Friday Facts #224 - Bots versus belts

[IronCartographer]

We have ideas like increasing the power consumption, decreasing the maximum stack size bots can carry to 2 items or buffing belts by adding a "stacked" belt tier.

Making bots take more power = Build more solar.
Making bots carry less = Build more bots.

Both of these ideas run counter to the "no-tedium" philosophy, and offer little novelty. The only justification is a heavy-handed attempt to make large bases use belts.

We've had a similar discussion about nerfing before: Turret creep vs. turret warmup time. Improving combat rather than weakening turrets worked brilliantly.

It's almost always more satisfying to improve alternatives rather than nerfing things which seem overpowered. Hopefully this philosophy continues.

(Obviously the fluid nerf from last FFF runs counter to this in a way, but it was fairly even-handed, all things considered. The flexibility argument was a good one. Losing side-loading compression and the ability to generate compression in a space too small to fit a splitter, however. . . )

It would be entertaining if you had to build belt layouts which flying robots would somehow learn from, but that's probably too much abstraction for a game about designing factories rather than ushering in the Singularity...

[ZombieMooose]

I think bot based builds are super interesting, e.g. that screen shot in the FF itself is very intriguing to me.

[irbork]

I like bots as they are right now.
The idea of removing logistic system now seems ridiculous.
All the presented bot nerfing ideas will only reduce UPS punishing players for wanting to play how they like. Just make logistic bots cost more, lets say 10 red boards per craft, and problem solved without QoL punishment.
And BTW designing good logistic bot builds is as challenging as doing it with bots if not more so. There are lots of issues that must be taken into consideration to create efficient bot based factories and most of them are hard to diagnose, unlike belt based systems. This is where the robot based base challenge lies.

[Yehn]

Play any other sandbox game akin to this-- Subnautica, Minecraft, Starbound, No Man's Sky, etc-- and you'll notice one thing.

There's very little practical automation. You're expected to gather all your resources manually (maybe you can put down some kind of "check on it later" auto miner at best). You build everything manually, and by hand. Their 'Automation' is maybe rigging up some controls for lights and doors. Most game devs in general would have a heart attack at the thought of the number of things Factorio lets you auto, before bots even.

Factorio is a game about automation (duh). As you progress onwards, your capabilities to do things automatically drastically expand. Including blueprinting and, yes, those evil logistics bots. But as others have noted -- there's still plenty of challenges and problems to solve, namely around design and 'other' logistics. Yeah the power requirements might seem trivial to YOU, but to a newer player they're not -- and then getting all the resources to build those big solar fields or whatever, and then to keep bots from bottlenecking on recharging -- there's still stuff here to solve. IMO having options plays to Factorio's strengths.

A late game, large, belt only base is enough of a hardcore challenge that I don't think it is appropriate for the 'average' factorio player.

But, a strong point of belts over bots is (as someone mentioned) they're vastly better at moving high volume over distances. Bots have incredible throughput over very short distances (nerfing the stack size to 2 won't change this since bots can overlap).

I would be in favor of high capacity belts, especially available earlier (and cheaper?) than bots. In general: Play to belts' strengths, but let bots have their place. My $.02. Lengthening underground belts was already a nice step in this direction.

[sunlis]

I think there's a lot of assumptions about bots and belts being made in this FFF. I can appreciate that some players don't enjoy bots, but removing them from the game is a jump that doesn't make sense.

I also believe that building belts is way more fun due to it's inherent complexities, challenges, and emergent situations.

I have trouble reading these arguments and not jumping to a strawman such as "I believe digging up resources by hand is way more fun due to it's inherent time investment, challenges, and emergent situations."

While overcoming challenges is absolutely part of the game, I believe that part of the game is the player's ability to overcome the challenges involved in using belts. If anything, I'd love to see a future where the player becomes even more overpowered as they advance beyond launching rockets. Silly stuff like teleporters that can move items or players around the map. Asteroid mining. Orbital ion cannons (for those pesky forests).

In the end, Twinsen's argument reminds me a lot of how I feel about PUBG: I don't care for the game. I have a number of reasons to dislike it that I'm not going to list here, but the root of it is that I think it would be a waste of my time to play it. An important distinction, however, is that I'm not trying to stop other people from playing that game simply because I don't like it, because that wouldn't make any sense.

[Tekky]

Since bots are unlocked very late-game, I see nothing wrong with them being very powerful.

