Friday Facts #159 - Research revolution

[Grossen]

Hello, I've been playing Factorio for a while but registered this account to get a few opinions in and maybe do a few other things while I'm at it.

Science

1. Research - I like the fact that it is changing, the current system is a bit linear or rather it's more like a parable when it comes to research difficulty. Branching sounds really interesting and I am really looking forward to a new system.
2. I think The proposed ideas are a bit too demanding in time and resources and I agree with the majority that they should require more complex items, not more resources and also I really do not like the use of finished products / buildings for creating research packs.
3. The branching idea could be extended a bit, and maybe an interesting option would be the ability to work on multiple researches at the same time since now we have more than 4 types of research packs and belting them all at one research facility could be kinda tedious without bringing much game-play improvement.
I would suggest that one technology would not require more than 4 different research packs and in consequence we could design multiple research facilities for example (Basic Research/Production, Military Research, High Tech Research). The research Lab speed technology could also unlock how many concurrent researches you can make (1,2,3).I also suggest adding an advanced military pack since the basic pack is too cheap to justify high level military research: Advanced Military pack example:(explosives/rockets, distractor capsule, flamethrower fuel, alien artifact). You could branch to: red, green, gray, black (M), red, green, blue, purple (P), red green blue, yellow / green blue yellow (H).

PvP Dynamic

1. I think the match should be dynamic and action packed from the beginning to the end so thing should be carefully balanced, I do not like the idea of using biters to attack enemies or alien farms. I think for such attacks, if it fits the game, the game developers could add a combat drone line, unlocked from Automation2.
These drones could have basic functionality that could extend over research progress and they would be used to add action to the game early on and also as a mean to hinder the other factions without hindering your own activity to cross the map and attack.
2. Late game when the factory is automated and you can afford to spend your time on attacks there should be ways to defeat turtle defenses like: long range artillery, drop-ships, sabotage, stronger war machines(tanks seem to pale in comparison to Power Armor MK2 and also the fact that you can use only either shells or machinegun nerfs it a lot).
3. In multiplayer games research costs should be cut in half for packs > red technology to make it faster paced since a match must end in a reasonable time-frame. For example,mobas suffer from the fact that matches may take more than an hour, imagine multiplayer that takes over 10 hours.

PvP Technology/Balance

1. In PvE Red mags become useless since biter armor makes them virtually immune to them, this will also happen in multiplayer and the only viable static defenses are laser turrets that require electricity and this is not good (Green Mags become obsolete in less than 30minutes, they should be removed maybe.) I suggest adding more turret types, as people mentioned (rocket, artillery, heavy) and with the addition of nuclear power plants a new mag type, depleted uranium rounds.
2. Combat drones specifications:
They should be equipable with custom weapons and ammo like player avatars are, can only be made in factories, require fuel, take a long time to build.
Scripting for more control if drone allows (target priority, conditional logic (hp,ammo,etc)=> action(retreat, change priority, use item))

- Basic (Burner drone) Slow, can be set to follow player or to attack a specific point, no priority.
Requirements: Automation 2 - (steam engine, green circuit, burner inserter, water).

- Intermediate (Engine Drone) medium speed, can be set to follow, patrol, seek and destroy (priority: buildings/players) Basic Scripting and item usage (grande/defender capsules) enabled.
Requirements: Drone Control Research, unlocked by Engines Research - (drone control unit(new), engines, steel, fast inserter)

These are just examples the list can go on.

Whoa, sorry for the long post, just throwing in some ideas, maybe some of them can inspire the devs.

[MiiNiPaa]

You bought me with that, really - yet I am afraid 5 locos might not be enough - 8 looks more like it

Do not forget about needing several parallel tracks to run it.

[AcolyteOfRocket]

what this gane needs is an artilliery emplacement

Yeah, totally. Good to finally find some twin soul And you know what else do we need? A rail cannon!

THIS rail cannon - Gustav <3

Ingredients:
5000 steel plates
1000 gears
5 locomotives

Ammo:
1 car (Shells the size of a volkswagen you say?)
1 rocket fuel

Not sure if the car fits the shape of the barrel, unless it really is a VW

Better to use a storage tank, better still one that is full of light/hvy oil - Kill the biters AND un-bung your oil production chain in one fell swoop - now thats progress

[Andrzejef]

Not sure if the car fits the shape of the barrel, unless it really is a VW

Better to use a storage tank, better still one that is full of light/hvy oil - Kill the biters AND un-bung your oil production chain in one fell swoop - now thats progress

Ahh, I love the smell of oil in the morning.

