Missions with different goals for replay value

[Gandalf]

So jeb_ dropped by on the subreddit and commented on his experience with the game. Part of his impression was what I've read from a lot of other people before: It's an awesome game but once you've reached the end goal you're pretty much done. There is little incentive to start a new game. Of course many people here on the forums are a lot deeper into the rabbit hole, using mods and whatnot to keep things fresh and challenging, but for a lot of players (I'd assume the majority) vanilla gameplay has to be enough.

I thought about ways to improve this and found that, at least for me, two things have the potential to keep the game interesting (nothing new here):

1. New features. I mostly get excited about big new releases in Factorio and that's usually when I consider starting a new map. Since these are rare, so is my motivation.
2. Missions. I'd like to expand on this a little. Obviously the idea of missions isn't new and if I remember correctly there are already a whole bunch of suggestions about expanding them. However I feel like there should be a lot more emphasis on them becoming a key component of the game rather than just a couple of tutorials.

Years ago I played a series of games that were in many respects very similar to Factorio: Caesar III and it's various spin offs (Cleopatra, Zeus, Emperor). Like Factorio they are 2D building games with production chains etc. Although the focus is different, one key parallel is that you have to build a sort-of efficient city from the ground up on every new map. However you actually have to unlock and complete dozens of maps (or rather “levels”) in order to finish the game. This kept me hooked for a lot longer than the freestyle maps. So how did they keep things interesting?

1. Each map offers special challenges. Sometimes you had to work with limited resources, sometimes the city had to fit into very confined space etc.
2. On each map there are multiple predefined goals to reach. These differ from mission to mission. They were usually defined in some key metrics (e.g. citizen satisfaction, money…)
3. Maps in different regions offer different resources that fulfil the same tasks. For example in a colder region you'd have other types of food than in warmer regions. These differences are mostly cosmetic but occasionally result in shorter or longer production lines.
4. On advanced maps you often start a little further ahead on the way of unlocking things. Even though you have to build a completely new city it never feels like you're completely starting from square one.

I believe these concepts could add a lot to Factorio, so to get to the point of my suggestion: Add missions similar to what I described above. In detail that means

1. Design some cool maps. I could see at least one developer doing nothing but creating cool maps. Of course the question is how to deal with finite vs infinite maps. One of the great things about Factorio is having the seed-based infinite worlds. But even though every generated world is different after a while they still all start to look the same. Maybe you could come up with a system where a map has a manual layout at its origin but than continues outwards through adjusted procedural generation? And not just those few parameters the map generator has right now to adjust the difficulty but actual gameplay changing differences.
   An essential part of this will also be that there should really be at least one more type of physical obstacle in the game: Mountains and/or plateaus.
2. Create more than a single end goal. Different maps should have different types of “now I'm done” reasons. The rocket silo is a great way to prototype this: Individual levels could require the construction (or supply) of individual special buildings. On a limited-size map vehicles could enter and leave to carry away resources to somewhere outside the player's reach (and of course the upcoming space station :3).
3. While there are a few resources in Factorio that are almost completely useless (wood, heavy oil) I noticed there is not a single redundant one. Some mods add this concept (e.g. Bob's mods) where you can use different ingredients to craft the same results. Adding this on a larger scale could keep things interesting.
   Now to clarify this, you don't want to have all the different potential ingredients on every map. Instead on different maps you'd be forced to use different resources and production chains to get the same outcome. That way even after many levels the player will still have to learn a small set of new receipts. This can also be used to enforce challenges. For example on some maps creating long transport belts could be so expensive (requires tons of lubricant because it's cold or whatever) that using trains is the only way to cover the long distance to resource X.
4. Pretty obvious: Once you've done the basic research a couple of times it gets boring. There are already mods out there giving you the personal roboport and some basic science from the start. Do that for advanced missions/level/maps (dunno what to call them anymore). Throw in a stack of solar panels for the really advanced ones. The challenge should not consist of having to mine coal with an iron tool over and over again.

So yeah, these are a couple of thoughts I had. Dunno if there's anything new in here for the devs but I'm hoping it's some food for thought. I'm well aware that the things I've described would be damn ambitious.
Also if you have never played any of the Casear III & co games I dearly recommend them (though I have no idea if they still run on today's Windows versions).