However, that should not make them better than belts in every respect. There should be late-game research making belts very powerful, too. One way to make belts more powerful would be to add loaders and belt sorters to vanilla.

[Peppe]

I would vote to remove logistics bots as well -- they never appealed to me like you say "win win win". I say swiss army knife solution. I only use them for player supply currently and that is a valuable role. If you have played modded minecraft I have the same opinion of the easy item, fluid, energy teleport era (tesseracts etc).

I think if you end up with a spaghetti base logistics bots smooth over the mess. But then the game becomes getting bots with any mess you like and then starting a bot base.

If you remove logistics bots base planning is more important. Maybe even trains move up in the player's toolkit to smooth over organization issues. Then you are using more of the games tools to solve problems, which is I think fun.

Another option might be to remove roboport logistics areas linking? Could see logistics bots working in small defined areas to help support max speed beaconed buildings (which being the UPS peak is the megabase design), but the logistics bots can't move out of their roboport's zone. Construction bots would deliver items across zones to the player / blueprints.
So in a small area logistics bots are the peak if item I/O, but to go any other distances belts and trains are the only option.

[Yehn]

Oh, another thought. There's already an achievement for not using logistics. So... I think this sort of thing is good? It encourages people to try out the belt only playstyle as is, without forcing it.

[eformo]

I like having choices.

I dislike it when someone else decides that I "ought" to like one style of play (belts or bots) over the other and then tries to force me to play that way.

Given the current arrangement of things:
If I think that belts are more fun than bots, I will build a factory that relies on belts. If I think bots are more fun, I will build a factory that relies on bots.

Given an arrangement of things where this choice is removed or the game is redesigned to try to prejudice decisions in one way or the other:
If I think that the favored method of moving stuff is more fun, I will build a factory that uses the favored method. If I think the non-favored method is more fun, I will stop playing Factorio and go do something else with my time instead.

Since that's the way I react to things, I would think it obvious from a designer's point of view that they ought to stick with the current arrangement of things. But I am not a designer, so I don't share the designers' sets of values. Some people are control freaks and I'm not likely to change that.

(An aside on how I actually do things: I use belts to take ores and process them into ingots (Bob's + Angel's refining and smelting mods) and then I use bots for most of the rest. There's WAY more than enough design time involved with this. I easily invest 50-60 hours of building new setups in each playthrough, so I don't have any desire to increase the amount of time spent designing belt layouts - and that's BEFORE I tweak ratios or optimize. This is heavily influenced however by the fact that I don't play Vanilla - haven't played Vanilla since ~v0.11. I wanted complexity and got it by using mods. Designing and building a 512 lane main bus (because of the quantity of products involved in modded Factorio) strikes me merely as tedious, not creative.

Bottom line: Since I've invested so much in modded factorio, rather than Vanilla, I want the basic game mechanics to stay stable, not keep changing. That was fine back in early Alpha. We're not there anymore. If I want to play the way you want me to play, I'll make that choice for my self and I'd appreciate you not making the choice for me. Please leave the basic game play alone at this point in development.)

[mozair]

> A universe where Factorio never had this feature. Construction bots were eventually added, and using bots to move items freely is nothing more than an idea that pops up from time to time but it's quickly discarded due to it breaking the game.

I wholeheartedly agree with the rant about logistics bots...

Actually the only part I disagree there is that it is "late game"... it is not that hard/slow at all to get into bots in a non marathon setting.

all in all, I do too think this would be a more interesting universe (the one from the quote) and I'd love to be pushed more towards trying to get overly creative with trains and belts instead. One may argue that this is just a matter of choosing to do so, but it is hard to choose something when it is so extremely sub optimal to do so.

as you said, it is so broken that it is hard to even nerf in any way that makes any difference:
- power nerfs can be 100% trivialized by solar panels/nuclear
- carrying capacity nerfs can be 100% trivialized by building more bots
- movement speed to some degree can also be mitigated by just building more
- making logistics areas smaller would just make people build more ports

what if bots didn't fly and/or required some sort of tracks? I suppose making them not fly would be super hard because of collision and all that, but maybe just requiring to build some tracks and/or other type of space demanding infrastructure for them to run along? i think something that occupies space could put enough pressure on designs to the point where it might become more viable to belt things in

[Monochrome]

If the problem is the use of excessive numbers of bots, then why not really increase their power needs? 5x, 10x, 20x even?

