Hacker News new | past | comments | ask | show | jobs | submit login
10 years of FTL: The making of an enduring spaceship simulator (arstechnica.com)
177 points by zdw on Nov 6, 2022 | hide | past | favorite | 80 comments



Quite a well designed rougelite especially for its time. You are able to win the vast majority of games and yet the game usually FEELS challenging the majority of the time. If you minmax it there's a few too many situations where stalling for extended periods of time is the optimal strategy, but even this can be remedied in quite a few cases using cheat engine to speed things up when you achieve a "known win" scenario. Once you master hard, playing hard without pause can extend the replay value further.

The best part of the game is the soundtrack. On twitch the FTL soundtrack has to be getting more playtime on broadcasts of other games than on broadcasts of FTL itself. It's great atmospheric space themed music.


I still haven't mastered hard. I can win on hard with crystals, but not with any other ship (though I've come close with the 4 square teleporter mantis ship).

Any non boarding strategy seems non tenable right now for me


It is a matter of luck(and patience). And the right augments, you cannot plan for, but rather have to adopt when you find them. For example a drone recovery arm will make me buy it, even if I do not have a drone control system yet. Hull repair is also a must buy(if you can save up drones), and so is weapon preigniter.

With them, it becomes managable, even on hard - with luck. If I don't have any luck, I can also die on easy.


> It is a matter of luck

but also valuation of spending and damage. Apparently, good players can win about 70% or so of the time on hard, because the game is designed such that the amount of leeway/mistakes you're allowed to make is near zero (but not zero).


On easy, I have to get really bad luck to die; there is just so much scrap, I can buy whatever I want whenever I want.


"I have to get really bad luck to die"

Or a moment of not paying attention and then suddenly your shields are all down and destroyed and before you know it that was it. But I haven't played easy in a while ..


Speaking of the soundtrack, Ben Prunty has done a lot of fantastic work!

https://twitter.com/benprunty


I haven't played FTL in years. What do you mean by "stalling" exactly? Do you mean visiting as many nodes in each region as possible and leaving just before the rebels catch you?

Or waiting for the opposing crew to run out of oxygen?


One strategy I remember was that if you got the right build against some foes you were literally invincible and could have low enough damage that you wouldn't destroy the enemy.

So you could use that to level up your crew to all locations, which was a pretty big gain, especially early on and before / without clone bay (with a clone bay your dupes are expendable and lose skill, so getting them to level 2 is less valuable).


Ah right, I do remember that actually. IIRC you could leave the game running in an asteroid field to level up evasion/shields until they patched it.


Another tactic that comes to mind...

My Shrike minmaxing strategy typically involves o2 depletion to whack enemy ships out and get maximum rewards. But it does frequently require a lot of waiting.


Bording is more fun and faster with the mantis/crystal.

Or burning them out (optional with rock bording party)


No criticism intended. But genuinely: What makes FTL a roguelike game?


I used the term rougelite instead of rougelike. This is explicitly to distinguish it from games like rouge and nethack. FTL has some elements of these games, procedurally generated levels, permadeath, exploration, PvE, RPG elements, crushing difficulty, but not others. Thus rougelite.

Anyways trying to define what qualifies what a "rougelike" is has started countless semantic slapfights (which is why I tried to use the cop-out word rougelite). Please consult a rougelike alignment chart to know where you stand in the great holy war over the definition of the word "rougelike" [1]. Nowadays most gamers think of indie games like dead cells, hades, enter the dungeon, darkest dungeon, crypt of the necrodancer, and ftl when they think "rougelike" and those games are all tagged with "rougelike" on steam and twitch.

[1] https://i.imgur.com/ZLDp5rU.png


As a fan of everything along the Roguelike spectrum, I’m happy enough with the “compromise” labels.

Traditional Roguelike for full-on Berlin Interpretation Roguelikes, like Nethack and Dungeon Crawl Stone Soup.

Roguelite for Slay the Spire, FTL, and similar. Much easier to say than Roguelikelikes or Games with Roguelike Elements.

And the word Roguelike? The battle was lost a long time ago, and most people will understand it to encompass both categories above. (Even though I personally have a tough time taking it to include the three great games in the mechanics radical column of the chart).


I have never thought of KSP as a roguelike/lite, but now that I see it on the chart it actually makes a little sense. I started with Binding of Isaac so I guess I have a fairly loose definition.


Rogue, not rouge.


Perma-death, with a procedurally generated world, combined with a clear destination / goal (to get to the next system). One can beeline straight for the goal or risk more time exploring for the prospect of a greater reward. It matches the "2010s Indie" definition of a roguelike.


