Belts vs Bots - A response to FFF #224

[oracleofepirus]

I really don't understand how the game is pushing you to use bots when you don't want to.

The push felt by some is because there is possibility offered by the game that v0.15.40 belts cannot reach. This alone wouldn't be an issue, as the entire reason for v0.16 is to make sure the belts can reach said possibilities, but for some reason, some of the bot-side players like to argue that bots deserve to be more capable than belts.

This is the core issue at hand. A sandbox game offers possibilities, but the furthest of these possibilities requires you to use bots. The problem with that is there are those on the bot side that say these possibilities should be kept away from those who use belts. If you told a pro-belt player that belts would eventually do all they wanted, they'd be thrilled. If you told a pro-bot player that bots are not the endgame, they're going to flip their shit. Those who don't ... might be reasonable people.

After that particular issue, lies the limitations of buff-only balancing. Buff-only balancing is great for players, but has its own limitations. One of those is severe power creep. Blizzard uses buff-only balancing for Diablo 3, and the original difficulty range that was Normal/Hard/Expert/Master/Torment 1 .. Torment 6 has gone up to Torment 13, with some people calling for Torment 15 or Torment 20. Buff-only balancing is still great for Diablo 3, because there are seven classes to play, which means that any given class can be adjusted without ever touching the other six, but Factorio doesn't have that. Everything in Factorio is connected to something else. There's a point at which you can't buff belts without buffing bots as well, and at that point, lies a bot nerf.

The bot-side players, of course, are just dead against anything that makes bots even slightly worse.

I myself, quite fancying belts, would be ok with some bot buffs if the bot-side would quit being so Republican about it.

[vampiricdust]

This is the core issue at hand. A sandbox game offers possibilities, but the furthest of these possibilities requires you to use bots. The problem with that is there are those on the bot side that say these possibilities should be kept away from those who use belts. If you told a pro-belt player that belts would eventually do all they wanted, they'd be thrilled. If you told a pro-bot player that bots are not the endgame, they're going to flip their shit. Those who don't ... might be reasonable people.

Wow, just wow. Nothing requires bots anymore than belts are required. If you choose to build a factory that requires them, that's on you. Bots are not the endgame, they are just a tool that comes late game. Launching rockets is the end game... but it seems you've lost the plot on this game. I have seen not one single post of people saying belts shouldn't be able to do what bots do, in fact nearly everyone is saying that belts should have the logistic flexibility of bots. If I could route resources on belts I would be thrilled as a bot loving player. I hope they give belts all the flexibility and usages of bots other than flying themselves of course as that would be silly.

However, I will argue that bots are balanced relatively to belts by research costs & straight up construction costs. Being overpowered would mean they are better than belts & trains in every use case. However, logistic bots are not dedicated to any one specific task. Any given logistic bot can do any given logistic request. This sometimes means that having too many bots causes worse performance than a fined tuned amount would. Nerfing bots serves no purpose other than to make people like you with irrational and condescending views of others feel better while it'll ruin the fun of people who like fine tuning a logistic system to optimal efficiency.

You're a bully who just wants to bad mouth others and break everyone else's toys so you're the only one having fun. Instead of having facts and reasoned arguments, you just troll and delusional assumptions about others.

[DaveMcW]

However, I will argue that bots are balanced relatively to belts by research costs & straight up construction costs.

Let's define endgame as "the point where you have researched everything and can build anything".

Therefore, by definition, research costs and construction costs don't matter in the endgame.

Therefore, bots are overpowered in the endgame.

[vampiricdust]

However, I will argue that bots are balanced relatively to belts by research costs & straight up construction costs.

Let's define endgame as "the point where you have researched everything and can build anything".

Therefore, by definition, research costs and construction costs don't matter in the endgame.

Therefore, bots are overpowered in the endgame.

I really do not like this argument that costs don't matter. They very much do matter. How much something costs determines how much production has to be dedicated to them. Bots also have infinite research costs which takes a great deal of resources to achieve the overpowered status. You're basically saying that building a factory for 40 hours doesn't matter. That the time we spend setting up the factory doesn't matter. So why the play the game if costs don't matter? Launching rockets don't matter, researching doesn't matter.... Factorio is a game that doesn't matter by your argument.

If costs don't matter, why should bots being compared to belts matter? Belts vs bots doesn't matter and is stupid debate on par with pistol vs artillery train.

