For a game that is in large part about sending Star Trek-esque science ships to discover, catalogue and solve the mysteries of the universe, it’s a little ironic that the thing that’s puzzling me most about Stellaris is why on earth it was released in this state.
Seriously. It’s utterly baffling to me that Paradox — who at this point aren’t exactly a struggling indie developer living hand to mouth — pushed Stellaris out in this form, which is just riddled with bugs and incomplete and/or missing features, and at this time, which is the start of what is possibly the most crowded summer release schedule in living memory. It is by no means as broken or buggy as Rome 2 was, but the fact that I am even drawing the comparison is doing Stellaris no favours whatsoever; like Rome 2, it is a game that really needed an additional 3-6 months in the oven before it could be called done. It’s so frustrating to play it and see all the low-hanging fruit that Stellaris just doesn’t bother reaching for; all the small features and improvements that result from a polish phase that have a huge impact on the game’s quality-of-life but which are conspicuous by their absence. Unfortunately that polish phase didn’t happen and Stellaris has been kicked out the door to live or die in a state that is a little malformed and more than a little malnourished.
Stellaris starts off promisingly enough. It’s Paradox’s attempt to fuse the vast storytelling scope of their historical strategy titles with a space 4X game that’s supposed to be simultaneously more accessible and less intimidating than said historical titles. Much of the empire management gumpf that fills something like Europa Universalis IV has been pared right back to the bone, but the element that does make it through — and which helps Stellaris a great deal — are the random events/event chains, which have been tied into the setting in a very organic way. You begin each game by sending out your science ships to survey nearby systems to uncover habitable worlds that are ripe for colonization and/or exploitable resources. Every planet you survey has a chance of throwing up an anomaly, which you can then direct your science ship to research; there’s a fair variety of anomalies in the game, and researching them is not without a risk of failure – which can sometimes lead to the catastrophic loss of the science ship involved — but successful research either results in a one-off reward, such as reactivating an ancient automated shipyard that then spits out a single ship that’s a couple of tech levels up from what you can currently build, or else starts an ongoing quest chain that often has several steps to complete before it is resolved. These quest chains don’t have particularly good material payoffs compared to the effort required to finish them, but they do have lasting effects in the game and provide a heap of flavour that other recent space 4X titles have been sorely missing.
Eventually you’ll research tech that lets you build colony ships, and that’s your cue to grab the 2-3 nearby systems that actually contain habitable systems for your race. Every race has a preference for a specific type of world, and that type of world can be colonised as soon as you can get a ship there. Depending on the preference (continental, desert etc.) they’ll also have lesser preferences for other types of world (so the humans have a strong preference for continental worlds, and lesser preferences for tropical and ocean worlds), but these other types will often require a specific technology to be researched before you’re allowed to colonise them, and you’ll also take a happiness hit (or to be more precise, a lower maximum happiness cap) if the planet type is too far away from your race’s primary preference – nobody really wants to go and live on Hoth, after all. Other technologies provide a small blanket increase to the habitability score of all worlds, and you’ll eventually assimilate other races that have a different preference and so will probably love living on the Arctic world that a bog standard human/reptile/horrible arachnoid will hate, but that’s something that happens much later on. At the start of the game you don’t really care about the other world types, though; all you’re looking for is the magic green globe marker that indicates you’ve found a world that matches your race’s preference.
Once you’ve ordered a colony ship to colonise said world you’ll immediately start experiencing a hefty drain on your energy income; this will persist until the colony has been in existence for a year, and functions as the brake on unfettered expansion that all modern 4X titles have. Once it’s up and running, though, you have a reasonably high likelihood of experiencing another random event chain that’s specific to the planet you’ve just colonised. One of my early planets was inhabited by titanic lifeforms that blocked a good third of the usable tiles; after a decade or so I got an event that gave me a choice between studying these lifeforms or wiping them out. Since I was playing a race of peacenik xenophile humans I went for the study option, and was eventually rewarded by communications being established with the titans, who subsequently provided a limited number of immensely powerful armies that came in useful later on. Another planet encountered a subterranean civilization, and the associated event chain for that went on for nearly a century. A third planet nearly got wiped out by an asteroid and it was extremely lucky I had one of my two military fleets nearby to blow it up before it hit the surface. Other 4X games have done this sort of planet-specific event before, of course, but never in such detail or variety.
