether they move first or attack and then move) The devs themselves clearly did recognize that this semi-equivalency approach was problematic and move away from it. It's too bad the devs have been pretty explicit they don't intend to ever do an FTL 2; having played Into The Breach and enjoyed it far, far more, an FTL-style game whose design philosophy (And, to a lesser extent, writing) is more in line with Into The Breach seems likely to be genuinely A Good Game.

      You're correct that attrition is assumed by the design, but fundamentally there's a couple design problems here. The first problem is the equivalency issue I get into heavily here; it's deeply problematic to have missiles in particular serve the design function of forcing attrition on the player when the game is designed so they can't have an equivalent utility in player hands, especially when trying to pretend they actually are equivalent. If the game wanted to ensure attrition was a thing, it should've baked that more fundamentally into the combat mechanics, rather than making the foundational mechanics allow for total invulnerability and then kludging in a widespread exception. (That isn't even adequate to prevent total invulnerability from happening...)

      Delete
    2. The second problem is that the game isn't really designed around attrition being a mechanic. I've played games that want the player to lose by a slow death of a thousand cuts, where each fight is almost certainly costing resources and the player has to judge whether the resources expended are worth whatever they might get out of it, where skipping a fight entirely can be absolutely a clear improvement to your bottom line over engaging everything you encounter. DoomRL, for example, where 98% of the time you're either burning ammo on fights or are a melee build and the game (correctly) assumes this means you're going to take more damage in every fight than if you were shooting things; you get experience for killing things in DoomRL and so ideally shouldn't skip fights too much, but this is a game where the design really does intend for you to make judgment calls like 'this fight isn't necessary and will be too costly for what I'll get out of the engagement; I'm just heading downstairs'.

      FTL, meanwhile, is your usual JRPG level treadmill where you're expected to successfully defeat every foe you encounter, but minus the ability to grind infinitely. Its gameplay curve is centered around trying to squeeze each Sector for all the loot its worth so you can stay ahead of the ever-improving enemies; your fail state occurs when you don't get good enough fast enough to avoid being outpaced by your enemies, and in fact running away is virtually always a mistake that is substantially reducing a run's chance of victory because you're skipping out on loot. Later ships in particular tend to drop so much loot that even if you take a fair amount of damage you're still net positive on Scrap value, where a given fight was only a mistake to stick out if that fight got you killed. A willingness to run at all tends to just result in a slower, even less fun death spiral, where you run from one fight because it's risky and in short order you're just constantly running from fights because they're literally impossible and it's basically impossible to escape this pit of doom.

      FTL's actual core gameplay is more comparable to Loop Hero, or Half-Minute Hero, or the strategic layer of the Firaxis X-COMs (Particularly War of the Chosen), in the regard that the player is trying to arrive at optimal growth speed without overloading themselves at any given step. Attrition is technically a non-terrible tool for supporting this dynamic (Loop Hero makes some use of it to decent effect, for example), but it's not necessary, and it's certainly not necessary to incorporate it in such a broken way. Much better would be for the game to provide the player better information on nearby fights and expect them to pick fights they can win and try to entirely avoid beacons they can't win against. (Or can win, but only at such a great cost they're in the hole) As opposed to the current situation where you have nearly no navigational information by default and Long Range Sensors are barely any help.

      Delete
    3. To be fair, I think a big part of the positive reaction to FTL is that it allows the player to live the Star Trek "redirect power from life support to shields" fantasy. It succeeds well at selling that, even if the balance of its individual mechanisms isn't quite there, to the point I suspect it was never really the goal of the devs. It then becomes a question of personal preference. For me, the lack of balance didn't prevent me from pouring in a lot of hours into it, but I recognize that it means it's not an infinitely replayable game because of it, where a better version of the game might have been.

      It's hard to tell for sure what the devs' intention was without asking them, but I'll hazard two guesses where I disagree with you. I don't think they introduced missiles for the AI as a way to kludge in attrition : I think they added missiles to have a workaround against "invulnerable" heavy shield ships; only after adding them did the current attrition situation surface, and the game was balanced around it. Honestly I have trouble seeing their heavy role in attrition as anything else than accidental / emergent. It's not the only way you get damaged during a run.