People now use the term 'roguelike' to mean more than turn-based, grid-based, permadeath, procedural games.

Usually any game that has procedural elements & permadeath now becomes labelled "roguelike" or "roguelite" by the community, depending.

I'm still a snob about it, to me roguelikes are nethack, ADOM, even newer games like Cogmind and Caves of Qud are roguelikes, but to me, FTL is not.


- navigate a collection of randomly generated 2D maps - pause the combat system to think about your next move - often don't know what's around the next corner - balance decisions about what items/upgrades to buy in shops using a few common resources you pick up from battles and exploration - when you die, your game ends

to each their own but like, it's _pretty_ close!


I'm an absolute purist, to be honest it's just pedantry and not really worth debating, I love FTL, and the entire "roguelike" genre, whether they adhere to my standards or not. The golden age of games is really upon us


Procedurally generated content and permadeath are two core concepts of roguelikes. Games that use these concepts are also called roguelites (with a 't'). Binding of Isaac and Spelunky are two other examples of roguelites.


The two may be jeg ingredients. But there must be something missing. It can't only be death as in death and procedurally generated worlds.


Well, that's not my definition, but certainly one that's widely used. I have seen quite a few heated discussions about what should be called a roguelike though. :)


I have the same feeling about the Fez OST. Fez was good. The soundtrack was great.


I listen to the soundtrack while working. It's truly amazing.


FTL is a great game. There is a super big mod called Multiverse that adds a huge story, 50+ new ships, new mechanics, new aliens, and more. It used to be a pain to install but now there's an exe for it. If you've bestent FTL, I'd highly recommend multiverse


Oh awesome. Ftl has been sitting in the queue for a long time now, might be time to redownload and get that mod


It's amazing. It turns it into a brand new game. More difficult but w waaaay more depth. It doesnt feel out of place in the universe.


The next game that they made, Into the Breach, is also fantastic.

I think Subset Games are true game design greats. Really admire their work.


They gave a great post-mortem at GDC one year:

https://m.youtube.com/watch?v=s_I07Iq_2XM

What was great about it is they fully admit to wanting to give up at one point. They simply couldn’t make the game fun. They stuck with it and eventually made it a great game, but it was persistence that led to success.

Game designers with a lot of success get put on a pedestal and labelled “great”, but they fail as much as anyone, they just push through it. Or they throw their games away, case in point Edward McMillen and Meowgenics.


Into the Breach is a puzzle game that doesn’t look like a puzzle game. As someone who hates puzzles, it was a sad surprise.


I feel the same way exactly. I admire the craft but I don't want to spend time with it. Logic puzzles feel like work after a while. I like games I can play imperfectly.


It’s not a puzzle in one critical sense: there is no single set solution to each level. I would say it’s an abstract, discrete (both turn- and grid-based) strategy game.


I think it's described as a tactics game, to emphasize decisions at the micro level, as opposed to a strategy game which usually is larger in scope and encompasses things like base building and resource gathering.


Yeah that's fair enough! The strategy-part is pretty minimal (stuff like choosing/upgrading your pilots and mechs).


It's a tactics game with perfect information and deterministic actions, which transforms it into a puzzle game if you are inclined to treat it as one. The fact that some puzzles may be unsolvable is incidental (and usually avoidable, if you solve previous puzzles well).

For some people, the puzzle aspect is where the fun is. For others, it's a recipe for frustration.


You can say that because we don’t have an agreed definition of a puzzle game. For me it’s as much a puzzle as chess or go, which is to say it isn’t. For my definition having only 1 intended solution is what separates abstract strategy games from puzzles.


At least to me it feels a lot closer to Sedoku than Civilization.


Sure, it also feels closer to Sudoku than to Tennis. But I think it's closer to a "traditional roguelike" (like Jupiter Hell: https://store.steampowered.com/app/811320/Jupiter_Hell/ or Brogue: https://github.com/tmewett/BrogueCE) than to either.


I enjoyed it a lot. In the harder difficulties I sometimes would spend 15 minutes planning many moves in advance, and it felt rewarding when you crush the game by doing so.

But eventually I grew very tired of the mental effort required. After a long day using my brain, doing it for fun was just not doing it for me anymore.


It's similar to a lot of Japanese "tactics" games in that regard (eg, Valkyria Chronicles). Games which have a lot of elements which make them look pretty free-flowing, but the scenarios end up being rather strict puzzles.