[Phyrwurx]

Hello. First post / just joined the forum. I just wanted to add my thoughts on this conversation, not that they matter to anyone else, nor should they. I really enjoy playing Factorio. The point of the game is to build a rocket, some say. I play the game to get my mind off of work, bills, appointments, etc. Speed run? Never. A default vanilla map will take me 70ish hours to get to where I can start to build a rocket.

So, where do I stand on bots vs belts? I like both. I love belts and the rough decisions on planning out a core base that is not a total wreck by the end game. Base planning and building are my favorite parts of Factorio. I use logistics bots to shuttle resources to lonely chests within my walls and resupply me when I return. (My babies love me. lol) But, I wouldn’t want to tear up my creative belt layout I accomplished at the beginning to rebuild a bot base. I like the limits of belts as they are, and I like the convenience of bots in the end game. I notice belt throughput limits, but I like the challenges they produce. I don’t need more on a belt – I just need more belts. Spaghetti for all! And, I’m totally cool with that. At end-game, my core base is mostly belts (15-wide main bus) with about 150-200 logistics bots and about twice that in construction bots. Are bots more efficient for gamers who love stats or max output? Sure. Am I one of them? Nope. I just play the game and enjoy it my way, just like everyone else.

So, add a game option during map setup to disable logistics bots outside of character resupply, if you want. Make them more expensive to produce, or keep it all the way it is now. It doesn’t matter to me. We each play the game how we want, and we’re all here because we love Factorio.

[m44v]

However, I will argue that bots are balanced relatively to belts by research costs & straight up construction costs.

Let's define endgame as "the point where you have researched everything and can build anything".

Therefore, by definition, research costs and construction costs don't matter in the endgame.

Therefore, bots are overpowered in the endgame.

I really do not like this argument that costs don't matter.

They don't matter, 99% of the resources go to science and modules, nobody maths how much a row of roboports is going to hurt in resources, that's how everyone builds at that stage even if you don't like it.

How much something costs determines how much production has to be dedicated to them.

Right, like a silver of what the silo consumes.

Bots also have infinite research costs which takes a great deal of resources to achieve the overpowered status.

That's a fair point, how much is that? 15 rockets? so like 3 hours of twiddling your thumbs with a single silo. A lot of effort there, nothing like all the effort you need to invest in if you want to go with belts instead of bots.

You're basically saying that building a factory for 40 hours doesn't matter. That the time we spend setting up the factory doesn't matter. So why the play the game if costs don't matter? Launching rockets don't matter, researching doesn't matter.... Factorio is a game that doesn't matter by your argument.

If costs don't matter, why should bots being compared to belts matter? Belts vs bots doesn't matter and is stupid debate on par with pistol vs artillery train.

you're making an hyperbole.

[vampiricdust]

you're making an hyperbole.

No, it isn't. The whole point is resource management... Factorio is a resource management game. Saying resource costs doesn't matter is saying the whole point of the game doesn't matter. Tell me how Factorio isn't a resource management game...

[Zavian]

you're making an hyperbole.

No, it isn't. The whole point is resource management... Factorio is a resource management game. Saying resource costs doesn't matter is saying the whole point of the game doesn't matter. Tell me how Factorio isn't a resource management game...

Well different people get different things out of it. To me factorio is a logistics and optimisation game (and resource management is only part of that).

But regardless of whether you think factorio is a resource management or a logistics management or an optimisation game, there are different resources to be managed. Raw resources like ore and oil. Power. Player time and enjoyment/fun factor. Unless you are concentrated on rushing the rocket launch asap, then I don't think anyone argues against the resource costs of construction bots in the endgame to make building things like solar fields faster and more convenient. Similarly if you really want to minimise infrastructure costs, then you only need one assembler, and no belts or inserters. You can just hand insert everything. But that would be insanely tedious for most players, so they build belts. And they upgrade yellow belts to red, not because red belts are more cost efficient, but because it is more convenient for the player to manage half the number of red belts, compared to double the number of yellow belts.