Scripting for more control if drone allows (target priority, conditional logic (hp,ammo,etc)=> action(retreat, change priority, use item))

That is something I mildly spam since they showcased embedded code generated of selected entities here
And since they announced that you could place roboport in vechicles.
Even the simplest script for drone behaviour -

Code: Select all

 if(Copper&&Iron&&Coal&&Water&&Stone) {Mine.all();} 

this sort of things, I would be happy, because this would allow - kind of - the automation of outpost creation

3. In multiplayer games research costs should be cut in half for packs > red technology to make it faster paced since a match must end in a reasonable time-frame. For example,mobas suffer from the fact that matches may take more than an hour, imagine multiplayer that takes over 10 hours.

Depends - there are people that enjoy long sessions, almost like old tabletop strategies, or DF succession games (been there, done that). Others would rather get in, subjugate the opposition, get out, and still savor the day. Both are fine, so I think rather of toggle option or "speed-up" slider 0-100%

2. I think The proposed ideas are a bit too demanding in time and resources and I agree with the majority that they should require more complex items, not more resources and also I really do not like the use of finished products / buildings for creating research packs.

I generally am against science packs, as they feel a bit forced and artificial. While I agree that they have some important purpose (force player into certain researches/way and thus making the player to progress properly), but I think that AndaleTheGreat idea would do fine too.

- Basic (Burner drone) Slow, can be set to follow player or to attack a specific point, no priority.
Requirements: Automation 2 - (steam engine, green circuit, burner inserter, water).

Some kind of steampunk-like floating plaftorm? I dig the idea

[rpek32]

what the reason of using 3 separated tanks on one platform?
looks more useful

[Andrzejef]

Why, hauling 3 different liquids, silly
EDIT: Well, the devs didn't find it as obvious
In that case, hastening the load/unload process, using 3 pumps simultaneously.

Plus, as we see here

the tanks are connected, with some pump-valve device (just like they were adapted to hauling more than one liquid).

[Antaios]

1. Lots of red/green advances you couldn't use until you got to blue science, like the much of the bot chain.

I don't see this as a problem, really. It gives your labs somethings to do while you're building up your blue science area to keep up properly. Early on I'm not making enough blue packs to go blue science full on, so I alternate, stocking up 100-200 blue science packs between sets of green/red research.

3. How about you address the monotony of combat instead - what this gane needs is an artilliery emplacement that allows me to destroy biter bases from inside my own compound - let it use cannon shells, it's not like we are going to want to put them in the tank is it ?

Not sure on this one. On the one hand this is factorio and we should strive to automate everything. On the other, sitting in the middle of the base and destroying all enemies with artillery from there sounds in a way even more monotonous that what combat currently is (there's also no risk...).

I don't think something that lets you remove biter bases that are in your way from the safety of your own heavily defended compound would add to the combat because it removes all risk and strategy, it's just a 'click to remove biters here' button.

If the emplacement has a max range and auto-targets like the current turrets do it might work, it would be more defensive and would serve more to counter the biter's aggressive expansion, similar to a turtle wall + turret line, and you'd still have to do the work yourself (for the most part) to go and expand for resources.

4. Speaking of the tank, can we have a "research multiplier" we can change to vastly increase research times, by, say a factor of 2, or 10, or 50 or more ?? This will allow us to build, use, and maybe even enjoy some of the intermediate weapon systems before the my million research labs make them obsolete.

If it's a setting in the game start menu, I'd be all for it, but for the most part I suspect the new research system will significantly slow down high level research on it's own.

Finally I realise its an alpha game, but you now have higher end achievements in the Steam system and making the game harder to complete in 8 hours might frustrate the efforts of players who might be able to win under the old system, but are unable to do it in the new system - game balance is more important now that you have those achievements in the system.

And? Seriously, achievements should come after every other consideration on the game design, they're way to overrated and these days, don't really mean anything to barely anyone. I'd definitely rather not miss out on a cool feature or balance change just because of stupid achievements.

what the reason of using 3 separated tanks on one platform?

This seems to get asked a lot and I still haven't seen anyone answer with what I thought would be the obvuious reason - to cut down on graphics/animations...

