Friday Facts #225 - Bots versus belts (part 2)

[Inv1s1ble]

Factorio seems to be a game which cares about realism for some extent. For me it always seemed super unrealistic that bots are able to fly through each other. So why not limit the amount of bots in a certain area at the same time or add pseudo collision detection so that bots need to fly around each other - maybe even with some safety distance?
It would nerf bots, yes, but not because of balancing, but rather realism.

[Piroko]

Hey there,

the splitter improvements seem cool but I'm a bit wary that splitters become 'clickable' now. I tend to use my belt paths as, well, paths where I can't accidentially click on assemblers, filter inserters and chests while walking. Don't let that stop you though, I'll adapt.

In my ~500 hours of playtime in this game I have never built a base that used more than 1000 logistics bots. Belts, trains and changing map settings have always been the thing that attracted me more than transitioning my bases to full robot bases (and crazy mods that only really work with robot bases for that matter).

For me personally the part that feels overpowered is that they have "perfect response time", as in they can keep Assemblers fed even at a large distance. I don't think this is 'fixable', because any limitations like 'maximum requests per chest' or 'slowed load/unload' will only lead to more awkward builds and potentially unintuitive behavior.

Giving robots a longer charging time would put a slightly stricter hard limit on maximum throughput within a limited space, I kinda like that idea even though it will primarily just make people spam more roboports since the space limit that is "having enough roboports to not have charging congestion" is so miniscule to start with. The additional power drain from all those roboports is probably more noticeable.

Maybe another change could be to lower the Worker robot speed bonus. The mining productivity bonus is only 2% per upgrade and it's still worthwile for me to research, the worker robot speed upgrade is something close to 20% boost relative even at research level 7? (+(305% + 65%)) That's a lot imho. Even having a speed increase of 300% in the first place is already a lot.

[Xycor]

Bots could have a carry weight mechanic. As loads are heaver range and speed decrease. Some items like furnaces and factories are simply too heavy. This lets bots move some items but not all items. The item weights can be tuned by the game designers so bots are helpful at reducing tediousness but not completely overpowered as they are now. To keep blueprints functional perhaps construction bots would slowly carry items in teams or require multiple trips for assembly.

[Xuhybrid]

Anyone who took the last FFF seriously clearly didn't bother reading it because it was very clear it wasn't a real plan. Reminds me of those people who only read the headline of an article and run with it. Oh wait, that's most people.

[rullie]

I happen to think belts and bots are fine as they are now. What needs nerfing is requester chest. Either by making them physically bigger or require research to unlock requedtable items.

Just an opinion from a manual inserter

[Froilen]

can you do the same with inserters ?

tell them to put the item in the other side of the belt?

please that would make them easier to build some complex belt layouts

Have you an example where you see magnificent problem without this inserter ability? To make an inserter fill the other side of the belt is basically an easy task with the use of some more belts:

this one in particular

[<NO_NAME>]

Factorio is not a competetive game. It is a sanbox game. It is fine stuff is not well balanced.

LOL! I'm saw that opinion a few times in this thread. It is basically like writing "Factorio is a sandbox game so is it fine if it sucks." - just a few levels down.
No, the game should be well-balanced. It will ensure that it will create an interesting challenge and be fun to play. It doesn't mean that it won't allow different play-styles, thought.

I'm understand where these opinions come from. We are afraid that the play-style that we are used to will be no longer possible or that are current factory will cease to work. To be honest, I am afraid of that too.
However, we have to remember that this is still early-access game and there will be changes. It isn't finished yet; it isn't fully balanced, and you can't do that without changing anything. So, if these changes brings more positive consequences than negative, we have to sacrifice some of our comfort zone and accept that. This will eventually come in our favor.

[hitzu]

Why not just restrict the number of bots per roboport just like personal roboports have? Possibly with the diminishing return inside one logistic network. Possibly with a minimal distance restriction between roboports. Possibly along with making the area of effect to be round. Possibly along with splitting roboports into logistic and construction roboports (and even possibly military roboports, but it's another story).

Another proposal: remove all requester slots but one from a requester chest.

[<NO_NAME>]

<Something that I am unable to describe>

This forum needs kudos so much.

[bobingabout]

Thumbs up on the filter splitters
I don't mind a nerf to bots, I mean, I can always undo it in my mods, just don't go too heavy handed with it.
Don't completely remove the bot carry capacity bonus, maybe reduce how many levels it has though. you did it with normal inserter stack size bonus when you introduced stack inserters. from a bonus of 4, to 2.
Speaking of stack inserters... you've plastered that FFF with pictures of you using loaders. Put loaders in the base game, they'll help buff belts.

I mean, Inserters can be used by Bots, Trains and Belts... but loaders NEED to use belts... if you want to load a train/factory from a belt, 1 loader between the train/factory, and the belt.
You want to do the same thing with bots that feed from a chest, you first need to use a loader to unload the chest onto a belt, then use another to load a train from that belt.

Faster belts, and inserters should be fine, just one more tier with relatable improvements to existing stuff sounds good to me.

Speaking of belts with stacking items.... definitely better than needing to pack or unpack an item, Perhaps consider this... ANY belt can carry a stack of items, but it must be created by a STACK INSERTER!!! As for graphics, well, each item would need to look different to stack well... so don't bother playing with the graphics, I mean, do stack inserters change the graphics when holding a stack?



So summery:
Love the splitter
Stack inserters should be able to add stacks to belts
a higher tier belt and inserters is a good idea
give us base game loaders.

[Mental_Flatus]

After reading several posts, and giving it some thought. I believe that no matter what you do about bots your gonna draw the ire of the bot lovers. They are powerful.

As far as belts are concerned. I think that the biggest point of contention is getting material down the bus from its point of origin, whether it be from smelter or train. I think that the stack inserters should be left alone. Higher tiered belts would certainly solve the issue of super wide buses. I would think that belt balancer / compressors would most definitely be designed with the given tools. There are two ways that material can be picked up by inserters at the assembler, 1. Have the super-fast belt compress to a point that the inserter can grab what it needs, or 2. Put down slower belts to slow the material enough for inserters to grab. Either process will work. Most people enjoy the satisfaction of seeing compressed belts and will work to maintain that status quo. It will most definitely drive creative design.

[Caine]

Preface

I thought it might be interesting to collect all of the sufficiently-specific suggestions made in the thread, and aggregate them into rough groupings, with references to the posts. So I went and did that, then thought I should go a bit further; prepare for giant wall of text.

You have the same idea as I and/or understood what I was going for. Thank you so much for doing phase two and three of this work. I dreaded having to go through all these pages to collect the information. I agree with your overall analysis and collection, but I need some time to digest your recommendations. I see you made MANY of them. I hope that a smaller solution will present itself.

but I'd collapse space efficiency and flexibility together, personally. Space efficiency isn't really a direct concern; it only matters insofar as it impacts your ability to freely transport items within a factory module (a set of machines doing single task or set of related tasks and conceptually grouped together by the builder during design/construction), and freely place factory modules relative to one another.

Some clarifications on the metrics that I described. I probably named them poorly:

• Space Efficiency: specifically the constraints that the logistic solution imposes on local sub-factory design.
• Flexibility: easy of altering and/or extending the logistic solution with new types of resources, new destinations, etc.

They are related but not entirely the same. Either way, it does not impact your analysis of belt/bots niche overlapping and the domination of bots in the shared niche.

Bots have very few space constraints and you can place them practically anywhere. I am not a megabase player so I cannot comment on the finer intricacies involved here. Trains on the other hand practically dictate the layout (see e.g. train smelting). Belts are somewhat in the middle. One reason to include the space metric is to seriously consider arguments which address beacons as they strongly impose a specific layout (the 8x8) which forces either logistic bots or belt braiding (or vertical belt stacking if that would get added).

Belts have pretty low flexibility, their contents are fixed by the inserters which put and take from them. Bots have no direct link to content or types, those are determined by chests and bots automatically adapt to them. Trains are fixed by their stops but can be extended or additional trains can be added. They can also easily share tracks, which bots don't need to at all and which is tricky with (sushi) belts.

Either way, I thought that a more structured analysis of the problem was required before we could dive deeply into solutions. Without understanding the issues and/or agreeing on what kind of game we want Factorio to be, the arguments will fall on deaf ears from both sides. So, again, thanks for the awesome summary and analysis work.

[Crow08]

A major sideeffect of nerfing bots via chargetime or cargospace would be a dramatic fall is convenience and usability for personal robots, which is not the actual target of the nerf.

My solution for nerfing bots would be not nerfing the bots themselves, rather nerfing the throughput to requesterchests.
If you limit the number of bots that can simultaneously fly to a requesterchest, you can make them worse for high throughput demands while preserving their effectiveness in all other areas.
You can try to offset this by using multiple chests but this brings it on par with belts and inserters.
Another problem may be the new filter-option of storagechests with this it is possible to substitute requesterchests with filtered storagechests.

[SamuelS]

What Do you think about locking higher tier products behind bot levels. Not creating new bots but after x amount of time or x items carried, the bot becomes able to carry new items.

Example tiers

Tier 1:
Iron ore
Copper ore
Stone
Coal
Uranium
Iron Plates
Copper Plates
Steel Plates

Tier 2
Green Circuits
Yellow Belts, Undergrounds, and Splitters
Inserters, Fast Inserters

This could be implemented by a Items_carried statistic for robots. Much like the Products_finished attribute that Assembling Machines have. After Items_carried reaches x number, the Bot can then Tier 2 items.

Other ideas

1. Maybe add a research tree to unlock the ability to carry new tiers?

2. Add weight to every item in the game, add a new attribute to bots called Weight_carried. After the total weight carried reaches x weight, allow it to carry heavier items.

3. Make heavier items slow down the bots. For example say bot speed is 10 miles per hour. A bot picks up 5 iron plates that weigh 50 pounds. That bot can only move at a max speed of 9.5 miles per hour until it has dropped off the plates. If it was carrying a water barrel (Cause thats heavy), it could only move at a max speed of 5 miles per hour.

[bobucles]

I've seen a lot of ideas on potential solutions. A lot of them simply aren't workable or cause huge issues in bad ways. For example

Bot air collision

Currently the CPU cost for bots is basically nothing. They pick a direction and fly X distance every tick. Mid air collisions would force bots to scan their surroundings every tick. That is VERY CPU expensive and would destroy UPS. ALL bases would suffer greatly. Bad idea all around.

Bot decay

The current inventory system can't really deal with "durability" as a concept. A bot with unique durability would need its own inventory slot, and a stack of bots with shared durability would all need to be fixed at the same time. You'd need a way to control inserters to make sure they pull out damaged bots while keeping the healthy ones alone. If bots break in spots without repair access you're in for a world of hurt because they'd clog roboports and basically shut your network down. It seems really punishing and would be incredibly difficult to manage. Most of all the status of a roboport filled with broken bots would be invisible, so it'd just not work without any clear indication as to why. Yuck.

Throttling chest speed

I guess it's an idea, but players would just lay down more chests until they get the desired throughput, right? This kind of idea would affect beacon builds on the super fast recipes, but ultimately would change nothing on 5second+ crafting recipes. There's also the issue if bots have to wait in line to use a chest, your network can be depleted of bots simply because they're waiting on a chest. Don't even think of deconstructing a full chest or the problem will be a hundred times worse than it is already. That's kind of nasty.

More energy expense
More resource cost

I build more energy. I mine more materials. This doesn't change the effectiveness of bots, merely the effort the set them up. In the post game it's safe to assume that players have "unlimited" resources and will simply choose their best option regardless of the cost. Money is no obstacle to late game balance, so all you will do is hurt legitimate pre-rocket play and nothing that you want to change will actually change. Pretty pointless.

Bot capacity nerf
Bot charging nerf

It'll work. Don't get me wrong. If bots suck more than belts then players will use belts. But that's probably fixing the problem from the wrong angle- the huge surge of platinum mad posters certainly don't like the idea and modders will casually "don't care" it back to pre patch status with mods. Not so great.

Cargo weight/"oversize items"

This screams all sorts of issues. The first is that if bots can't carry X item, then you can never get supplied from the logistics network. If bots can not construct with heavy things, then the blueprint system breaks down. There's no obvious way to know what a bot can or can not carry, and if you decon a chest with immovable items what will the bots do? What a nightmare.

Better belts

Ultimately I think this is the best solution. Bots have a scaling power beyond the end game. Train networks swell in power with high tier fuel and huge stretches of post game territory. Conveyors just sort of peak at red belts and then don't get much better. Trains can handle post game resource demands, and bots can handle post game demands, but blue belts definitely struggle in a major way. They clearly need some love. The idea of a stacker belt is really sexy, and smart splitters open up exciting new possibilities as well. There is no shame in making a belt upgrade recklessly overpowered. In fact it's probably better that way. If all post game logistics are overpowered, then nothing is.

Tldr: I think stacker belts with stack loaders will have what it takes to deal with late game resource demands.

[Ishimuro]

After removing the splitter magic, that splitter update is a nice way

But it feels a bit strange, first removing some handy belt options and then going into a belt vs bot story. (After these compression removals and stuff i tend to go more for bots)

Most stuff i read is about doing some weird stuff with logistics but keep construction bots, but both are QoL, both are flying things so why handle em different cause one side is QoL we want, the other is explained as op stuff.
Like many others i look at bots as the next evolution step for my factory. So i want to use them, i want them to be a way better than belts.

Factorio is about building a factory using my imagination and i want to use my imagination within this sandbox game, but it feels more and more like i have to use your imagination? We need more options to play with, not even more restricted ones.
I want to build a factory with as much options as possible, as free as possible.
I could run with energy free, space consuming belts, which, without planing ahead, will be such a hugh mess i have to redo, to upgrade, to fit my needs.The last updates got us some nice QoL, improve that even further, if working with and maintaining these belt layouts become easier to handle and less work intense, people stick with belts.
But i could also run with bots, less space consuming, but energy hungry, i dont have to care about the mess i made with belts and the work intense process to upgrade and redo them, i dont need to care about moving my whole base to fit some more lanes to supply one part of my factory, but i have to maintain a constant grow of my bot stock and also the energy supply. It's the next level, getting all my stuff into the air, and the sky is huge.

I like the flow of stuff at belts, like veins keeping the beating heart of my factory alive, but i also like these botswarms like a beehive, never resting.

I don't like bots vs belts, i like to see them together, not cause i have to, no, cause i want to.

If it feels like belts should be used more, go on making em more handy, easier to scale and maintain.
It doesn't feel like a good way to just get bots down to the level belts are at the moment, but should be that belts get up to the level bots are.

Sometimes i want to puzzle, spending hours just to solve some small issues, sometimes i just dont want to care about it at all, or at this time.

I dont't know, some ppl say pls do something to bots, it isn't necessary to use anything else, but thats a sandbox, for the own creativity, if these ppl don't have the creativity to solve a problem or don't have the effort to do so, is it the job of the devs to help them, to force them to go another way?

I'm months into a game that could be finished within a Day, cause there are so many ways, so many tools to play with, pls expand our color palette but leave it to use which color to use.

Vote for the evolution, get stuff into the air and beyond.

TL;DR

Feels like comparing bots vs belts and next time we start with steamengine vs reactor.
Keep the spirit of evolution within the game (logistics).

[PiMan]

I have an idea, although it doesn't really address the problem: Make the slowest belts free.
As in, don't require them to be assembled at all; give every player an unending stack tied to either the first or last belt slot from the beginning of the game.
It wouldn't change how belts or bots are used in optimal factories, but it would make the game feel more belt-ish for new players. (And would cut the start of game/pre-belt tedium a little)

As for addressing the problem at hand, I'd like to see the effectiveness of making a fourth belt class; one that worked at infinite speed. The moment something is dropped onto it, it immediately moves to the end of the belt's path.
There are two things that would need to be done that differ from other belts: It couldn't move players standing on it, and it probably shouldn't have splitters in the same belt class.

[OrangeFlame.476]

In order to visually do the item stacks on belts you can make the item into the appearance of an open-top bin with the item-stack's icon inside that bin. The height the icon is placed inside the bin represents a percentage of the stack limit for that item, or the stack limit for stack belts depending on how that is implemented. I don't think stacks should try to combine if they can't move forward on the belt without the help of inserters, or maybe a splitter. Bin appearance replaced with single item when there is only one, and single item changed to bin when there is more than one. The tech cost can be explained as them being plastic bins. Ignore the fact that they just magically appear and disappear as the magic of video games for the sake of fun.

[Slayn25]

If you guys do go the stack belt route it might be good to add a checkbox to inserters to only cycle for full loads.

[anarcobra]

The more I think about this the more confused I get.
After reading more or less 20 pages I've completely lost sight of what role bots are supposed to play in the game.
Long distance single items delivery to player and other low throughput items?
Short distance high throughput? If bots are supposed to be only for long distance single items and delivery to player then they are kind of useless for such an expensive tech.
If they aren't for high throughput, then what's the point of cargo capacity and robot speed research?
What's the reason for those at all, and why is robot research so expensive and late in the tech tree if bots are not supposed to replace belt?
I'd like to see from someone an explanation of an optimally fun™ factory where belts bots and trains are each used to achieve maximum fun™.
Sarcasm aside, I mean just explain in terms of an actual factory with circuit builds, smelters etc., what part bots belts and trains would play each.

Anyway, that said, I've been thinking about a way to nerf bots that isn't so transparent. (not that I want bots to be nerfed)
Some other posts gave me this idea by saying that bots should be limited in how many can be active in a network in some way.
Rather than simply adding a straight up cap that would clearly only be there so belts don't suck as much in comparison, I though it could be done by adding a new resource for logistics networks that I call "Urgent Processing Solution" or UPS for short. Each bot, smart inserter, whatever entity that interacts with a logistics network would consume some amount of UPS available in that network.
Since bots need to fly around, check which chest to go to, etc. they would consume more UPS than an inserter that just needs to check the chest it's inserting into.
Meanwhile, UPS would be provided by buildings within the network. For example, a roboport could generate 10 UPS, an inserter 1, an assembler 20, etc. (whatever, I'm making these numbers up as I go)
There could be bonuses for certain conditions, like a chain of assemblers direct inserting from one to the other, or penalties for machines that are not doing anything.
They goal would be to limit bots in a way that can still be completely negated through designing your base around it that doesn't devolve into just plonk down an array of UPS structures, or feed resources into science, like solar fields, or infinite bot research. The numbers could then be tweaked to force players to build fun™ bases. (couldn't resist)

I don't know that this would address whatever the devs really want, or that it would even be fun looking at it now that I've written it out.
But it was a bit different from most of the suggestions, so I wanted to post it anyway.