Your opinion on how to calculate an efficient expansion path

[Impatient]

Hi everyone!

I would like to hear your opinion on how to calculate an efficient expansion path (EEP). I am interested to read on what basis you decide what and when to build what and what and when to research what.

To distinguish, this thread is NOT about how an EEP in your opinion looks like, or what your EP is. It is about the parameters, which in your opinion, are important to take into account when planing the EP for your game.

I plan to summarize all ideas coming up in this thread, in this first post here.

I will give you my first thoughts on the topic in a first reply in this thread.

[Impatient]

In my opinion the best way to plan an EEP is to look at the development of the base backward. Meaning, I decide where I want to get at, and then I plan what I need to build and research from the end to the start.

Example

aims could be:
- be able to produce components for rocket, and rocket itself in one hour
- have science production capacity of 1000 science/h each
- have science consumption capacity of 1000 science/h each
...

IMO, important parameters are restrictions or conditions while on the EP. Like going for steam achievements (no robots, no lazors, etc. ...). Or something different, like having double the amount of resources necessary at all time, to be able to play around with.

Also the game parameters have to be taken into account. Number of enemies, availability of resources, ... .

[Impatient]

Another thing I have not yet wrapped my mind around, is the question when to increase production capacities.

Eg ...
Everyone would agree, that when launching the rocket, the production capacity of the base is x-times higher, than when starting research on red science only technologies. Naturally, you would say, it would not make sense, to build up iron plate production capacity to a level that can support the production of the rocket components at that in-game progress. Agreed. The question though is, WHY not? What are the implicit criteria we apply to make that decision?

One thing that comes up immediately is the progress in technology. IMO the availability of a new technology is a point in the game when it looks feasible to think about updating production lines and maybe also extending production capacities. Another criterion we implicitly may apply, are the necessities to accomplish the immediate next steps. Why should I build up iron plate production to rocket-state levels, if 10 furnaces are sufficient to get red science research and yellow belt production up and running?

But this does not completely satisfy me. Apart from the possibility that there are some more, maybe minor criteria we implicitly apply, when deciding how much to increase production capacity, I also have the feeling, that there are reasons which justify the extension of production capacities far beyond the current necessities. And what might these be? Is the topographical base layout one, or isn't it? Is it the tediousnes of upgrading the production by one furnace at a time?

[DaveMcW]

There is one popular parameter, time to produce your first rocket.

Also called the speed run.

Watch some speed runs and you will find answers to your other questions.

[FrodoOf9Fingers]

I concur with the end-objective perspective.

However, I personally have two modes: goal oriented and expansionist. Goal oriented is already described (create just enough to satisfy target demand). However, in the expansionist mode, I work nearly the opposite direction. For instance, I'm currently building a train fed system that converts ore to plates. The system is a line of ~14 (IIRC) train stops, where a train will stop at each stop for 1 second (the idea is to have the train empty by the last stop). I then make it so that I have enough furnaces to process that amount of material. Then I expand my mining to feed the system, and then finally I calculate how many plates I'm making of each type (my capacity), and build out the rest of the science system such that it is under the furnace capacity.

TL:DR Sometimes I like to optimize one specific aspect to some maximum potential, and then expand the rest of my system to match the maximum output.

[Frightning]

Hi everyone!

I would like to hear your opinion on how to calculate an efficient expansion path (EEP). I am interested to read on what basis you decide what and when to build what and what and when to research what.

To distinguish, this thread is NOT about how an EEP in your opinion looks like, or what your EP is. It is about the parameters, which in your opinion, are important to take into account when planing the EP for your game.

I plan to summarize all ideas coming up in this thread, in this first post here.

I will give you my first thoughts on the topic in a first reply in this thread.

I think it's very important to look at how quickly and cheaply a setup can be built versus its rate return once built. This lets you determine what sorts of setups to go for at various stages of the game, for instance, at the start, the best thing to do for plate production is simply plop down a Stone furnace and manually feed it materials, but this is prohibitive beyond the early game as you need to handle too many plates/min and the micro management would be untenable. Likewise, a simple furnace line can take you pretty far, but eventually one belt of ore is just not enough, and that's when things like train-fed smeltery blocks make more sense to build, and using logistics robots instead of belts can be worthwhile, etc. Understanding how much you can expand a setup before it will run into scaling issues is very helpful knowledge, but not so easy to learn for an inexperienced player.

[featherwinglove]

I tend to play heavily modded and often relatively blind in terms of primary resource access (i.e. Bob's w/o Angel's Ores <--- mod reduces Bob's 15 primary ores to 5 or 6, not sure about lithium, thus solving the problem ---> low frequency in map gen options putting the patches far apart.) This means that expanding my base involves reaching and processing a new primary resource, or fetching more of an old primary after secured patches run down or out. This results in different technological priorities depending on whether I need to automate something (which would result in researching an automation or electronics tech, sometimes a primary processing tech if I need, say invar or titanium for the next tier), or expand (which would result in researching weapons tech.) With primary access and a lack of map knowledge being an issue, efficient expansion often involves a literal path which must be first discovered with radar and then charted through the crawling horror to its destination through the lowest number of spawners while still being rail network compatible, and finally conquered, uncovered, and exploited. The result is very different from the speed run approach, which generally relies on having not only recipes and ratios memorized, but the map and build as well. When you don't know where the resources are and don't have a rehearsed build in mind, the recipes and ratios are still useful to keep close at hand, along with some basic blueprints to help implement them more quickly and easily, but memorising the map and overall build is quite impossible. (In the Muddy Mountains game, I thought my battery plant was going to be just east of home plate and thus I started a sulfur warehouse there. It actually wound up four kilometres to the west simply because that's where the oil was. It was easier to move everything solid, including the sulfur, across the map, and most of the sulfur would come from the oil anyway. The lead is four kilometres east of home plate, as far as you can get away from western outpost while staying within the defense perimeter, but it does take the battery system a rather long time to use up an entire train load of lead.)

[MisterSpock]

In my opinion the main parameter to messure efficiency is time needed to reach a certain goal.
This can be your first rocket or a substainable base. This is acutally very close to the idea of a speedrun.

Sure there are ressources in this game and you automatice stuff. But like reallity its the human which makes stuff working, not the machine.

The second parameter are enviromental or ressources relatet parameter. For example you only have limited uranium, which means your limited by this.

Factorio is a game which doesnt develop when the player is idleing.

Addendum:
You can actually tranfer this mentality of "everything is worktime" into real economy. Building is spaceship (for example the Spaceshuttle) is mainly time consuming, not ressource consuming. With this, the answer to the question "How much cost a Spaceshuttle?" is a number of working hours. Thats the time you need to programm and plan stuff, assemble it and gather the ressources.