If the wagon is six tiles long/tall, then the pump that attatches needs 24 animations to attatch to a single point in the middle of the wagon. (6 tiles each side, 2 sides, 2 orientations -> 6 * 2 * 2 = 24)

horizontal example,

[a01][a02][a03] [a04][a05][a06]
{-----------------()-----------------}
[a07][a08][a09] [a10][a11][a12]

If there are three points where the pump can attatch, then it would only need 8 animations (2 animations each side, 2 sides, 2 orientations -> 2 * 2 * 2 = 8)

horizontal example,

[a01][a02][a01] [a02][a01][a02]
{----()-----------()-----------()----}
[a03][a04][a03] [a04][a03][a04]

Checked the images from FF 157 of the pump animations and it looks like the tank ports don't line up with grid spacing , in which case I'd say it's probably so the pump doesn't have to extend ridiculously far to reach a fill/empty port. The idea of putting more than one liquid in a tank wagon doesn't make much sense to me - just have three tank wagons, and they'll store more to boot .


Now I'm wondering why the devs didn't space them so they could reuse the animations...

[Andrzejef]

3. How about you address the monotony of combat instead - what this gane needs is an artilliery emplacement that allows me to destroy biter bases from inside my own compound - let it use cannon shells, it's not like we are going to want to put them in the tank is it ?

Not sure on this one. On the one hand this is factorio and we should strive to automate everything. On the other, sitting in the middle of the base and destroying all enemies with artillery from there sounds in a way even more monotonous that what combat currently is (there's also no risk...).

I don't think something that lets you remove biter bases that are in your way from the safety of your own heavily defended compound would add to the combat because it removes all risk and strategy, it's just a 'click to remove biters here' button.

Hence the rail cannon suggestion
I agree, obliterating everything from your cozy factory seems tempting, but kinda pointless. Seeing some colossi people make, you might even not realise there are biters around (well, unless they start to bite, in which case artillery fire will also damage factory, and we don't want that). Max range and auto target (in other words mortar turret) seems sensible.
Tho I'd seen it rather like a some kind of hybrid of trebuchets from Stronghold and AoE series. Pick one or more, set up (mark the "danger zone" the projectile will impact), order to fire, and watch them missing the target you had in mind. Then pack it, put on train, relocate, unpack, set up. Rinse and repeat
Of course there could be researches to reduce the radius of "danger zone", within some limits.

[silver_26]

i like the idea that changes the science...
however i do see a minor problem...

and i have a few questions

1... can you make the science a bit more ( copper ) expensive ? caz you dont use copper that much...
or some other producs that uses more copper...

2. about the science... science 4 ... no more alien artifacts... ?'??
do you remove them from the game ?

caz.. well on my 450 hour map... i have 50.000 of them and how much do you use to compleet all science ? 2000 ? 2200 maby ?
and my map is not even cleared... i can get 150.000 more with out a problem
i would like some other use for them ...
Alien power ? alien guns / bullets or something

3. ar the start ( i think ) at .13 late .12 you where talking about space... that it would com in 14...
you have shown us a plaform in space but how is that going ?? i have seen / heared zero progress on that part for the past year orso

4 and lastly...
will you add the mod Infinite Research so ppl have something to research when thay are done ?

[Drury]

I could actually see FTL-esque PvP working on a basis of artillery wars, and it's more about who builds huge artillery shells faster than doing things with tanks and rifles and such. I mean, people are going to be turret creeping anyway, why not give them huge cannons that can fire all across the map and make those fun to supply.

[Andrzejef]

I could actually see FTL-esque PvP working on a basis of artillery wars, and it's more about who builds huge artillery shells faster than doing things with tanks and rifles and such. I mean, people are going to be turret creeping anyway, why not give them huge cannons that can fire all across the map and make those fun to supply.

So we actually came round right back to my idea - build a rocket silo, and launch a rocket, just as single player game. Only that rocket will not go into space, it would crash relatively far away, leaving NYC size crater (filled with hopes that the enemy factory is not bigger than that)

[Aru]

As for the PvP map generation, to spawn players radially as shown in the image on Friday Facts #159, you could just make the entire map radially symmetrical. 3 players get 3-way radial symmetry, 8 players get 8-way, etc.

[AcolyteOfRocket]

I don't see this as a problem, really. It gives your labs somethings to do while you're building up your blue science area to keep up properly. Early on I'm not making enough blue packs to go blue science full on, so I alternate, stocking up 100-200 blue science packs between sets of green/red research.

Sry, I could have sworn robots needed blue science but I just double checked and they don't seem to - maybe scrub my last on this

I don't think something that lets you remove biter bases that are in your way from the safety of your own heavily defended compound would add to the combat because it removes all risk and strategy, it's just a 'click to remove biters here' button.