      I also don't think Into the Breach is an "answer" to FTL. Their similarities seem "accidental" to me, in so far as they arise due to the same people working on it, not out of a conscious decision to iterate on their original game. They're just very different games in practice, with ITB being much, much closer to a puzzle game than FTL is.

      I agree that escaping from fights should be done very sparingly, but escaping from one or two fights won't doom your run. Or at least it will doom less runs than choosing never to escape. It would for sure be better if the player had more control over what ships they face, but instead you are expected to be able to face any ship composition, with "escape" as an option meant exlusively for when the RNG is severely unkind. I am not sure why all shorter (i.e. <2h per run) roguelikes all seem to trend this way, while the genre is inspired by games in which you had a lot more control on which engagements to take and when (necessary when death can wipe 10-20h of playtime).

      Delete
    4. I don't think FTL works for the 'it's like I'm playing a Star Trek episode!' experience, like... at all. I can sort of see how for example Power as a mechanic was constructed the way it was in an attempt to create a situation where the player *could* do the 'redirect power from X to Y' decision, but the actual mechanics and incentive structures are such that a player should basically never actually be doing that. So I don't think this argument holds water, or more precisely I feel this argument is a far more damning indictment of the game than anything I'm going to be saying in these posts -if the goal of FTL was to create that experience, or if a player goes in wanting that experience, FTL is about as complete a failure at this task as is meaningfully possible. (By which I mean that, say, Tetris is even less alike to Star Trek, but nobody's going to look at Tetris and criticize it for not being Star Trek-like; to recognizably fail at such a task, you have to succeed at least a little, so people can connect what's in front of them to the goal at all)

      The thing is, it's completely unambiguous that FTL hates the idea of the player achieving even a limited form of invulnerability; pretty much every time there's a context in which it could or would make sense for the player to be untouchable for free, the game makes a conscious decision to say 'no' to that possibility, even if it requires SEVERE cheating to make the 'no' work. Boarding events and how they interact with Zoltan Supershields is the most direct example; the raw mechanics of how Supershields work tell us that they ought to make the player permanently immune to boarding events. Nonetheless, the game simply ignores your Supershields with a flagrant lie about how the enemy 'must' have a Supershield Bypass -which said Augment didn't even exist in the base game, and which is a stupid lie regardless since literally every enemy in the galaxy has the convenient ability to bypass your Supershield exactly one time. (If a boarding event is attached to a ship fight, the enemy ship won't be able to retrieve the crew until your Supershield is down... assuming they even have a Crew Teleporter in the first place!)

      Missiles working the way they do is a direct extension of this broader philosophy of the game, including the part where it's specifically constructed as a player-hostile feature: FTL is perfectly happy for ENEMY ships to be completely invincible, and in fact the mechanics of the fleet following you basically require the system be able to construct more-or-less unwinnable fights. (Because otherwise the player would just grind infinitely on the fleet pursuit fights, rather than running in terror from the fleet... well, frankly I think it was a mistake to have the consequence of the fleet catching up being a fight at all, but that's getting into the weeds)

      Which means yes missiles were put in to force attrition on the player, because that is in fact a broader goal of the game and missiles are merely the bluntest and one of the most obnoxious expressions of this goal. (I actually hate the 'Supershield Bypass' nonsense even more, honestly...) Maybe the devs did not consciously set out to implement missiles to fill this function, but they certainly were perfectly happy to accept it filling that function instead of going 'wow, this is awful', because it was in line with FTL's broader philosophy/design goals.

      Delete
    5. I don't think Into The Breach was imagined as a conscious iteration, no. It doesn't matter, though; Into The Breach has many, many decisions that are unambiguously responses to FTL's problems, with the firm switch to a PvE model being merely the most directly pertinent to this conversation. Even little mechanical elements are examples of this, like how a power management concept returns and is reconstructed to be more mechanically meaningful. Same for major narrative elements -I'll be getting into this in a later post, but FTL's narrative justification for the infinite hordes of enemies is HORRENDOUSLY bad, and whaddya know Into The Breach reuses the infinite hordes of enemy framework but provides a completely different justification that works far, far better.

