FTL Analysis: The Rebel Flagship

The Rebel Flagship is perhaps the misstep with the most widely problematic implications in FTL's design, and it's sad because the Rebel Flagship is a reasonably competently-designed final boss... for a school of game design that FTL is not designed to fit into.

For the game FTL is actually designed to be, it's a horrendous mismatch that subtly harms most every run well before the point a run actually has a chance of fighting it, its long shadow darkening even the earliest portion of the game.

So to properly contextualize what I'm talking about, I need to walk readers through a set of design patterns I'm sure most people who play games a decent amount are familiar with, but which I'm not aware of having an established distinct term for it. For the purposes of this post, I'm going to call it the 'Traditional Linear Progression' school of design.

The Traditional Linear Progression school of design is, broadly speaking, any game in which the player more-or-less-permanently gains new capabilities over the course of the game, in which the design is bounded such that the devs can design specific elements in the expectation that the player definitely has certain capabilities when, for example, reaching a specific boss fight, and definitely do not have certain other capabilities at a given juncture.

This is probably a bit hard a description to connect to any kind of concrete reality, so let's give some concrete examples of reasonably well-known games and walk through them a bit to get across what I'm talking about. I'll start with what I feel is a particularly good representation of the strictest definition of what I'm talking about, then move to a looser definition so readers hopefully have a better understanding of what the essential elements are.

For reference, I'm not going to be discussing spoilers for either of my examples, if that's a concern to you.

Anyway.

So let's talk about Metroid Fusion.

Most Metroid games are pretty representative of the formula I'm trying to articulate, but Metroid Fusion is notable for being the entry in the series with the most strictly linear design. There are no major powerups that are strictly optional. There are no cases where multiple bosses become available to fight simultaneously where different players could go through them in different orders. There are no cases of major powerups being possible to be acquired in a variable order. You can quite firmly lay out a very exact, specific order of what a Metroid Fusion run does, in terms of where it travels, the bosses it fights, when major powerups are collected, and so on. (There's one highly-demanding secret 'sequence break' that can be done to get a special message, but aside that the game is strict on all this)

In turn, area design and boss fight design gets to reliably account for the player's capabilities at any given point. If a given powerup would make a given boss far too easy a fight, that's okay, because you're not allowed to have that powerup when you fight that boss anyway. If a given boss is designed so a given powerup is basically mandatory to be able to clear the fight, that too is okay, because it's not possible for the player to reach the fight without having acquired the powerup in question. And if a given boss is designed so that a given powerup has no relevancy to a given fight you have the powerup in -well. There's negative things that could be said about that, but it's certainly the case that the player will be able to cope with the fight regardless, even if they might personally be unhappy that their favorite powerup is useless in this one fight or the like.

So that's the strict-definition example: a game where the player has no influence on order of operations whatsoever when it comes to the game's major elements, and in turn the devs can fully leverage this control over the possible range of player experiences to make sure everything works.

For my looser-definition-example, I'm going to talk about Starcraft II, more precisely its campaigns -though first I have to back up a bit to talk about a general RTS convention of relevance.

See, classically in RTS design, it's been a convention of campaign design that once a player gets a New Thing in a given campaign -a new unit, a new building, a new research, whatever- that New Thing is permanently available throughout the rest of the campaign. That is, in the original Starcraft, once the player has acquired Firebats in the Terran campaign, Firebats are available for construction in all later missions (That construction is relevant to, of course), even if a given mission's design is such that the player has no real reason to want to break out Firebats in that mission.

I call this a 'convention' because in most cases there is no meaningful internal sense in which the player actually permanently acquires new capabilities: it's not that completing a Starcraft mission causes the game to internally decide the player now has Goliaths and so unlock access to them in any future missions the player undertakes, it's simply that the player is sent linearly through a series of missions which are constructed and ordered such that Goliaths are available in all missions starting from a specific point and no sooner. If you hacked the game and reshuffled the mission ordering, it wouldn't cause Goliaths to be unavailable in missions they're intended to be available in and available in missions not intended to let the player build them; it would just break this illusion of 'unlocking' things as you advanced.

If you notice some parallels to my commentary about Metroid Fusion: good! These parallels are existent and relevant.