However my suggestion for an artilliery emplace ment is serious, not to add to combat, but to add to the game by removing combat.

Combat in Factorio is always poor, and mostly bypassed with turret creep anyway. In the beginning it adds some threat to the game and pens you in until you are ready to break out - thats fine.

But late game combat totally sucks, and the crappy Ironman-in-a-robot-cloud model means the devs aren't much interested in the combat side either. Better to let us use the part of the game that is good (automation and resource gathering) to just blast the biters away without all the boring grind. You still need to make ammo for the gun, so there is still a reason to advance to get more resources.

And I am requesting a static emplacement, that can fire out to the far range of radar scanning. The "Rail" gun ( ) is an entertaining idea and would work for me if I could park it somewhere and automate its operation. But I don't want to fire it, I only want to feed it (using belts and inserters).

[malecord]

Dunno. I approve the science rework overall. But I also feel that the alien artifacts (although unbalanced) were going into the right direction: forcing the player to research and build offensive weapons.

If I had to design this I would make alien artifact a rare item that is only spawned by special nests which have to be reasonably distant from starting point and surrounded by a dense defense. Then there should be a bunch of "blocking techs" that require 1 artifact (or the product made out of the artifact: like artifact -> alien cells -> alien tissue -> alien pack) to be researched. In this way the mindless farming would be removed (no more hours spent going around killing hundreds of nests manually): in an entire game you would need like to kill 10 of the special ones to research all techs. But one would still need to build and research offensive weapons (and possibly scouting techs) to be able to obtain them.

Basically: instead of having 1000 player time where 20% is spent setting up production line and 80% is spent to manually harvest artifact like now, you would need to invest 100 player time where 90% is spent on setting up production line and 10% into manually grab the few artifacts needed. And ofc: turret creep should not be an option.

[Deadly-Bagel]

2. I think The proposed ideas are a bit too demanding in time and resources and I agree with the majority that they should require more complex items, not more resources and also I really do not like the use of finished products / buildings for creating research packs.

Part of the reason behind research being what it is is to force the player to automate certain products, such as belts and inserters. Enterprising players will set this up so that the research takes what it needs and excess is sent to production of fast inserters and such for the player.

This encourages new players to get into good habits of automating products rather than relying on player crafting. I'll admit even now I usually don't get around to automating assembler production until I start making MK2 power armour.

3. The branching idea could be extended a bit, and maybe an interesting option would be the ability to work on multiple researches at the same time...
I would suggest that one technology would not require more than 4 different research packs and in consequence we could design multiple research facilities for example (Basic Research/Production, Military Research, High Tech Research).

This was my first thought, have separate research "modules" that research a specific type. However this is inefficient, you will have labs doing nothing because you don't have any military research you need right now, or you might have one research in one module not using green packs while another module doesn't have enough. It is better to incorporate all research in a single lab setup.

Additionally by researching two things at once, they are each going to be completed in twice the time. Say two researches take 10 mins to complete, you could do both at once and after 20 mins both would finish. Why not instead research one, have it finish at 10 mins then finish the other? Both are still finished at 20 mins.

I'm actually looking forward to coming up with a few different types of lab setup. I think logistic chests could be used quite efficiently to tile labs in groups of 3 around a chest, or alternatively a single belt with circuit logic to limit the number of each type of research pack, with chests acting as buffers.

I think the match should be dynamic and action packed from the beginning to the end so thing should be carefully balanced, I do not like the idea of using biters to attack enemies or alien farms. I think for such attacks, if it fits the game, the game developers could add a combat drone line, unlocked from Automation2.
These drones could have basic functionality that could extend over research progress and they would be used to add action to the game early on and also as a mean to hinder the other factions without hindering your own activity to cross the map and attack.

There's no incentive to attack early right now, yes you might be able to hinder (one of) your opponent(s), but the time spent away from your base is absolutely not worth it. That's time your opponents will be using to progress to either launch a rocket before you, or surpass your military research and hit you back. Even in a 1v1 you would have to win from that first assault or do a LOT of damage.

The thing about turret creeping is you're creeping up to turrets with the same hp, range and damage. The moment you put a turret down in range it will be shot to hell before you can load the ammo. Only way this is different is if you have more advanced research, ie armour piercing rounds and damage / shooting speed research.

Therefore targeted capsules are definitely a good idea, which would immensely benefit single player too.