When you actually start building a megabases resource costs aren't really a consideration for most megabase players. What matters is player time, player convenientence and ups. Regardless of whether you are going for a belt or a bot based megabase you will want productivity modules in pretty every assembler that can use them. You will almost certainly want speed beacons as well because they reduce the total amount of modules you need, and reduce the number of active entities. At that stage of the game the game you already have automated mining and smelting, you have already built power to run your existing setup, so that as long as the game is running (and not paused) the only cost of more resources is player time. (Sooner or later you will need to build more mining outposts, hence the player time cost). Let me rephrase that the only actual cost of building more bots is waiting a little bit longer for them to be built/occasionally building new mining outposts. And normally you aren't sitting there waiting whilst the bots build, but are busy building your base. So whilst the bots do cost resources, those costs are basically irrelevant to the megabase player, once he has actually started building his megabase. (The only costs he cares about are his player time investment, and the ups costs of his design. Those are the costs that typical megabase players are trying to minimise/optimise).

[oracleofepirus]

No, it isn't. The whole point is resource management... Factorio is a resource management game. Saying resource costs doesn't matter is saying the whole point of the game doesn't matter. Tell me how Factorio isn't a resource management game...

The short description on Steam says

Factorio is a game about building and creating automated factories to produce items of increasing complexity, within an infinite 2D world. Use your imagination to design your factory, combine simple elements into ingenious structures, and finally protect it from the creatures who don't really like you.

Nowhere does it say Factorio is a 4x. Also nowhere does it say bots are endgame.

Costs for infinite research is not a justification for keeping bad balance. It takes 122 050 copper, 151 475 iron, and 22 100 plastic just to get to infinite worker robot speed. That's 5000+ robots. On top of that, each unit of infinite worker robot speed costs the equivalent of 10 000 more worker robots. Clearly, cost isn't an issue. The issue is that belts don't have such equivalent research, and just increasing belt speed does nothing. Just because it costs over a billion dollars to build a casino doesn't mean Trump should get a huge tax break.

[vampiricdust]

Alright, well. I've had my fill of debating people's opinions. You guys dont bother reading and facts dont matter. Well, hope they nerf bots into oblivion so at least all you guys are happy.

[Dizmal]

The problem is that bots go from '0 to 100' the moment you get more advanced logistics chests. Part of this is because the technologies leading up to the chests are overpowered, so there's no middle ground by the time a player gets the chests, everything is in place to abuse logistics.

If I, as a user with one single post had undisputed reign, I'd suggest that roboports and broad-coverage networks are actually the issue - not the bots. I'd consider...

Nerfing Logisitcs networks by decoupling the Roboports;

• Add a 'tier 2' blue cable requiring either red circuits or processors. Blue cables would serve as the logistics network. If you want a chest to be covered by a roboport, it must be blue wired.
• Adding a logistics antennae/beacon. It would cover an area much like existing roboports, but smaller. Just big enough for groups of chests or chained at train stops. They wouldn't link like current roboports and would need to be wired to a network.
• Move logistics chests to earlier in the game to compensate for reduced logistics coverage.

Modifying Roboports;

• Introduce a 'tier 1' roboport. Mid-game, 1x1, holds/deploys one bot.
• Significantly slow down charging speeds. Make charging speed an upgrade.
• Make logistics networks track how many bots they have 'out' vs how many bots can be docked for charging. Logistics networks will not deploy more bots than can be charged at once.

The goal is to make an early roboport which acts like a longer-range inserter still practical for solving certain problems. E.g. these would be useful for reloading turrets, loading odds & ends onto trains, repairing defenses/trains, or slightly streamlining an unweildy belt situation. Without the overpowered network it would effectively end full-base automation unless a player invests significantly in wiring.

[olafthecat]

The problem is that bots go from '0 to 100' the moment you get more advanced logistics chests. Part of this is because the technologies leading up to the chests are overpowered, so there's no middle ground by the time a player gets the chests, everything is in place to abuse logistics.

If I, as a user with one single post had undisputed reign, I'd suggest that roboports and broad-coverage networks are actually the issue - not the bots. I'd consider...

Nerfing Logisitcs networks by decoupling the Roboports;

• Add a 'tier 2' blue cable requiring either red circuits or processors. Blue cables would serve as the logistics network. If you want a chest to be covered by a roboport, it must be blue wired.
• Adding a logistics antennae/beacon. It would cover an area much like existing roboports, but smaller. Just big enough for groups of chests or chained at train stops. They wouldn't link like current roboports and would need to be wired to a network.
• Move logistics chests to earlier in the game to compensate for reduced logistics coverage.