So wrapping back to Starcraft II, one of the things its campaigns does is try to formalize this convention into an actual mechanic; instead of each campaign being completely linear, the player has choice about what missions to take in what order, and then missions do in fact unlock units (And other things) that are then permanently accessible to a given run for the remainder of that run, where different runs doing missions in different orders will result in different lineups of available capabilities in a given mission. The first campaign of Starcraft II (Wings of Liberty) in fact has a large fraction of its missions entirely optional, where a given run can skip Mission X, or Y, or Z yet still complete the campaign, where the final missions have to account for the possibility of there being any number of possible holes in the player's arsenal!

Which is all a cool concept, but Starcraft II quite clearly struggled with it on a more practical level. Wings of Liberty is the only Starcraft II campaign to have optional missions, and the later campaigns are more tightly bounded in general: for example, in Heart of the Swarm the general format is that you have a couple of planets to pick between, and picking a planet then locks you into a 3-mission chain you must complete in a fixed order before being returned to the planet selection layer. Furthermore, the planets are organized in 'blocks', where the player MUST complete all planets from a given 'block' before they can reach the next 'block'; the final missions of the Heart of the Swarm campaign are designed under the assumption the player has all the capabilities one can unlock in it, because the campaign has been designed so that this is in fact guaranteed to be true.

And, well, yeah. The concept is cool, but the wide array of possible lineups in what the player may or may not have at any given point in Wings of Liberty is basically impossible for the game to account for in a meaningfully-balanced way. There are missions that can be basically trivialized by putting them off until you unlock a specific unit from a mission that takes a while to gain access to, missions that can become basically impossible if you've skipped several key capabilities, and entire categories of concept even the Wings of Liberty campaign gave up on entirely; stealth-using enemies essentially don't exist in the Wings of Liberty campaign, because Wings of Liberty can't count on the player having detector units to be able to fight stealthed enemies at all! Which is an ugly solution that makes detector capability on player units functionally a lie, because there's nothing to detect!

Hence why Heart of the Swarm and Legacy of the Void tightened the leash and made things much more linear and predictable; Wings of Liberty's situation was simply untenable.

The troubles Wings of Liberty had in turn wrap back pretty nicely to the problems with the Rebel Flagship's design; the Rebel Flagship is designed as if it's been placed in a game like Metroid Fusion, while actually existing in a game with even less bounding than Wings of Liberty's campaign, and so in turn it suffers many of the same problems the Wings of Liberty campaign does.

But more so.

Much more so.

So let's start with probably the most egregious example of the Rebel Flagship's design being horribly out of place to the game it's in: its response to killing all of its crew.

Normally in FTL, there are two possible responses to an enemy ship having zero living crew: response one is that nothing happens because it's an Autoship and it never had crew in the first place. You can't crewkill Autoships, simple as that. Response two is that crewkilling a ship is better than wiping out its Hull points: crewkilled ships generally give better rewards, no enemy ship will ever respond to crew deaths by trying to run the way some are willing to respond to Hull loss, and a number of ships (Such as slaver ships) have special responses to being crewkilled that are even more of an improvement than general crewkilling already results in.

The Rebel Flagship takes an awful, awful third option and punishes the player for wiping out its crew, becoming an Autoship in response. This is, in an immediate sense, already quite awful as it encourages the perverse behavior of leaving exactly 1 enemy crew alive so the ship doesn't turn into an Autoship but is still crippled by lack of crew, but actually the bigger thing I'm driving at is how the game is otherwise clearly designed so all-inning on a pure boarding strategy is supposed to be a viable way to play... and surprise! Not against the only fight that matters for beating the game!

To contextualize my point, I'll return to something I said about Metroid Fusion as my strict-definition-example: that in a design like Metroid Fusion's, it's actually okay for some of the player's capabilities to be entirely irrelevant to some fights. Even to the final fight, potentially. In Metroid Fusion the player can't commit to a strategy based entirely on a specific set of tools and then be unable to beat the game because those tools are irrelevant to the final fight of the game: even if the player likes using some tools a lot more than others, they definitely have all the tools they need to beat the game, because the game made sure they did.

