Stellaris grand strategy space game by Paradox discussy thingy thready thingy

Also tried the free weekend to see if it got any better since the previous one, and I don’t… hate it? It still has many, or all, of the issues it had before. But the ride is smooth enough that I want to see what happens in another decade of in-game time, or in the next game with a different set-up. It’s satisfying even though I could write a good few paragraphs about it’s problems and much less about what does work. And the AI works as a challenge, though it doesn’t do much diplomatically.

(I did write those paragraphs before realizing most folks here have been playing for years and don’t need me to explain the game).

It has always suffered from one key issue, the end game. By the mid game process, most games you know whether you have won or not. The end game crisis adds a bit of spice, but once you know how to deal with each crisis, that also looses it’s appeal.

I have played lots of games and quit almost 99% of them at the mid-late game point because either you know it’s a cake-walk or an impossible mission.

The first half of the game is great, the latter half not so great. Unfortunately, once you’ve played the first half of the game numerous times, you’ve seen all the scenarios, and events so many times that they loose their appeal too.

It’s a “good” game (that I keep going back to when a new version comes out) but adding continual DLC is still putting a plaster over a significant wound and doesn’t help. I think the only way forward is to work on a sequel and resolve the root problem with the AI/gameplay.

I’ve got ~200 hours in, but have mostly not played games to the official end year. But one thing I did notice in my last game where I stubbornly played to the very end is that my empire which was generally in pretty good shape for most of the game basically went to hell in the final years. I don’t remember all the details, but I had to reverse everything I had been doing earlier in the game. Most notably, I had to beg everyone to please, please stop having babies.

I admit it’s possible I just had no idea what I was doing. But even with ring worlds, I was running out of space for everyone.

hm, this does sound like a positive. That’s how chess works, too. Usually at some point after midgame, you know if you have lost or won. Most chessgames are not played to the end. (trying to see the positive here).

That’s one reason why chess isn’t considered well designed in modern gameplay terms. Chess’s greatness comes from the massive possibility space, and the fact that it’s completely unintuitive to our brains - a bundle of pieces from both sides sits huddled close together and eventually resolves to one side winning. There’s no intuitive sense of armies and supply lines and it has very little do with war in general. It’s more of a two-player abstract competitive puzzle game.

I read your intent, and I know what you mean. However, chess is a strategy game (long term) with a highly tactical component (short term). There is some “puzzle” aspect, if you want to find the “best” move in a position. But in real, the best move is the one you chose that fits your strategy or tactical needs. You can outplay an weaker opponent just on strategy alone.

It could be an issue in Stellaris where you play for many, many hours in several sittings and than have to kill the game because you know you lost or you will win easy.

I think this is true for me in most 4X or grand strategy games I can think of. I usually have a good idea that I’ll “win” in EU4 by the 1600s and rarely play past 1700. Same with Civilization and it’s descendents, I rarely if ever play the games through to the end as it’s a foregone conclusion.

I know that’s what the end game crises were attempting to address but I don’t think it’s that successful. As you say, once you figure out how to handle each you’ll know if you are going to win or not before it happens.

Why is this a problem?

I’d give two answers:

  1. Because many players don’t like to arbitrarily decide they have won (or lost) and so keep playing even when it’s stopped being much fun.
  2. Because if you do stop playing you miss out on the end-game content. Charting the course of a civilization is a bit disappointing if you never get past knights.

For chess, neither of these is really a concern. Both humans and AI will resign the game when they can no longer see a path to victory. And there are enough games that go the distance that you should get to experience the end-game pretty frequently.

As far as 4X games go, Master of Orion (the original) does perhaps the best job of avoiding case 1. It has a good idea of when a player has won the game and will indicate this with a victory screen (though of course you can continue if you wish). It’s still rare to get to play with the end-game toys, though.

I wish more 4Xs did better with “late starts”. I remember being sad that I never got to play with the American UU in Civ4 (the generally underwhelming Navy SEAL) and specifically started a game in the industrial (or something) era on an archipelago. While there were definitely some flaws (missing all the early game buildings is weird, to say the least), it was a blast.

If there was one thing that would completely change Stellaris, it would be the AI from SOTS. It adapted to your ship layouts and if you won a battle heavy with projectile weaponry, they would come loaded with armor for the next battle. It would also add variety to the end game crisis (which are extremely predictable once you know each ones weakness) and give you a reason to actual continue playing.