Modifying Roboports;

• Introduce a 'tier 1' roboport. Mid-game, 1x1, holds/deploys one bot.
• Significantly slow down charging speeds. Make charging speed an upgrade.
• Make logistics networks track how many bots they have 'out' vs how many bots can be docked for charging. Logistics networks will not deploy more bots than can be charged at once.

The goal is to make an early roboport which acts like a longer-range inserter still practical for solving certain problems. E.g. these would be useful for reloading turrets, loading odds & ends onto trains, repairing defenses/trains, or slightly streamlining an unweildy belt situation. Without the overpowered network it would effectively end full-base automation unless a player invests significantly in wiring.

I would agree, with this.
You could also add ground bots, slower and run on fuels.
The better the fuel, the faster the bot.
They also have to go round obstacles.

[xng]

I won't pretend I've read everything written here, but I am one of the ones that would appreciate if logistics bots were optional per world.

As it is now, as soon as you reach logistics you don't need any other blueprint than this:

Screenshot20180123_145837.png (47.26 KiB) Viewed 9249 times

You, still can do more, do your own logistics instead of letting the game play itself, but why go through the hassle if the game is designed to be over at this stage.

If bots only did building and player logistics, you would still need to think a little and could still have fun for hundreds of more hours on the same map instead of having to restart.

The key word is to have the option, not make a change that breaks every botbased map. And with this change/alternate game mode, you could also have bots a lot earlier as they no longer break the game. They only take away some of the really bad parts of the game, like clicking until your mouse breaks. Planning blueprints and making it fit and be repeatable in the factory is so much more fun to me than only using one blueprint to make every item.

I'm pretty sure that if that boardgame (from original post) autoplayed the cards while you were sitting back watching it, it would not be as fun either, and if that started happening after playing for a while it would also feel like the game was more or less over, just waiting for the result screen.

It doesn't have to be black or white, the changes could easily fit in a mod, but as mods are considered cheating (when achievements are concerned) it's prefered to have all the good stuff as part of the base game.

[Kilitar]

Well, one possible solution instead "nerf robots" might be just tech like "logistic central control tech" with unlimited tech update.

This tech would limit Robot total number CAP for factory. Which simply motivate player to where he should effectivelly use robots as their total number is limited. But with investing more and more to unlimited tech robots total CAP would definitelly grow. I think such approach would promote unlimited belts/trains transportation on one side but also leave current bots effectivitiy in critical parts of factory.
So something like Robots researched - robot CAP limit 100. Next tech level +100 etc.

If you want whole mega factory "robotized", you simply need invest even lot more of resources into tech to have big swarms of robots.

[midknightwraith]

So here is my perspective from somebody who is coming to the party late.

So I found this thread while looking for information on how to best set up a bot network. Basically because every time I get to bots, I feel like they aren't worth it. Even construction bots. Because, they are too slow. And I keep thinking to myself I must be doing something wrong, because everybody talks about how efficient bots are, and how they are so much better than belts. And I see a bunch of potential, but never quite realize what I think that potential is in an actual game.

I have never built a mega-base, 3rd tier modules are something I build to have the best armor in the game. However, my own limited experimentation with modules, even without beacons, tells me there is something fundamentally wrong with belts. Especially in the late game where certain assemblers require 10s of multiple products relatively quickly to produce at full output.

To my way of thinking, you need 3 legitimate competitors each with strengths and weaknesses to make any system balanced, and that balance has to be maintained over the course of the entire game. The problem is, over the long haul of the game (once you get to infinite research) it becomes obvious that the three different available logistics systems are not equal. The simple fact is Bots have infinite research, while Belts/Trains/Inserters do not. Hence, once you get to the infinite research stage, Bots will eventually dominate. It does not matter what else gets changed, that will ALWAYS remain the case while bots get infinite research and the other options do not. And thus there really are only two solutions to the Bots vs Belts debate, if, as Kovarex suggested in his followup, that the issue is the magnitude of the efficiency difference.

1) Eliminate Infinite Bot speed research
2) Instantiate infinite Belt/Train/Inserter speed research.

Heck even mining productivity has infinite research, which again pushes things to bots. Because, you could get to a point where a single miner can fill/flood an express belt by itself. Of course, one could argue you could push that to the point where a single miner can flood a Provider Chest too, right?

The problem with option 1 is that Modules exist, and because of the speed/productivity boosts involved only the infinite speed research bots can hope to keep up with the resource requirements, which is necessary for Mega base construction. Or at least that is how it is presented by the Mega-Basers, as I have no experience with this myself. But as I alluded to, even without beacons I noticed that keeping modulized assemblers fed with components is difficult.