I’ll say that Into the Breach was what I had hoped FTL would have been. But I don’t like rogue-likes and love puzzles. (I suppose logic puzzles were more like what work was instead of debugging AWS configurations…)


This is what I loved about it! Felt like a more fun version of a chess tactics trainer. But to each their own.


I thought I hated puzzles, it turns out that I hate poorly presented and slow puzzles and ended up loving ItB.


I enjoyed the game. However when coming back to it there is always a long relearning curve as I never remember what the mechs do. And then the muscle memory of how to plan/chain attacks takes a while to return... and when it does the fun begins!


I am not a hardcore game player but I enjoy non-graphics-heavy games, and I loved FTL.

It's hard but not frustrating, varied, and short enough to play while waiting for something :)

Runs fine on Mac too!


That's one of the reasons I like it as well, it plays like a board game. You usually start a fresh game, play it to finality, and that's it. Done for the night. Most games are egging you on to play over and over, longer and longer, but FTL is satiating in one game.


I got into board games lately and played quite a bit of On Mars. I wish there was a similar game but more focus on building and upgrading your spaceship.


Do you get FPS drops in asteroid encounters too? I played the game ten years ago on my old 2012 MBP and it ran great. Now running on my 2017 The game slows to a crawl in astroid fights. I'm not sure if it's an issue with the game version, my MacOS version, or something else entirely.


I did not notice it on my 2019 MBP.


In hindsight FTL helped propel the roguelite genre to the heights it currently enjoys, with Hades coming almost a decade later and easily being one of the best games of 2020s (yes I’m calling out now in 2022)


Not really comparable IMO. While FTL was real-time, it played almost like a turn-based game, there was no twitching or real action. Looking at Hades, it seems to be another stressful action game.


Just now replaying Hades, having finished it last year.

In general, I dislike action games, don't want to learn key combos, and like to be able to save my progress.

Yet for all that, Hades is probably (for me) the best game I have ever played.

Action within each room is hectic, but in between (after you clear one and go to the next) there's time to think, power up, adjust your strategy, and then plunge into it again.

The game handles death so well I actually wanted to die many times just so I could go back to the starting point and check new powerups.

It was frustratingly hard (for me, casual gamer that really doesn't play action games) but even on those times I kept pushing because of the story. It's surprisingly good foe a rougelite (and action games in general).

All this is to say, if you haven't played it, give it a try (at least a few runs, and explore the other character). You might be surprised.


I guess real time doesn’t really matter to me in a roguelite. The idea remains the same: start, die, repeat until you get good enough to win. Hades just takes that to the extreme - the player gets permanent bonuses for playing and the story requires multiple deaths to advance.


> Hades just takes that to the extreme - the player gets permanent bonuses for playing and the story requires multiple deaths to advance.

That's pretty standard for roguelites, the metaprogression is usually necessary.

I guess FTL is the one one out there, because the metaprogression is only ships being unlocked IIRC.


You can turn FTL into that by not pausing.


Turn/RT is orthogonal, both exist in roguelites.

Though you're completely correct that Hades is on the more ARPG-side of things (very much in-line with previous Supergiant games).

It's not that twitchy compared to an Enter the Gungeon or Binding of Isaac.


FTL is great, even if it becomes somewhat easy to beat once you figure out the best strategies for every set of weapons/devices the game is throwing your way. If you manage to get past Sector 1, you are almost sure to win.


The one size fit all strategy I found when I played many years ago was to put all my crew in the infirmary and asphyxiate the boarding enemies by opening all airlocks :)


The free expansion added cloning bays as an alternative to medical bays which prevent you from healing your crew but revive them from the dead after a short cloning phase. I remember getting stuck in one situation where my O2 system was destroyed and my crew were stuck in an infinite loop of cloning and running through an oxygen-free ship in a desperate attempt to repair the life support. Funny stuff but survivable if your cloning bay is upgraded (so that it clones faster).


The hard AI is also coded to prioritise targeting your oxygen room when low on oxygen, which in various situations can cause a soft lock where you can't repair the oxygen room without dying and get stuck in the infirmary. Always hilarious when this happens.


That is just one obvious thing to do, of course. However, you need to destroy the enemy ship first. There easiest are teleportation-oriented and lasers-oriented. Teleportation oriented is extremely easy with Crystal crew, easy with mantis/rock crew, a must for 4-pad teleporter ships. Lasers oriented, best works with those lasers that use 2 power and have 3 shots, get 3 of them and enemy flagship is dead very soon. And if you are playing a drone or missile oriented game, you just need to figure out how to use the devices you happen upon on your flight to compliment your deficiencies.