FTL, however, is not such a design; the only guarantees on a selected ship are what it starts with, and FTL is perfectly happy to have ships start with, for example, literally no offenses outside boarding. In the type of design FTL is actually using, a final boss needs to be designed to be beatable with a broad range of possible strategies, and preferably shoot for being roughly even in difficulty for each possible major strategy. 'You can get to the end of the game on the strength of boarding, but then will automatically lose if that's your entire strategy' is a major design fail in FTL's framework, doubly so when the game itself so clearly signals that boarding-centric strategies are meant to be more or less equally viable to other strategies; I'd be much less firm on this criticism if all player ships started with actual weapons, where it would still be possible for a player to commit to a pure-boarding strategy (By selling off their weapons and all) but it wouldn't be implicitly endorsed by the game as a viable thing to do.

Perhaps more significant but less obvious of a problem with the Rebel Flagship is the very fact that it's a fixed 3-stage fight whose loadout does not vary across runs. Yes, whether advanced edition content and hard mode are enabled or disabled changes its layout and capabilities, but bringing that up is missing the point here.

Again: in the Traditional Linear Progression framework, this predictability is fine, in fact it's good. It means the devs get to tune the final fight of the game carefully to make an interesting challenge that's roughly as difficult as they want it to be for as many players as possible. Having the final boss of Metroid Fusion be randomized at the start of a run would be a very janky, problematic thing to do.

But in FTL, it means the Rebel Flagship casts its shadow over the entire run in terms of viability of choices, and FTL clearly does not recognize this fact; I have seen Roguelikes and related games have a fixed final boss where the design intends for the player to be actively working toward being able to deal with the peculiarities of that final boss, where it's understood by the design that anything that doesn't help against the final stages of the game is only okay in the design if it's intended to be something the player makes use of temporarily and ultimately displaces with things relevant to the endgame. FTL does not conform to such a framework, even though what it did with its final boss necessitates such a design mentality; among other points, FTL defaults to giving the player little wiggle room on purchase power, where eg buying a tool that will ultimately be useless in The Last Stand or against the Rebel Flagship is difficult to justify unless it's going to be amazing in the meanwhile, and then such tools are... not designed to be that. (eg the Augment that delays the Rebel Fleet advancement by exactly 1 Jump per Sector -which is a really mild benefit in general, and then does nothing in The Last Stand. Exactly why am I supposed to think this is a worthwhile investment of Scrap?)

Among other points, the Rebel Flagship being fixed directly undermines a lot of the potential for run variety; I've had runs in FTL that had really long streaks of atypical situations (eg fighting a lot of Autoships across the run, or fighting multiple Drone Control ships per Sector), where in theory this could've produced interesting pressures toward trying non-standard strategies that would be better-suited to these unusual conditions, but in practice it didn't matter because I still needed to prepare for the Rebel Flagship and so my choices for what to prioritize fundamentally still had their 'center of gravity' in the same place as any other run.

(Though this also touches on a slightly different design problem with FTL; even if the Rebel Flagship were randomized, having a long streak of Autoships is no reason to expect to fight still more Autoships and adjust plans appropriately. FTL would've needed to both give the player more information for planning ahead, and also been designed from the ground-up to be able to create unusual compositions the player can then be informed of ahead of encountering them, for it to even be meaningfully relevant that atypical distributions of enemy capabilities can in theory lend themselves to pursuing divergent strategies)

Perhaps most subtle of how the Traditional Linear Progression thought process poisons the Rebel Flagship's design is the consideration of how poorly-considered its learning curve is: the 'turns into an Autoship if you crewkill it' thing is a really clear example, where a given player is liable to have a run crash and burn by simple virtue of the game having taught them 'crewkills are good' and then in this one case they're bad and make the fight harder and there's absolutely no reason to expect this.

But this is all over the place with the Rebel Flagship. Its four 'weapons' are actually Artillery; the only other enemy ship to use Artillery is an incomplete Rebel Flagship that can only be fought in Rebel Stronghold Sectors (Of which a given run has at most 1 such Sector, and there's no guarantee it generates, and no guarantee the player runs into this under-construction ship even if they go to the Sector), and the only player ships to use Artillery are the Federation Cruisers, and the Artillery icon is really easy to mistake for the Weapon Control System icon, all of which means plenty of players are going to lose runs due to these Rebel Flagship 'weapons' not behaving the way they expect -especially since a sub-component of this is that the Rebel Flagship's missile launcher doesn't run out of ammo, a fact that is completely invisible in actual play and that a player has zero reason to guess on their own.