The final category of event is what I’d call the social ones, and it’s here that Stellaris starts to get really interesting since I haven’t played another game that’s really tackled how certain sci-fi concepts would affect an interstellar empire – and since Stellaris is basically a grab-bag of every single science fiction concept in existence there is a lot of scope for this sort of event. Take Uplifting, for example. Uplifting is a technique whereby an advanced species genetically modifies a pre-sentient race so that it achieves sentience, and which then becomes a client race of the uplifting species. Uplifting is in the game (once the appropriate technology has been researched) and is useful if your borders happen to contain a pre-sentient race with desirable characteristics, such the ability to settle the radiation-blasted Tomb Worlds that are created whenever a sentient, pre-spaceflight species wipes itself out through atomic conflict. However, an uplifted race that has had its development artificially accelerated might not take to its newfound status all that gracefully, and there are events in the game to reflect this. Then there’s the ability to build observation posts to monitor the development of those sentient but non-spacefaring species. The base behaviour of passively observing them nets you a small research bonus, but if your empire’s outlook permits it you can abduct specimens for active study, or even try and covertly replace the species’ leadership so that you can integrate them into your empire; that last one can have some blowback, however, as there’s a chance they’ll figure out something’s up and set up an equivalent of XCOM to thwart your plans. Finally there’s some treatment of how certain techs would affect your existing population; after researching gene modification I got a pop-up saying that the colonists on one of my planets had modified themselves to be better suited to the existing environment and were now referring to themselves as Trans-Humans. I understand there’s also a detailed number of options and policies for how to treat synthetic lifeforms, although I never saw any of them since I was never offered the relevant technology.
This is as good a moment as any to take a paragraph out from the main thread of the review to talk about the research system. Stellaris has opted for the same sort of blind research I first saw in Alpha Centauri all those years ago — where instead of getting a free pick you’re instead presented with a random choice of one of three available techs — except here it’s super blind since you’re not even able to see a tech tree of any kind. Indeed, I don’t think there even is a tech tree in Stellaris; the fact that the techs you are offered depend somewhat on your empire’s ideology (so xenophobes won’t get the tech that allows aliens to become leaders, for examples) and who knows what other factors would make this more than a little bit difficult. Instead it’s more like being dealt a selection of cards from a deck, with certain rare techs having a lower chance of popping up. The RNG is somewhat ameliorated by research being split into three areas that can be researched simultaneously – Physics, Society and Engineering — and you can also unlock options that hang around permanently until researched by studying space battle debris (so that you can extract that cool weapon your enemy was using and turn it on them). Still, it was frustrating to be surrounded by Ocean planets with a 60% habitability rating (this is quite high in the early game) and not be able to colonise them because the Ocean Colonisation technology repeatedly refused to come up as one of my three picks in the Society category. Since techs take absolutely forever to research – about five years on average, which is around 10-15 minutes of time on the fastest speed setting — you can be sat there for a long time waiting for the option you want to come up. I think blind tech research works really well for a game like Stellaris, where a large part of the appeal is not knowing for sure what the game is going to throw at you next, but there’s a difference between being able to direct your research in a limited way and not being able to plan ahead at all and I think Stellaris comes down way too heavily on the latter side.
Anyway, the Stellaris early game is good – great, even. You’re slowly expanding into adjacent star systems preceded by your science ships and firing off all sorts of interesting events; never mind that the colony management is a bit flat (thankfully you only have to directly deal with 5-ish planets at a time, with the rest being devolved to AI governors in recognition of the fact that micromanagement at scale is deadly boring) and that the economy is weirdly balanced, the events are exactly what I bought Stellaris for. Eventually your science ships run into AI empires, including a specific type called a Fallen Empire — this is another one of Stellaris’s great ideas that’s been pulled straight from the pages of Iain M. Banks. “Fallen” in this case is a bit of a misnomer; the Fallen Empires are actually hyper-advanced elder races which have lost interest in expansion and have become super-isolationist. They could crush you like a bug if they were so inclined, but they mostly consider you beneath their notice — unless you do something stupid, like claim systems close to their borders or settle their holy worlds. They function as a sort of sleeping dragon that you have to tiptoe around, as well as providing pretty much the only source of end-game challenge: teching up to the point where you can fight and defeat a Fallen Empire is kind of like smashing a sixer in conkers.