Hear is a crazy thought, what if, Beacons could also impact Belts, Trains and Inserters? And even Bots/Roboports if infinite bot speed research is eliminated. If done right, it should create a situation where the module production step ups end up matching the logistic network step ups, so that things always remain in balance.

Then you really do have 3 legit transportation network choices. Trains become the bulk long distance haulers. Belts are the quickest, and more flexible (than trains) but more work than Bots, and Bots are slower (than belts) but the MOST flexible delivery system. The suggested limitation on simultaneous Bot chest access is probably also a good idea, otherwise bots > belts no matter what the speed difference is (assuming it isn't so low as to make logibots unusable to begin with).

One last thing, eliminating infinite bot speed also has a negative impact on Construction bots. So I might suggest giving them infinite speed upgrades is okay. Side note: Is it just me? or do construction bots not take advantage of the additional cargo capacity when building stuff? That also feels like an issue that should be addressed.

[GlassDeviant]

I see a lot of opinions about a lot of things that just don't make any sense.

First this crap about launching the rocket. Most people have never launched a rocket, or if they have, they did it once and have never done it again.

Second, far from "everyone" switches to [implied logistics] bots as soon as they are available. Read the forums. Watch vloggers. Lots of people never bother to use logistics bots, though I notice a lot do use construction bots, and many of those use a mod like TinyStart, often never making any more for the majority of the game. I almost never see logistics bots used outside of megabases, and sometimes not even in those. Personally I completely reject logistics bots and for the same reason I am guessing a lot of people do likewise: using the belts and trains is more fun.

Going to stop there or I will be here all night pointing out silly assumptions, and just get to my main point: There is nothing wrong with bots, and it's each players choice whether to use them or not. Removing them takes away the option. Options are good in games. If you want to make there be "only one way to play the game", not only are you limiting the freedom of choice of your users (and potential users), you are preemptively choking the growth of your user base.

[HurkWurk]

i think a better baseline understanding of what role bots are supposed to play is needed first.

1. I think everyone understands this is about logistics bots only.
2. I think it should be obvious that logistics bots are an evolution past belts. to me, they should be better overall.
3. I think that the game is progressive in that way, in that early on we do things one way, and later on a different way because the results are better.

with that said, to me, 99% of what drives bot use is density for beacons and chips. in a vanilla game, bot loaded/unloaded machines are much more dense than they can ever be compared to using space for belts, especially using space for enough belts to properly beacon. based on some napkin math, belt mega bases need about four times the total space. and an optimal belt base would use about 16 times the space.

the reason for that is beacons. beacons in a double row make the compressed area require bots to feed/pull fast enough, while an optimal belt megabase would use few beacons. because of that, it would require much less power, shrinking the massive solar fields, and using more factories to achieve the same output.

________________________________________

I think a good test of bots/belts would be a megabase that is producing 1000 science per minute. then compare/contrast the two.
This would allow people to visualize both scenarios.

[oracleofepirus]

2. I think it should be obvious that logistics bots are an evolution past belts. to me, they should be better overall.

And there is the reason why the developers think bots are unfun.

If you introduce a player to a game mechanic, then turn around and say that mechanic is entirely obsoleted by a mechanic that only exists in endgame, you've alienated that player.

No new player is going to spend a ton of time on a belt-based factory, and then continue to play after finding out they have to tear down everything and rebuild for endgame. Even if this strategy is executed well, the best you've done is confuse them. You've made everything before endgame into a tutorial, then given them something that has nothing to do with the tutorial.

To say that bots should be better than belts is to say that nothing before that matters.

[tolomea]

TL;DR

One way to nerf bots would be by limiting how many can access a container at once.

What ?

Sometime back the devs discussed the OP nature of bots and how they are superior to belts in most ways, thus heavily encouraging a late game switch to bot based factories.
I was thinking about that and it occurred to me that one way to limit bots so they are inferior to belt/inserter setups would be to limit the rate at which bots can empty and fill containers so that stack inserters would always be faster. The simplest conception of that would be something like only one bot can access a particular container per tick. (values to be adjusted to achieve the desired numbers).

[Koub]

[Koub] Merged into (one of the many) older topic (s) where this has already been suggested.