      Even when devs actually do make radically different games, lessons are learned and carried forward. And in this case, Into The Breach is more directly, recognizably a follow-up to FTL than some sequels I've seen are to the games they're ostensibly a sequel to.

      I formulated my point about escape slightly wrong, in that I left out the part where if escape is legitimately the best choice in a situation, that itself is very often a sign your run is already doomed. I've had only a bare handful of runs where fleeing worked out; 98% of the time, either fleeing directly lead to the death spiral, or that first need to run didn't doom me because the actual problem was that I hit the point of 3 Shield bubbles being normal with zero ability to overcome such and so my run was doomed the second it hit that Sector in such a sorry state. The final conclusion of 'running is basically always worthless or counterproductive' is true, I just worded in a slightly-too-narrow way.

      I assume part of why shorter roguelikes are bigger on do-or-die is precisely because the shorter roguelike framework makes it possible to meaningfully explore the space of death having real consequences. Long-form roguelikes that go 'It breathes. You die.' (I'm literally quoting Angband here, for reference) are potentially throwing weeks of playtime in the trash for daring to play the game without spoilers. Non-roguelikes tend to end up with death being annoying but not terribly consequential -being killed by a Donkey Kong Country boss just kicks you to the world map to start the fight over from the beginning, and running out of lives entirely just kicks you to the menu to load the file back up and walk to the boss and start the fight over from the beginning. A short-form roguelike is the only currently-existent genre where death can be meaningfully impactful without it also being horribly obnoxious. Thus: it draws game designs that pursue exactly that space.

      Delete
    6. "There's literally dozens of less common yet still awful ways for this to screw over the player, getting crew killed, potentially delaying the player's escape from hostile environments and so creating a compounding problem that rapidly results in an unavoidable game over due to one unlucky hit, ad nauseam."

      There are players who go on win-streaks on the hard difficulty. There are even players who often succeed at a self-imposed challenge to play on hard without ever pausing.

      This wouldn't be possible if missiles caused unavoidable game overs as you say. Of course it's possible to be unlucky and lose your run due to missiles, however blaming the game over only on this is, in the vast majority of circumstances, a mistake. There ARE counterplay options available, even if they aren't direct hard counters 100% of the time, and utilizing them will allow you to at least mitigate the effect of missiles to an extent where the "unavoidable game overs" they cause is an afterthought.

      Delete
    7. 'Missiles can cause unavoidable game overs' is not a statement busted by 'people have win streaks on the highest difficulty and have won with self-imposed challenges'. It's perfectly possible to play a game with unavoidable game overs and get a lucky streak where the conditions never arise; this is true in any case where random factors are involved, such as in FTL itself. Attempting to argue that such lucky events happening proves unavoidable game overs *must not exist in the system* is just obviously nonsense.

      I'm also not sure why you bring up the hard difficulty at all given the changes it makes don't really factor into why missiles can cause unavoidable game overs.

      As a bonus, you directly and explicitly admit missiles do cause unavoidable game overs in your final sentence, and just present it as if it's a minor issue.

      So... what are you even trying to actually argue here?

      Delete
    8. The point was that it isn't nearly as common as you were implying. Certainly not worth listing as a major flaw, in any case.

      Delete
    9. Having had literally dozens of runs terminate due to one bad missile hit chaining into a complete and total loss, from full Hull, where there was literally nothing I could've done to prevent this (Happened too soon to Jump out, RNG denied me doing damage to their Weapons System or I did damage but they kept the missile launcher active anyway, etc), up to and including that the game had given me literally no opportunity to buy Defense Drones or other anti-missile measures so even the 'you should've prepared better' argument is clearly simply false... It's common. It's unreasonably common.

      Again, 'some people get win streaks' is not a cogent counterargument. I had plenty of win streaks, up to a dozen runs in a row even. Nonetheless, I also had plenty of cases of a run being instantly trashed where there was *literally nothing* I could've done differently at any point, due very specifically to missiles.

      Notably, the devs themselves pretty clearly disagree with you; when Into The Breach came along, it conspicuously failed to have anything resembling FTL's missile design, neither for the player's end of things nor for the enemy's, in spite of there being so many design parallels between the two games.

      Delete

Post a Comment

Popular Posts