Similarly, its third stage is the only ship in the entire game to break from the usual rules on Supershields, and does so twice over; it has twelve Supershield HP instead of 5, and the Rebel Flagship will intermittently regenerate the entire thing in full.

Then there's how its second and third stages both have 'power surge' attacks, which are a mechanic unique to the Rebel Flagship and which have individual learning curve issues of their own. (eg the second phase's 'power surge' attack is to throw a horde of Drones at you -so surely attacking the Drone Control will help, right? Wrong)

And on and on, where a player is liable to lose a dozen or more runs not to making any real mistakes but to being presented with a puzzle boss full of unique exceptions that are impossible to learn about in-game except by fighting exactly that boss.

This is a common dynamic with final bosses in games of myriad sorts; final bosses often, for various reasons, are set apart from the standard flow of gameplay, sometimes very far apart. But in most games that have anything remotely approaching this degree of uniqueness and especially of 'gotcha!' trap design? The player gets to load a save that's thirty seconds away from confronting that final boss, where wiping and retrying five times means losing twenty minutes to working through all the unintuitive rules before finally arriving at a successful strategy.

Whereas in FTL you're going to need to spend somewhere between an hour to two hours to get a run from its fresh beginning back to the Rebel Flagship. That's a lot of time-wasting, and it exacerbates the learning curve issues still further by spacing apart the opportunities to study the Rebel Flagship and make sense of what's being seen; instead of making an attempt, failing, loading a save from thirty seconds before the boss, trying again while the first fight is fresh in your memory, and probably making a good amount of progress on making sense of what you experienced the first time, you're going to have to spend at least one full FTL session to get back to the Rebel Flagship, and realistically for a learning player it's going to be quite a few more attempts than that to get back there, probably spaced across multiple days so the memory dimming is even worse...

... so where this kind of final boss design is merely moderately annoying in a Traditional Linear Progression sort of design, in FTL it artificially adds a really long learning curve where the player has basically mastered the core gameplay systems but is still dying to the final boss because they're still learning its long list of unique qualities because their opportunities to learn are spaced so far apart.

So yeah. The Rebel Flagship was designed, effectively, for a completely different game, one where its design would at worst have been a bit obnoxious and at best would've been a pretty solid final boss... but it's in FTL, so it's a miserable, unfun slog that hurts the design on several important levels.

The sad thing is the dev's response to this situation was, when making Into The Breach, to basically just abandon the concept of a distinct final boss. Oh, it has a final mission that's got some unique bits to it, but... it's pretty same-y with the rest of the game's flow ("Survive for 5 turns, twice, while not losing a widget" is only notable for the 2-missions-in-a-row aspect), and makes for an unsatisfying endpoint to a successful run of Into The Breach. Which is unfortunate, as FTL's Rebel Flagship isn't bad boss design, it's boss design that's a bad fit to the game it got put in.

Alas.

(That said, the missile and laser Artillery inexplicably jumping from Level 3 in the first two phases to Level 4 in the third phase is just bad, making absolutely no sense, being highly counterintuitive, and serving no useful design purpose)

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

Though now we turn to the narrative elements of the Rebel Flagship, as these are... much harder to be understanding and sympathetic to.

I've touched fairly often on the fact that FTL signals in many way that Star Trek is a primary influence. What I haven't touched on previously is the exceedingly strange decision that the central plot is, essentially, a Star Wars plot.

To be clear, I don't mean that in a judgment call sort of way of 'I'd expect to see this plot from an official Star Wars video game and not an official Star Trek video game'. I mean that FTL's core plot is literally a cut-down version of an actual Star Wars plot: the very first one, Episode IV: A New Hope. You're a single ship carrying critical information on the key weakness of the enemy's spacefaring superweapon, traveling through hostile territory to deliver this info to the leaders of the people opposed to the enemy, and the conclusion of the story is the destruction of this superweapon after having delivered this information.

In and of itself this is a bit strange as a possible decision; while the 'Star Trek vs Star Wars' fandom rivalry memes are exaggerated and generally pretty silly, the memes survive in part because Star Trek and Star Wars really are looking to the stars and starfaring civilizations with very different attitudes on many, many topics, both major and minor. The disparity is sufficiently vast that it's genuinely difficult to try to fit the core kinds of plot threads of one into the other, because the foundations are simply too different in too many ways.