Unfortunately the point at which you run into the other empires is the point at which Stellaris starts to slow down considerably. As more and more planets are claimed by one empire or another, and as your science ships start to run low on unexplored systems to survey, the supply of events begins to dry up. There are almost no events that will be triggered by a mature empire; they are only triggered by exploration, expansion, or by you actively doing something like Uplifting a race of pre-sentients. Once you run out of places to explore or expand into you run out of events, and this is where Stellaris starts to encounter some major problems because as it turns out there’s actually nothing else to do in the game except tech up — which is impossible to plan as mentioned previously and so boils down to the one-of-three choice every 5-10 minutes — and fight people. There are a whopping two victory conditions in Stellaris: you can either kill everyone else (which involves fighting them) or you can try and claim 40% of the systems in the galaxy (which is going to involve some fighting no matter which way you slice it). Either way you’d better start gearing up for some of the dullest, most frustrating wars you will ever fight in a 4X.
Let’s start with the nitty-gritty of the battles themselves. Stellaris has a pretty good ship designer that takes full advantage of the rock-paper-scissors system between missiles, lasers and mass drivers that all space 4Xes seemingly have to have these days along, with some extra bolt-ons like point defence and torpedoes. You can make just about any ship that takes you fancy – fast torpedo boat corvettes or picket ships with high-powered sensors, destroyers configured for shooting down enemy missiles and fighters, battleships that are converted into carriers or which are simply very, very good at killing other battleships. Unfortunately since every battle immediately degenerates into a huge blob of ships firing at each other I had absolutely no idea whether my ship designs were doing what they were supposed to do or not. Were the point defence destroyers shooting down torpedoes aimed at my battleships, or were they just protecting themselves? Were my battleships preferentially targeting enemy battleships and cruisers with their large weapons, or were they wasting their shots at fast, nippy corvettes? I just couldn’t tell; it’s nigh-impossible to figure out what’s actually going on because the game doesn’t surface it in any detail besides the aforementioned ship-blob that can often contain a hundred vessels on each side.
Tactically the business of fighting a war is a bit of a wash, then, but strategically it gets even worse. Stellaris takes a leaf out of Sword of the Stars’ book by having distinctly different propulsion types in the game: hyperlanes, wormholes and warp. Wormholes are little bit niche; your ships themselves can’t travel between the stars and have to rely on special wormhole stations to send them on their way, which sounds interesting but I never saw anyone actually use them. Hyperlanes are the standard collection of fixed travel paths connecting stars a la Endless Space; if there’s a star close by that doesn’t have a hyperlane route to it in your territory you can’t get to it despite its seeming proximity, which provides lots of strategic potential for fortifying chokepoints generated by the hyperlane layout. Unfortunately this is completely undone by the third propulsion type, warp drive, which simply lets ships travel anywhere in a set radius. You can imagine how this works when handed to an AI that simply does not give a fuck: it’s very good at running away when its ships are outnumbered, meaning that if one fleet gets into your empire (which it will, because warpdrive ships make a mockery of borders and can go anywhere they have the range to get to) you end up chasing it around endlessly. You can build defence stations that have warp interdictors in them, but the strongest station I was able to build had a military strength of 1.5K; a small mid-game fleet will have a strength of at least 5K and will rip said station to pieces long before a relief fleet can get there. Warp drive reduces invasions of your empire to a gigantic game of whack-a-mole, and it is awful.
How about when you invade another empire, though? Well, the actual business of taking and holding planets isn’t too bad as you soften up with a preliminary bombardment and then drop invasion armies onto the planet surface. The problem with invading other empires is more a political one: one of the things Stellaris has inherited from the Europa Universalis series is the War Goals system, where you state at the outset of hostilities what you’re trying to achieve by fighting this war. You goals will usually be for the enemy empire to cede you a certain number of planets, but — bizarrely — each planet you set as a goal has a warscore value associated with it, and you can only set goals up to a total warscore of 100, which equates to about three planets. (Curiously the total vassalisation of the enemy empire only has a warscore of 60.) Even if you are absolutely shredding the enemy empire in combat and have occupied all of their systems, when the time comes to sit down at the negotiating table you can only ask for whatever you set as the original War Goals – i.e. three planets, because that was all the game would let you ask for.
It would be something of an understatement to say I have some fundamental objections to this system. For starters, while War Goals might make sense in a game about 16th century European politics where everyone is moderately civilized even while they’re trying to kill each other, they make absolutely none in a game that has entirely different species warring against each other. If I’m a race of militaristic xenophobe arachnoids and I want to enslave the neighbouring race of filthy peace-loving mushrooms because I am literally a slavering monster from outer space, I’m entirely out of luck. The interstellar community is far too genteel for something that should by all rights be fantastically diverse and contain some out-and-out villain races. Where are the religious wars of extermination? The wars where the plucky underdog bands together in an alliance with other races to break apart the evil empire? For a game that purports to be all things to all people Stellaris is fantastically restrictive in what it allows you to achieve in a war, which is a shitty payoff when you take into account what a pain in the ass actually fighting the damn thing is.