I don’t play enough to notice these things myself, but it would first need a balance pass and some mechanics tweaks for alternate layouts to be viable. People seem to like them mods to change it up, though; if nothing else, the cpu-melting insane crisis from Gigastructural Engineering seem to be a thing.


I think having an end-game start in Stellaris with a randomised fully colonised galaxy, existing alliances/fleets/wars/whatever - would be really fun as a sort of puzzle game. Much shorter than normal games, but endless variety and many different challenges.

I fully agree, a starting point just like in EU or CK or Victoria or all their other games. That’s my fundamental problem with Stellaris. Asymmetric starts are just way more interesting than the usual 4x where everyone starts with one city/planet. Also makes the game much more of a challenge without needing to resort to giving AI cheats.

While starting with one planet seemingly fits the theme of the game, as it is like Earth starting to explore the stars, it makes no sense for all intelligent life in the universe to start exploring space at the same moment. Fallen Empires goes a little way to alleviating this, but it would just be better if you start the game with an established universe and you learn more about that universe as you explore.

You could pump “Advanced AI Starts” up to maximum so you’re the only newcomer – I can’t remember if you can do that for every AI in vanilla but it would be trivial to mod. That’ll give you what you’re looking for.

For me, no thank you – the “Explore, Expand, Exploit” parts of these games are my favorite and as soon as the map is full I’ll probably just quit.

Well not really as then every AI will be roughly the same size and I will be one planet (or has that setting changed from when I last used it?). EU4 has a big variety of large empires, medium sized empires, small empires with alot of potential and inner strength, and small empires that don’t have alot of potential.

Stellaris is not balanced or designed for this type of setup either, I am just daydreaming about a diferent Space Grand Strategy Game.

As an aside, I am reading the Three Body Problem trilogy by liu cixin and it is amazing. Does anyone know if over the years whether Stellaris has added anything that resembles Dark Forest Deterrence or anything like what you read about in these novels?


I don’t think Dark Forest Theory makes sense in Stellaris because

  1. Peace between civilizations is possible.
  2. You can’t really make the kind of fast technological leaps that make other civilizations so threatening in Three Body Problem. If you’re way ahead of somebody in tech, you’re not going to all of a sudden find yourself so behind them that they could wipe you out in an instant. You’re going to stay ahead, or worst case, they could gradually catch up.
  3. Tech isn’t so important relative to production and economy. There’s nothing possible like in Book 1 where Earth had dozens of ships and the Trisolarans had one, and they destroyed the Earth fleet in ten minutes. Even if someone is way ahead in tech, they would need to be able to pump out a whole lot of energy and minerals to be a threat.

Which all makes sense – Three Body Problem wouldn’t make a great game, I don’t think – destroying every civilization you find until one finds you and then they shoot you with their superweapon and the game is instantly over. Also, you’d need to roll up a galaxy with thousands, if not tens of thousands of stars for the game to make any sense – everyone needs to be hard to find!

Or maybe it would be fun, I don’t know. For better or worse, Stellaris is mostly relitigating Star Trek.

I suspect you’re right, but it would be really interesting to see a game tackle this type of conception of the universe! Until then, a Stellaris mod that tries to change the three points you outlined would itself be interesting.

Well perhaps we wouldn’t call it ‘peace’, but I would have thought coexistence would be possible between civilizations assuming each had the technology necessary to transmit coordinates? Like the US and the Soviet Union. If they transmitted your coordinates you could probably detect the transmission before anything happened, which means you would transmit their coordinates (and they would know that).

I think it could have more nuance than that. If you were a space-faring civilization operating under a Dark Forest attitude, you would want to colonize other planets and star systems ASAP to make yourself less vulnerable to one strike. If you meet other civilizations in space you will be doing your utmost to prevent them from learning where you live, while they will be doing their best to prevent you from learning where they live. I don’t know, I just came from reading the second book and find it so fascinating (and dissimilar to Star Trek / Wars / etc.) that I am curious if it could be made into good gameplay.

This really should have been the case in Stellaris. It would have solved EU and CK’s problems too, which are mainly that one simply cannot simulate the dynamics of real-world history to any serious degree. With a fictional universe, none of that matters.