Late game when the factory is automated and you can afford to spend your time on attacks there should be ways to defeat turtle defenses like: long range artillery, drop-ships, sabotage, stronger war machines(tanks seem to pale in comparison to Power Armor MK2 and also the fact that you can use only either shells or machinegun nerfs it a lot).

Yes but the cannon shells of the tank have ludicrously more range than any tower or handheld weapon. In that respect, with enough shells you could demolish any base without ever being in range. If you see the player running at you, even in power armour just back off and attack from another angle.

In multiplayer games research costs should be cut in half for packs > red technology to make it faster paced since a match must end in a reasonable time-frame. For example,mobas suffer from the fact that matches may take more than an hour, imagine multiplayer that takes over 10 hours.

There's a lot of requests for a research slider in the game generation settings to make research cost more or less. Again this is a tool that solves this problem and benefits single player as well.

As for your ideas on combat, the whole system needs an overhaul. The combat shotgun is far too powerful compared to the other weapons and rockets are a joke. In single player, once biters reach critical mass there isn't really anything you can do other than walk around with a bunch of destroyer capsules which is just tedious.

[alan2here]

I like the way that right now research eventually caps out and various sorts of clever/optimum defense choices can be reasoned about and iterated on, the rules not then constantly changing as research happens.

On the other hand, endless research would be cool in late game.

A compromise could be to move further and further towards a fixed point of weapon power, that was a significant distance above the current endgame level, giving diminishing returns for each new upgrade, so that continuously upgrading with an ever increasing base size and production rate would yield more power, but also allowing a player to consider they were sufficiently approximately close to the optimum and leave it at that while still being at more or less full power.

[alan2here]

Let defeated players play as biters and their bases or just specific groups of them.

And/or allow other players to recruit these ghosts to play assisting them, with a new type of turret that has to be individually ghost player controlled, and so relies on a good player darting around the map to be much good at all, but has some other advantage, such as being very long range.

[Selvek]

Science:
100% agree. It always seemed silly that blue flasks were such a nightmare and purple ones were practically free (while at the same time being literally one of only 2 things in the game that couldn't be automated!). Of course, 90% of the nightmare was just getting oil production set up correctly, balancing the oil fractions, etc. Removing the battery and smart inserters from blue flasks will help a little, but there will still be a pretty huge gap between green and blue I think. I do like that this will give motivation for belting things like assemblers, pumpjacks, furnaces, engines, turrets.

One complaint is that turrets become pretty much obsolete as soon as laser turrets come out, so we will be manufacturing turrets only for science purposes. Re-balancing such that laser turrets aren't the end-all of defense would be nice. Perhaps a high-tech armor-piercing round that stands a chance against a behemoth biter? Or perhaps just a certain kind of enemy with laser resistance? Something to make it a tradeoff between the convenience of laser turrets vs the logistic effort of getting ammo to a large network of gun turrets (or flame turrets, for that matter).

PvP:
What if the map is just generated to be 4-way symmetric? That way things like far-distant mining outposts are guaranteed to be fair as well. Also, 8-team PvP seems a little insane to me... getting that many people on for a long of block of time sounds pretty much impossible.

On that note: Other PvP games work because either 1) you can drop in and out, because you are only battling for yourself and there is level balancing so players who play less don't wind up battling out of their league (EX: WoW), or 2) a full match, starting from equal conditions, takes < 1 hr so you don't need a huge time block. I think the only way to make factorio PvP work would be to start you with a headstart factory so the game can play out fairly quickly. Otherwise the winning team will be the one who is able to convince its players to do nothing else for 8 hours straight.

[Andrzejef]

One complaint is that turrets become pretty much obsolete as soon as laser turrets come out, so we will be manufacturing turrets only for science purposes. Re-balancing such that laser turrets aren't the end-all of defense would be nice. Perhaps a high-tech armor-piercing round that stands a chance against a behemoth biter? Or perhaps just a certain kind of enemy with laser resistance? Something to make it a tradeoff between the convenience of laser turrets vs the logistic effort of getting ammo to a large network of gun turrets (or flame turrets, for that matter).

It might just sound silly, but it also mighy just work. What about making gun turret an ingredient in laser (or maybe even any later) production?

[Unoxxis]

Gun turrets are already very valuable in terms of raw demage output, if I recall correctly, because they benefit from multiple line of upgrades (the turret ones AND the magazines ones). If endless reasearch gets a thing and is just more tiers of the research we already have - than the gap could become increasingly larger, that at some point it might be really worth it to go for them despite the lower range.
I personally would not rule out the gun turrets yet.