Research Takes more Time and Labs Increases Speed after a While / Acceleration of Research

[ssilk]

TL;DR: Each technology will take more and more time to research. The higher you go with the tech you need much, much more time than yet. In exchange of this extreme nerf the behavior of labs change. If a lab is in continuous research for some time it increases it's basic-speed. It "accelerates". With each acceleration-step the power usage increases, too (of course). Once the lab needs to stop research (no power, no items) this "speed-bonus" will be reset. This will lead to a much more interesting mid- and end-game, without changing the basic core of Factorio.


I coming from this post: posting.php?mode=reply&f=6&t=28900#pr197216 Add a slider to increase research and build times

When I read Kevin's thought about the research speed I thought this is a good idea.

I found that increased lab power made the game much more enjoyable as it forces me to build more power, along with the need for more materials and the need to expand the base.
PROBLEMS: Increasing all science x100 makes starting the game not as fun since you are just waiting to research basic automation. Thus, I think there must be an option to exclude technologies that only use red science paks... these early techs would have no increased cost or time so you can get to the real game sooner.

My idea about this problem is, that the game should start as now, but with each tech you research the following tech takes more and more time.

In exchange to this nerf of technology tree the labs are buffed in a - not equal but - similar way:
If a lab is in continuous research for some time it increases it's speed. It "accelerates". This is limited by the current existing technology Lab efficiency (which still increases also the basic starting speed), but this speedup will be lost, once the lab needs to stop (cause no power or no items or just stopped).

The point is here: With each acceleration the power usage increases, too. So after some minutes (1-5 minutes?) of continuous research the lab speed is as high as "usual" (without this change), but the energy usage is much, much higher than yet.


In short:
- each research increases time-cost of the next research by some factor (could be explained with "the more research we already did, the more complicated it gets to find the next"). The result is an exponential (*) scale of time for the research cost.
- in exchange a continuous resarch-time in lab increases lab-speed exponentially.
- The lower bound of the lab speed is as now. There is no change.
- The upper bound needs to be calculated (some math) to be in a fair balance with the the above research time-factor and increases of course with lab-efficiency tech (as now).
- The bonus is lost when the lab stopped. Due to power shortage, item shortage or no order. This cries for a research-planner!
- The bonus is not lost if the lab stopped only for a very short time. I think bonus gets lost after some seconds (1-10?) of no power/items/order.
- The time without loosing the research-speed-acceleration-bonus could be increased by some research.

(*) I'm not sure, if exponential is right, perhaps we need other curves that are more flat in the end... in every case this curve needs to be balanced of course. So please replace in mind every "exponential" with "some fun-function".


Why is this useful?

You are in the mid-game, and you have managed - after some tries, that your labs have reached maximum research-speed (for now). The needed power surprises you, it's far beyond of what you ever imagined. Now you know, why there is a nuclear power plant in the game and you are proud for your working nuclear supply-chain.
But this is still a little bit too slow, each recipe increases the time a bit, so you research the lab-efficiency (again). Now this is overwhelming: The power need is over 2 gigawatts! Alone for the labs! Some short power fails everywhere! You need a second nuclear plant.
As you are looking around, you see an alarm coming in: a massive wave of biters. Now you need to make a decision: Turn off some of the labs to have more power for the lasers or keep the labs all on in hope that the power loss is short enough to keep the speed bonus.

...

- I think that a power-loss in the mid- to end-game is a very adrenaline kicking (and so enjoyable) part of the game.
- I also think, that exponential need of power is a very cool game element: you cannot longer estimate how far you can go, even experienced players will be surprised how fast things can change.
- Currently, after you reached some point in the game you don't need to care about the right order of research. With this change it is important again.
- There will be dozens of solutions/blueprints to control your research-center in a more or less optimal way, inclusive mods to telecontrol the research progress.
- There will be lab-layouts, that guarantees, that labs keep their research bonus as long as possible, even on item shortage.
- Your research-decision you formerly made changes the time (and power usage) for the following research. Which is a very interesting game element.
- You can balance this like so, that we don't need so much labs than yet, cause there is no sense in having 20 labs. I think 5-10 labs should be enough for anything that comes. If you build more you will come faster into power problems.
- You will care much more about if the labs have enough items, cause you will lose the whole speed-bonus, once a red potions is not available.
- And if you loose the bonus: Don't care - this is self-regulating.