Which... FTL nicely illustrates through creating dozens of major problems for itself by trying to do it anyway.

The bluntest example is the relative attitude toward official authority figures and/or widespread social norms. Star Trek classically has as its central cast people who are the best and brightest products of their government/society, and as part of this tends to start from a fairly optimistic perspective on things like the goodness of social norms and governmental oversight; the Federation of Star Trek is generally presented as a governmental model that broadly tries to do right by its people and on average succeeds at this goal, instills good moral principles into its most elite members who tend to hold to those values, etc.

Star Wars, meanwhile, gives us a universe that, in the original trilogy, is held with an iron fist by an explicitly bad government that oppresses and abuses its people to advance the interests of the people at the top. The protagonists are people rebelling against the rightful-but-awful government, the most significant antagonists are top political figures in this government, and even what the Empire doesn't rule tightly tends to be at a similar level of bad, where for example the portion of Tatooine we see is ruled more by a literal crime lord than by the Empire, and Star Wars makes no attempt to suggest Jabba the Hutt is a force for good or anything. Even when we go to the prequel trilogy, before Palpatine converts the Old Republic into the Empire, the Old Republic gets presented as perhaps well-meaning but if so it's not very effectual, with entire planets in its territory practicing slavery, corporations able to legally run military campaigns while the government itself lacks a standing army, etc.

All of which is why it's natural that the original trilogy has our protagonists be an underdog force vastly outmatched in a conventional sense by their enemies and having to pick their fights strategically and so on.

So then FTL wants to do A New Hope's plot, but it wants to place it in a more Star Trek-style universe with more Star Trek-style protagonists, which leads directly to the ludicrous scenario of the official government forces being the underdogs vastly outmatched conventionally by the infinite legions of rebel ships.

I've already broadly implied this, but to be explicit this includes that the Rebel Flagship is blatantly taking the narrative place of the Death Star lining up to blow up the Rebel base on Yavin IV in A New Hope, up to and including that FTL's main plot beats insist that you carry critical information on a key weakness of the enemy's doom machine that you're taking to your bosses. (Which, by the way, FTL gets that wrong too, as it doesn't establish what this weakness is and makes no effort to explain why you go fight the Rebel Flagship entirely by yourself) So the Rebel Flagship is the Death Star, only not making the slightest bit of sense.

I genuinely don't understand the thought process here. Grabbing a widely-known plot that functions and then actively sabotaging all its obviously important underpinnings is the sort of thing I'd expect from a parody story actively trying to construct a nonsense plot, to then make fun of how little sense it makes. FTL seems to have just genuinely not seen that it was annihilating the foundations of why this plot works, while systemically performing such destruction.

It's baffling.

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

As an aside, Crying Suns is a game that looks very much to me like the devs saw many of the narrative problems in FTL I'm talking about and then went and used those problems as a foundation for their own FTL-like game. Its narrative framework shares a lot of ground with FTL, such as you being the underdog agent in a hostile galaxy while being one of your government's top agents, but the framing is very different. For example, FTL seems to genuinely intend for the player to believe that the FTL Federation is in fact a fairly utopian government that does right by its people etc while having constructed a setting where that's ludicrously unbelievable, while Crying Suns makes it clear that the Official Line in-universe is that your government is utopian but there's always been problems and critics and so on; when Crying Suns has state propaganda pressing the idea your government is wonderful and great, there's a clear undertone that this isn't the writer talking directly to the audience, and so any evidence it's false is fine because you're not supposed to think it's an objectively true statement.

Crying Suns also, humorously to me, makes more of an effort to cleave to substantive Star Trek tropes like away teams, bringing the scale of the ship more in line with Star Trek and having your cast be primarily officers, and so on, even though it's much less aggressive about aping the more superficial elements and directly referencing Star Trek episodes than FTL is.

I wouldn't be willing to declare Crying Suns a better version of FTL simply because their respective combat systems are so fundamentally unalike (And, honestly, for all FTL's flaws, including with its combat mechanics, I find its core combat mechanics more compelling than Crying Suns' core combat mechanics), but I would argue that Crying Suns is better at being what FTL seems to think of itself as being than FTL is.

That it exists is interesting to me.

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

So that's it, we're done talking about FTL. On to the next project.

See you around.