Factorio Forums

The most recent Friday Factorio Facts brought up a good way of describing game enjoyment (EEI). While that post was mainly about making sure players have a good EEI in the first few hours, it is still a valid concept even at hour 50 (or 200, or whatever crazy number people are at).

One of the issues that hurts my late game enjoyment is running out of research. But even before you research everything you get to a point where you have researched everything that you plan on using (Combat robot follower count, and shotgun upgrades are always the last ones for me). At that point you keep your factory running and you do make some progress but it isn't very satisfying. It feels like just checking off boxes on a giant checklist.

Here are two goals that I believe would make late game research much more enjoyable.

• Enough research options that you feel like you are making a real choice
• Reward specialization so that you have an incentive to make this factory different than the last one

One key item for both of these is that the tech tree needs to be deep enough and expensive enough that it isn't practical to get everything (of course people will just make even grander factories to get everything, but that is a good thing to encourage). Currently it doesn't feel like you are making any real choices since you will unlock everything fairly quickly anyway (except robot follower count, which does have the sort of depth that I am looking for). And if you are going to easily unlock everything then you don't really have to make any choice (solar vs coal, bullets vs lasers) since at any moment you can switch to the other without any reward/penalty.

One easy way to achieve this is to have incremental upgrades, which the game does for combat (bullet/laser, speed/damage) and some logistics (robot speed/count/capacity, and inserted stack size). But I think this concept could be applied to far more aspects of the game.

One issue with incremental upgrades is that they don't seem very exciting (robot follower count 15...). To counter this I suggest having a unique bonus every 5/10 levels or at the end of the tree.

A few examples:

• Radar - Each level: +25% faster scanning speed/increase max scanning distance by 1 - Every 5 levels: Extend continuous scanning range by 1 (8x8 chunks, then 9x9 chunks)
• Energy distribution - Each level: -2% power consumption for all entities - End of tree bonus: Extend power pole radius by 1
• Solar Power Efficiency - Each level: +5% power generation from solar panels - End of tree bonus: Solar Panel Mk2
• Steam Power Efficiency - Each level: Boilers consume 5% less coal (or coal provides 10% more energy) - End of tree bonus: Steam Engine Mk2
• Inventory - Each level: Increase player inventory by 2 slots - End of tree bonus: Double item stack size
• Mining Drills - Each level: +5% mining speed - End of tree bonus: Increase mining area by 1 (5x5 -> 6x6)
• Logistics - Each level: Increase max underground belt/pipe distance by 1
• Oil - Each level: +5% speed for pumpjacks - End of tree: Improved cracking recipe
• Tank - Each level: +1% movement speed, End of tree: Tank Mk2 or perhaps just boost the stats of the current tank (armor/health)

I think nearly every item/concept could have some sort of 10-20 level deep specialization tree. The example bonuses that I provided probably aren't the best and hopefully the community can come up with better ones. But in the end I really am not too concerned about what the actual bonuses are, I mainly just want to make sure that I always have useful research options so that I feel my factory is always getting better (and I keep getting that little boost to my EEI).

Although having a few end of tree bonuses that alter core items (like increasing power pole radius, and perhaps underground belt distance) would provide a couple of interesting benefits.

• It would encourage you to keep discovering new layouts and new designs throughout the game. Like the change from small power pole layouts to medium power pole layouts.
• If these techs are really deep enough and expensive enough it would be unlikely that you unlock all of them in every play through so you wouldn't just have the same set of layouts that you use every time.

I'm thinking that when you've researched everything else, a unique research tree pops up which isn't predictable of what it does.

Instead, when researched, you get a random bonus to something.
Like, suddenly your rockets are 30% harder hitting, or your solar panels produce 30% more power, or inserters consuming 50% less power, etc.
These bonuses should have diminishing returns if randomly hit again during later researches.

Fairly powerful bonuses that are expensive to research, and you never know what you're going to get! All those 5000 research packs (or whatever number) and the present you're opening on completion might be awesome, or it might be "meh". It's the combination of the "Future Tech" concept from Civilization with the random loot system of MMOs.
Since your factory is more or less complete at this point, and you could've completed that rocket to space ages ago, you don't really have to care about the bonuses unbalancing the game. You're already the master of the game, now comes the time to spice things up with some RNG to extend that EEI curve.

I like the idea lancar but you should be able to research out of it like if buildings use 20% more power you need to do 4 researches of 5% power reduction before it gives a bonus

I agree with (most) of this, and would love to see something like this in the next update.
On another hand, how about a research that can never be completed? That will allow for you to keep your science factory running for far after you finish all of the research.

With regards to specialist research...

It would be nice if 'this-OR-that' choices are present throughout the technology tree, meaning that you have to specialize throughout.
It would additionally be nice of the choices you didn't pick then move to the end of the tech tree so that you can still get them later on, but at a much increased cost.