Speaking of alliances, these are one of the elements of Stellaris that are currently totally broken. When you’re in a bog-standard alliance with somebody you’ll always be dragged into a defensive war – as it should be – and you get a vote on whether to fight an offensive war, with the war only occurring if the majority of the alliance members vote yes. Which is fine. Unfortunately once war is actually declared you’ll start noticing some… odd behaviour on the part of the allied fleets; namely that they will drive every single one of their fleets over to sit on top of your biggest fleet and follow it around for the duration of hostilities. There is no alliance warfighting AI to speak of, just this puppydog behaviour where it blobs together everything with your big fleet, even if your fleet is safely docked at your homeworld on the other side of the galaxy well away from the fighting. This means your allies basically rely on you to fight the war for them, which is more than a little bit galling when they started it and they dictated War Goals to liberate three planets; it’s you that’ll have to actually go over there and take them, and you don’t even get anything out of it. There’s an advanced form of alliance in Stellaris called Federations, but these are a hundred times worse than the bog-standard alliances because you no longer get to vote on whether you get dragged into a foreign war. The Federation presidency will rotate between all the Federation members, and whoever currently holds the presidency can declare offensive wars on behalf of the whole Federation. What this means in practice is that you’re totally at the mercy of the idiotic diplo-AI, which will declare wars out of your control and then expect you to fight them for it.
I haven’t even gotten to the worst part about Federations yet, though. Since only the “primary” attacker (i.e. whoever the Federation president is) or defender in a war can actually negotiate for a peace, and since the AI will absolutely not sue for peace unless it can achieve 100% of its War Goals, this can lead to some game-breaking bugs. In my game as the humans the Federation president declared war on a nearby Confederacy, citing the liberation of four planets as its War Goals. Before I could even get my fleets over there the Confederacy ceded two of those four planets to another, much more powerful race in a separate war. This rendered the War Goals unachievable, but the AI didn’t understand that and, even when I’d occupied every other planet the Confederacy owned, was still stubbornly holding out for those two planets that were now happily part of a completely different empire. Even leaving the Federation didn’t help, as I was still locked in a never-ending state of war that only the Federation president could resolve; I wasn’t allowed to negotiate a separate peace. Eventually I resorted to hacking my save file so that I could play as the president race, making peace, and then hacking it again so that I could play as the humans and leave the Federation ASAP, as it’s nothing but a diplomatic deathtrap.
(Another comedy warfighting bug I should mention: Fighters and Bombers in Stellaris have a very short weapons range stat. I assumed this meant that the Fighters would fly out from their carrier and engage enemy ships at point-blank range. What it actually means is that that’s the range into which you have to get your carrier before the fighters will even attack; you basically have to ram it down their throats in order for it to do any damage. I built four expensive carriers before I realised how broken and useless they were, which was surprising considering there are whole battleship sections built around supporting fighters and battlefleet systems built around defending against fighters; they’re a major lategame weapons system and they don’t fucking work.)
Basically, wars are fucked right now. Since the latter 50% of Stellaris is entirely focused around fighting wars, that means Stellaris is fucked right now unless you avoid fighting wars entirely and wait for one of the big galactic catastrophes that have apparently been coded into the game, like an AI revolt or an invasion by an extra-dimensional race. I say “apparently” because despite running the game for an additional fifty years past the point where I got bored I never saw one of these catastrophes; I suspect this is partly because some of them are reliant on somebody researching one of the red “risky” techs like Jump Drives or AI Sentience, and thanks to the tech RNG I was never offered these technologies. As a consequence the last 5-6 hours of Stellaris consisted entirely of me running the game on the fastest speed setting and browsing the internet on the other monitor while waiting for my techs to research. The only other popups that occurred during this time were the ones about elections (these are useless for reasons that we’ll go into in a second) and an occasional plea from the Federation I’d abandoned to rejoin, since they presumably needed me to fight more wars of aggression for them. The fact that I could do this for several hours is an awful indictment of just how non-existent Stellaris’ endgame really is.