[person3triple0]

I really like this idea, and I have a suggestion for how the power aspect would work:

First, the bounds:
Base_power = 300kW //or whatever the number is, I've been working with too many machines....
Max_power_needed = Base_power*(1 - Evolution_factor)^( (some varying number) )
//The power requirement would be based on the evolution factor: the farther you progress, the more power you might need to do research.
Next, the loop that'll check if science has stopped: (please realize I have no experience with lua and I'm rusty with my Java)

Boolean continuous = true; //This will tell us if the lab is still running after a certain amount of time (say, 600 ticks?)
int previous = 0;
while(continuous) //Boolean that checks whether the lab is doing any research
{
{} //the loop that refreshes the use of a new science pack, which will change "continuous" if the lab isn't running for more than 600 ticks
//otherwise, the loop adds 1 to "previous"
if(continuous)
{
{} //the loop that tells the power system how much power this building uses
//wherein the number of science packs used continuously, stored in "previous", is used as a multiple in the power equation
}

[bobucles]

Honestly just removing lab speed research will go a long way. Research DOES get slower, but the speed increase blows it all away

I think that higher tier potions could increase lab energy use. Like red still does 60kw, green 200kw, blue 500kw, purple 2mw. It lets one lab scale in multiple ways so you dont need multiple lab types.

[ssilk]

@person3triple0: I think it is not a good idea to couple evolution factor with research speed. Evolution factor and research speed have no commons. The evolution of the biters is a factor of time, pollution and destruction. The research speed (as in this suggestion) depends on what you already researched. It would be unfair to link it with the time; why should beginners have a more expensive research than experts?

Honestly just removing lab speed research will go a long way. Research DOES get slower, but the speed increase blows it all away

Right, that was also my thought, but I described how to change as less as possible. Hm, yes, lab-efficiency research could be removed to make it all easier and to enhance this acceleration-effect.

I think that higher tier potions could increase lab energy use. Like red still does 60kw, green 200kw, blue 500kw, purple 2mw. It lets one lab scale in multiple ways so you dont need multiple lab types.

Yes, that is another way to have similar effects. I can think immediately to about 10 or more ways to achieve this effect.

Hm. I'm far away to tell the devs how to program this, but maybe this is how I think this could be implemented:
As developer I would create about 3-6 mechanisms to affect the power usage of a (any!) device. Like
- the speed-up effect described in this suggestion,
- dependency on used items (as described by bobucles)
- dependency on how narrow you build similar devices.
- ... some more, that have this "exponential" effect...

And then I would make 3 game versions with quite different balancing that I tested a long time and tell other to play it and score, how that feels.

[person3triple0]

Again, it was just a suggestion. The evolution factor thing was kind of a limit thing in my mind? (I think My equation was the wrong kind anyway, it should be *(1 + Evo_Fac) ) Like, you can't research FASTER than a certain point in the early game. It would be highly OP for constant red research to speed the labs all the way up. You just play at THAT speed, assuming you have no power outages...

[Kevin Ar18]

As for how much power increases, you could go with a linear scale... or exponential?


An exponential increase might look like this:
Each science pak color increases the power use.

For example:
base lab power = 6kW
red pack = 10x power (60kW)
green = 10*10 = 100x power (600kW)
blue = 10*10*10 = 1000x power (6MW)
purple = 10000x power (60MW)

maybe with different numbers, but you get the idea ... perhaps might be too much for default game though, unless nuclear is really high output ...

[person3triple0]

maybe with different numbers, but you get the idea ... perhaps might be too much for default game though, unless nuclear is really high output ...

Not at all. My current solar/accumulator setup can do about 600MW, and I can easily double that with the resources I have.

[bobucles]

A power increase for higher tech tiers is nice, but energy is only a tiny part of the research challenge. The VAST majority of effort goes into potion production. Once that is built up, players will research whatever they feel is most cost effective.

It could very well be that some topics are too easy or hard to invest into. It is better to address those concerns first, rather than placing a huge number of arbitrary research rules.

For example, all the purple topics are only a fractionally bit more difficult than blue science. Obviously this is because only a single extra assembler is required to build purple potions.