Interesting, I have never felt that I could be sure I was going to win in advance... (losing is another matter)

But then I haven't played FTL unmodded for a long time...


"the whole captain simulation system hadn't really been explored before"

There was a science fiction story a long time ago, probably from the 70s or 80s that I don't recall well enough to identify, but in the story people were playing a war game where each human player was sort of like a general without direct control over their side's forces. The simulation was extremely detailed and the battles were recreations of historical events, but the gameplay was very high level.

When WarCraft and StarCraft came along, I wasn't that impressed because the skills required seemed so micro-manage-y and unrealistic.

Has anyone ever come up with an AI that lets players be commanders at a very high level? Where things unfold in real time, with elaborate graphics, but without the ability to control each individual unit. Even if it was dumb, it could cause interesting problems with command and control.


I think your request needs more clarification. By the letter, Total War (tactical level) would qualify since combat at least nominally happens soldier-by-soldier, but the player has control over larger groups of soldiers only. But I don't think this is quite what you're looking for. The player is still freely able to order units to take action and they will at least attempt to obey instantly, albeit fallibly due to mechanics like momentum, close engagement with the enemy on one or more sides, or routing. In fact I think TW is an anti-example because the units having failed to obey an immediate directive cannot be made to accomplish a general goal without repeated direct commands.

There's also something like Gratuitous Space Battles (or worse, Stellaris' combat mechanics) where the player has no combat control at all, but it also doesn't seem quite right.

I fear the formula may simply not be very fun. GSB may push the envelope on how much you can take the player out of the loop without losing the plot, but even there I suspect it borrows quite a lot from strategic games like Stellaris (or any Paradox grand strategy) where the tactical layer exists solely to validate or falsify the players' strategic decisions in a slightly more dynamic way than "bigger number wins".


Mount & Blade II: Bannerlord. You can play Starsector completely hands free during battles.


FTL's design sems like it would work well on a touchscreen, bu I have never seen a port. The closest we got was the asm.js humble bundle, but I was not able to extract it.

I wish they open-sourced the game, like VVVVVV was. Oh well, at least Fex-Emu, Box86 & co are progressing well :)


It works great on the iPad. In fact, it is a superior experience to keyboard/mouse.


FTL is in the app store on iOS. I remember playing it on an iPad years ago - looks like it's been available there since 2014.


It is currently available for iPad only. The developers apparently felt that the iPhone screen was too small.


If only they had known...


I used to really enjoy watching people stream FTL, but my favorite FTL streamers have become so good at the game that in 80% of the fights, it's not actually a fight. The enemy ship is incapable of damaging them. It makes the game boring to watch when the main point of excitement is whether or not the enemies will be able to sneak in 1 or 2 points of hull damage this entire sector. There's just no chance of losing a lot of the time, no sense of danger, no struggle.

I know this is a very minor late game complaint. The game is tough, and this only effects the very best, but it's something I've thought about many times and I was wondering if anyone else has thought the same?


It took many many years to get to the point the game got played out to the point people could feasibly streak 100 games in a row with random ships without pausing. People just got too good at exploiting game mechanics. People broke Issac long before they broke FTL and I never thought Isaac was a bad game.

I think that FTL has largely been abandoned because it has been played out but it did take quite a few years to get there. Designing a game that is consistently hard without just regularly throwing you seeds that are impossible to win is difficult to impossible and at some point I think you need to move onto a new game.


Ah the good old DolphinChemist and his Completing the FTL Hard/No Pause/All Ships Challenge come to mind :)

https://www.twitch.tv/dolphinchemist/video/28582788?filter=c...


random thoughts: i think some games more execution focus. "harder" is more finesse in execution. some are more knowledge; getting better is arcane knowledge about the smallest details. most are some mix.

I look at what "being good" is, and what devs have to do to add a player-demanded "higher difficulty", and what players have to do beat it. What's fascinating is how the experience changes; Sometimes that means "breaking" the game (check-pointing and cheesing), sometimes not.


Dying in space has never been so much fun.

I've recently discovered this gem of a channel about philosophy and ftl https://youtu.be/L2HpA5u5qKc


"War! Never been so much fun!" https://www.youtube.com/watch?v=PiYuq6Ac3a0


This is why I love HN.

Watching right now !


If you'd like to play something roguelikey in space with more freedom to roam, the Trees Brothers have an incredible series of games that are just super fun (and punishingly hard, if you decide to choose the more difficult modes.)

Check out "Star Traders: Frontiers" here: https://www.tresebrothers.com/




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: