Hacker News Reader

Comment by the_af | Hacker News Reader

See my sibling comment about Stellaris. Please elaborate!

These days I would enjoy a 4X empire building game that focused on the absolutely highest level issues I imagine a ruler of a galactic empire would focus on. That is, managing the empire and maybe space fleets at the highest level, not issuing orders to individual ships or planets. Maybe dropping down to a particular theater of conflict or diplomatic crisis if the situation calls for it. And at the same time, making it enjoyable, of course!

Nothing kills my enjoyment faster than old-school micromanagement. R&D, ship building and handling, resource gathering -- my serfs should be doing that, not me, the emperor. I'm no longer a teenager with lots of spare time :)

View on HN

Stellaris doesn't have much tech or fleet micro. However, there is a _massive_ amount of planet (or more specifically, population) micro. You have to manually build up each planet and move the extra pops that grow to wherever they are most needed. Once you get to the late game and you have 20-30+ planets the game really becomes a slog to micro and game performance drops a ton. I still love Stellaris despite these flaws, but they are pretty big flaws.

Thanks! Pretty big flaws indeed. Not the game for me then.

Civ 5 (or was it Civ 4?) attempted this by giving you city governors that you could give "general directions" for what to focus on, and then ignore. The workers were also automatable in this way. It works pretty well for casual play, but playing higher difficulties still requires micromanagement. Since you can micromanage, you can zoom in on the nitty gritty as desired.

For a certain kind of self-actualized casual gaming, it might actually work. You have to trust / live with the bad choices of your automatons, yet also remember to tell them when you want to change direction.

The main limitation to the AI governor model, which has been tried in many games to varying success, is that the solver necessary to make them play like an optimal player tends to start encompassing more and more parts of the game mechanics as you add in tricks like "use a worker to rush production" or "plan improvements to coincide with tech advances". The player, likely to be of a controlling or anxious mindset, will reject the idea that these details can be left up to an unaware AI, and given the option to do so they will micro-manage the fun out of the game.

When the game has relatively few options that can be played to a high depth, like Chess, of course, it's easy to write such a competent AI with a brute-force approach. But that's usually not how 4X games are structured: there are a ton of decisions involving overlapping timers and resource bonuses and force committments, and so the top-end strategy revolves around leveraging everything in a just-so way to accelerate growth and minimize losses.

The move towards lower micro and symmetric balancing in newer games has the upside of downplaying much of this; your advantage, where you get one, is much less likely to come from careful micro employed to replicate a 50% advantage in a single scenario hundreds of times, and AI benefits as well. But it also loses some of the charm of weirdly unbalanced 90's strategy.

You summed up more perfectly than I could why the AI governors are really only suitable for a certain kind of casual (relaxed) style of play.

>See my sibling comment about Stellaris. Please elaborate!

Basically, it's a game emphasizing and focusing on roleplay. The early game is absolutely brilliant. But it has way too strict rules and there aren't many meaningful (combat or diplomatic) interactions with the AI past mid-game once you've exhausted most events * .

The war AI is a joke once you've grown enough, and diplomacy offers very few options - maybe the latest DLC will help with that.

So you are left with either slowly conquering the galaxy (because the wargoal system is stupid, I'll get to that), or basically doing very little - diplo is only slightly more fleshed out than MoO, Federations only make sense for some Empires and don't do much yet - and waiting until the Crisis, which is bugged enough that it's no longer that much of a threat.

There are also the drastic performance issues some players have mid-late game, or how the economy revamp (a good idea in theory, tiles needed to go) didn't reduce micromanagement, but arguably increased it. Again, early game Stellaris is awesome, but later... not so much.

---

As for the wargoal system, unless you're playing a genocidal (FP/DE) race which almost removes diplomacy, you're limited to wars with very limited goals with 10-year ceasefires * . There's no way to break those ceasefires, despite it not making any roleplay sense.

By comparison, in MoO sometimes an AI race will attack one of your planets without declaring war, and sometimes I just have to take in the chin, because I'm nowhere near strong enough to fight. At which point I might begin methodically planning revenge while being nice-nice diplomatically.

This is actually something that can also happen the other way around, with the player being the aggressor. AFAIK, there are no fixed rules in the MoO code for that - this is just an emergent property.

Stellaris, despite the large number of fixed events and mechanics can't support that kind of interaction, because the mechanics are implemented way too strictly, and you can't just send out ships without formally declaring war with fixed pre-set goals limited by Influence and if you did have a war, you'd always have the large diplomatic malus with the faction and no good way to try to appease the AI.

* Events are basically fixed storylines choices written by Paradox, where every choice makes a small difference. Compare GalCiv which also had that mechanic, but GalCiv's choices are far less detailed and GalCiv had a moral scale mechanic which wouldn't have been a good fit for Stellaris

* Well, there's the Vasselize-Annex method, which is almost as slow because it costs a lot of Influence.

This is a great summary of Stellaris, but I have one tiny nit. If you get the Apocalypse DLC you can build a Colossus which gives you a total war war goal that lets you immediately annex conquered systems without spending influence or having to wait for the war to end to gain control of the systems.

I didn't know that. I knew that Cololsi gave wargoals but didn't connect the dots that they are total war goals. Never bothered buying Apocalypse, because it didn't seem to offer that much I cared about. Perhaps that was a mistake on my part.

I should have also mentioned that Stellaris has some DLCs which are de facto mandatory, which is just the way Paradox does things. Utopia is all but officially required. Apoc I guess I should have bought. Leviathans is IMHO very useful too. Not sure if anything else can be called mandatory?

I actually think that Apocalypse is the weakest of the main DLCs. Other than the Cololsi there aren't really any big features in it. Synthetic Dawn is mandatory if you want to play as a machine empire (which are THE meta right now.)

TBH other than Utopia all the "big" Stellaris DLCs have felt rather lackluster to me. I'd much rather have Distant Stars or Ancient Relics or even the Lithoids species pack than Apocalypse or Megacorp. Of course I've bought them all except for the Humaniods pack anyway on sale because I absolutely love Stellaris and have put a massive amount of time into it. Even with all of its flaws and quirks it and KSP are still my favorite PC games.

The 10 year ceasefires are because Paradox likes to make it impossible to take out an enemy in a single war. Each war can only take a bite of a certain size, then you have to wait a while (during which they can rebuild their military) before you take another bite.

In theory it means a loss isn't catastrophic, but in practice if they can beat you once they can beat you again, your only real hope is to find allies--and when you're weak it's unlikely you'll find allies.

Reply on news.ycombinator.com