“Still,” you might think, “at least Stellaris is 50% of a good game, right?” Well, not quite. The opening hours of a campaign are different and interesting, true, but Stellaris also happens to be saddled with an absolutely atrocious UI, with basic features I took for granted 20 years ago completely missing. You get a lot of pop-ups in the early game, but they’re not logged anywhere; if (for example) you discover a pre-sentient race and want to go back and see how they’re getting on a few years later, there’s nowhere where you can go back to look for the original event to give you some clue as to where to find them. Instead you have to bring up the Species pane of the Contacts screen, filter on pre-sentient races and then click through them one by one until you find the one living in your territory. There is a thing called a “Situation Log” but this is just a quest log, really. If you then take the option to Uplift said pre-sentient species this will generate an anomaly marker on top of the species’ planet; since you have spent the entire game sending science ships to research these markers you’ll naturally send one to look at this one too on the assumption that this is how you Uplift the species. Spoiler, though: it isn’t. The science ship will just bug out and repeatedly fire its survey beam at the planet until the Uplift “event” times out. It turns out that selecting the Uplift option also generates an associated quest in the Situation Log, and it’s here that you have to do the research in order to successfully Uplift somebody. It took me three separate goes before I figured this out, it was that counter-intuitive.
The list of UI woes goes on and on. Elections periodically occur which elevate one of your existing leaders to President, but the effect this has in-game is to snatch away one of your scientists from their research without telling you there’s an empty spot; every time there was an election I’d frantically click through my research and all of my fleets to try and figure out who had been elected and fill their vacancy. Leaders are elected with a mandate (usually something like “Build X research stations”) but you’re not told what this is in the election popup, you have to go into the Situation Log to find it. You spend most of your time on the galaxy view, which shows systems as individual stars, but if you want to select a planet within a system you have to click on the star to go to system view and then find it within the system. There’s no right-click option to select it directly from the galaxy view. There’s a welcome Outliner pane on the right that provides shortcuts to planets you directly control, but if you have a large empire probably 4/5ths of your colonies will be controlled by AI governors and you’ll have a bugger of a time finding anything that falls within one of their sectors unless you go through the Empire tab. There’s no central repository of known colonisable planets; this is something that Master of Orion 2 absolutely nailed, but it’s totally absent from Stellaris. Instead, if you want to colonise a planet, you have to:
- Locate a system on the galaxy map that has a green globe next to it. This will tell you that there is a habitable planet in the system, and if you mouseover it it’ll even tell you the type! Unfortunately what it won’t do is tell you which of the races in your empire it’s habitable for.
- Click on the system to go to system view and find the planet; this is not always as easy as it sounds if the “planet” in question is actually a moon. If you mouseover the little green globe symbol here it’ll finally tell you which race it’s optimally habitable for.
- If you happen to have a colony ship lying around with said race on board, then there is a colonise shortcut on the planet screen that’ll let you select it from a list and order it to colonise.
- If you don’t have an appropriately-populated colony ship you’ll have to build one. This entails finding a planet in your empire upon which that race resides and then giving it a build order – thanks to the planet-selection problem above, this is a similarly long-winded process.
- Finally, colonise the planet.
It’s just so much more convoluted than MOO 2’s process, which if you had a colony ship built would let you give an order to colonise a planet in two mouse clicks – one to go to the colonisable worlds screen, and one to dispatch the ship. These are solved problems, and they have been solved for literal decades. There is absolutely no excuse for Stellaris to get it so wrong, especially coming from an experienced developer like Paradox.
So Stellaris is a bit of a mess right now. Thankfully since this is Paradox we’re talking about I’m pretty certain it’s not going to remain a mess. They have a well-deserved reputation for constant patching of their games, and they’ve already been confirmed to be working on certain missing features I’ve bitched about above like the colonisable worlds screen. In a year’s time it’ll likely be as good a space 4X game as there’s ever been; there is a lot of potential here, a lot of imagination and a lot of new ideas that I’ve just not seen before in a strategy game. Unfortunately the execution doesn’t even come close to matching up with the ambition — this initial release of Stellaris is sloppy, incomplete and doesn’t do those ideas the justice they deserve. On top of that there’s way too many bugs; I’ve heard it said this is par for the course for a Paradox title but that doesn’t make it at all acceptable, especially for a game that by all accounts they were hoping was going to break out into the mainstream a little more. It needed another three months of development to squash the bugs and fix the UI issues, and probably another three after that to flesh out the endgame. If you haven’t bought it yet I’d recommend waiting at least that long until picking it up. Otherwise you’ll just end up as disappointed as I was.