I feel that these two design choices would help to not only add variety, but extend the late game more by simply allowing other avenues to be pursued (though with far more difficulty).

Truly insane specialty trees can then be unlocked when two opposing specialty trees are entirely unlocked (meaning you had to max a tree, and then max the choices you hadn't initially chosen in that tree at the increased costs).

Smarty wrote:I like the idea lancar but you should be able to research out of it like if buildings use 20% more power you need to do 4 researches of 5% power reduction before it gives a bonus

Well, regardless of the levels I think it's important to keep it random. Otherwise you lose the excitement factor that you most likely are starving for at that point in the game. Remember that your factory is likely huge at this point, with several outposts. Enemies are a non-issue with your established defenses, and all you're doing is just magnifying your build for the sake of it.

Randomness, I believe, would help to keep you invested at that point. Granted some people can build on the same map almost forever, but I doubt most are like that. They want some sort of easier emotional return for their investment, and random research "loot" would help with that.

The problem I see with random research options in a game like Factorio is that the game isn't built around re-playability as a core feature. NOW don't get me wrong I'm not saying you can't replay nor that replaying holds no value; but rather that the game allows itself to expand toward larger and larger factories and in a sense rewards the player for playing in a single game more so than it rewards them for restarting over and over. Plus with no actual "end" game as such the game is pretty much open ended.

As a result random research would harm the game; players with huge factories would feel disadvantaged if the research they hoped would be there once they get to nearer the end of the research tree, just isn't there.

Random research in games I can only thing of Sword of the Stars doing this and for that game it was mostly about weapons and some upgrades that were random each time. Which was no bad thing as if one set of weapons were not present all the way to their top level there were other suitable options alongside; and that's without trading with other factions for missing technologies.

In factorio there is no one to trade with; even in multiplayer.


I do think that research should continue, but things like +% increase in performance is ok but really at the end of the day its a very bland upgrade. Nothing much really happens and in all truth half the time you won't notice for a long while. It's nice, but ask yourself - do you research those +damage upgrades on turrets or research a new building/object to use. Chances are most of us research the latter until such point as the latter is so expensive its too "big" for our factory to do quickly; or where it requires research bottles we don't yet have production setup for.


Adding more buildings and more options is one way to keep research going. However its always going to be a finite result. That said more research options would help; sometimes I find that by mid-point of the research tree a factory could do with a rebuild to be more efficient; but at that point the player is looking at a big time investment for no actual net gain bar efficiency which is only going to mean that the "fun" ends all the quicker - because generally once you've done all the research there isn't all that more to do.


Indeed I'd say its a weakness of Factorio that research is the primary gameplay. It's a GOOD thing early on; but mid to late game something new needs to come into the mix to balance up the fact that the research is going to end or become a bland case of +% improvements on things. The new space-era elements should help alleviate some of this problem; especially if its expanded upon. However I think that we need something that also requires research bottle production long-term. Land terraforming is one option that won't end (within practical terms peoples computers will grind the game to a halt or the map will go on forever before the player has "terraformed" all of the map). There might also be other ideas tied into the space era.

Really neat would be using research to research asteriods/space objects which could unlock new resources and structures in-space etc...

Overread wrote:The problem I see with random research options in a game like Factorio is that the game isn't built around re-playability as a core feature. NOW don't get me wrong I'm not saying you can't replay nor that replaying holds no value; but rather that the game allows itself to expand toward larger and larger factories and in a sense rewards the player for playing in a single game more so than it rewards them for restarting over and over. Plus with no actual "end" game as such the game is pretty much open ended.

As a result random research would harm the game; players with huge factories would feel disadvantaged if the research they hoped would be there once they get to nearer the end of the research tree, just isn't there.

I don't think so. This random endgame research would always come when all other research had been finished. If they add more types of research for anything, that would have to be completed before you get this random research reward.

It is a bonus, and nothing more. Kinda of like New Game+ mode. The research continues, but you cannot control it and the bonuses it gives can either drastically alter your factory throughput, or not. It's never the goal of the game to get it, rather a treat for those who wish to go on regardless when all has already been said and done.

Besides, nothing is infinite (cue Einstein quote) . Do this research enough and you'll eventually get all the bonuses to their max value.
Thinking of it some more, it might be good to have it truly be the new game+ mode. When you've launched the rocket, built your space platform and rescued all the colonists, the random research option unlocks. That would work, right?

Don't forget that sometimes players just want to see their factory run at full tilt. Pounding out research is a major part of sinking those resources for much of the game. A post game research system would be pretty nice to have, if merely to keep the factory running.

To keep my factory running endlessly, I use this mod.

https://forums.factorio.com/forum/vie ... 93&t=15308

It auto launches rockets.