True mega-bases could still use them (and have definitely earned them by that point in the game), and a late-game smaller base would still be able to use a few bots to supplement their belts if they needed to.

Edit : numbers

[psiphre]

there are certain things that i don't think logistics bots moving around are "game breaking" - very high end stuff like processing units, all the things that go into the rocket, modules, and nuclear fuel and spent fuel are things that i would hate to lose the use of logistics bots for.

construction bots definitely need to stay in. maybe we could get them earlier based on engine units and then upgraded construction bots and logistics bots could use electric engines? fueling construction machinery with "the good stuff" (coal, oil) has a real world precedent.

maybe adding weight to items would make logistics bots not feel so OP. why should a little flying drone be able to carry a nuclear reactor and a circuit board with the same difficulty?

[Yehn]

Another option might be to remove roboport logistics areas linking? Could see logistics bots working in small defined areas to help support max speed beaconed buildings (which being the UPS peak is the megabase design), but the logistics bots can't move out of their roboport's zone. Construction bots would deliver items across zones to the player / blueprints.
So in a small area logistics bots are the peak if item I/O, but to go any other distances belts and trains are the only option.

This is already a strong use case for bots as is. Their throughput is extremely high over short distances, so if you have a smelter array fed by 16 blue belts, it's a good case for using logistics to get them into the smelter and then output the plates back onto belts.

What this nerf would really do is buff laser turrets even more. Ammo and other oddjob, low volume items over a distance would be painful.

[Zool]

Well, imho, you could simply add a couple of options to the „new game“ tab:
- Disable Logistic Bots
- Disable Construction Bots
- Disable Solar Power
- Disable Atomic Power

This way, people can simply choose the playstyle they prefer, and enforce it when creating a multiplayer map.

What I strongly disagree against is, to make any recipes LESS complex ... the situations where you really have to think about how to manage 3, 4 or even 5 belts between a set of machines is one of the most fun experiences in the game ... especially in lategame, many recipes are just „more of the same“, just with different items.

[Kyralessa]

We have ideas like increasing the power consumption, decreasing the maximum stack size bots can carry to 2 items or buffing belts by adding a "stacked" belt tier.

Increasing the power consumption? In my current 0.16.x game I have some 28,000 bots, and even though I'm still getting bot arcs showing I need more roboports, more than half my power consumption is just the roboports. That seems like plenty to me.

I think the revamped research where you have to wait a lot longer to get bots makes things well-balanced. You have to have red, green, blue, yellow, and purple science in order to get the full logistics system. That leaves the player with plenty of time for doing belt design until then, and gives plenty of opportunity to decide whether to go belts all the way, or breath a sigh of relief when bots finally become available.

Instead of nerfing bots, how about buffing belts? One thing I've noticed is that at a certain point, even blue belts are too slow. Maybe we need a green super-fast (and even more expensive and/or complex to build) belt.

[mozair]

I would vote to remove logistics bots as well -- they never appealed to me like you say "win win win". I say swiss army knife solution. I only use them for player supply currently and that is a valuable role. If you have played modded minecraft I have the same opinion of the easy item, fluid, energy teleport era (tesseracts etc).

I think if you end up with a spaghetti base logistics bots smooth over the mess. But then the game becomes getting bots with any mess you like and then starting a bot base.

If you remove logistics bots base planning is more important. Maybe even trains move up in the player's toolkit to smooth over organization issues. Then you are using more of the games tools to solve problems, which is I think fun.

Another option might be to remove roboport logistics areas linking? Could see logistics bots working in small defined areas to help support max speed beaconed buildings (which being the UPS peak is the megabase design), but the logistics bots can't move out of their roboport's zone. Construction bots would deliver items across zones to the player / blueprints.
So in a small area logistics bots are the peak if item I/O, but to go any other distances belts and trains are the only option.

really like the idea of removing the port linking

[PhatzDomino]

Um is it April 1st already? Please don't nerf or remove logistic bots, if you really must make the research requirements higher. For me, it would completely kill end game megabase building.

[Neemys]

Bots are boring.

It is satisfying (and fun) to make belt based production chain, even more when you try some specific thing like filling a blue belt with assembler, beacon and modules (speed+prod). Using bot is only a matter of laying roboport, assembler+request/passive storage and having a good bot production and power.

It doesn't mean I don't use them, they have a niche use in most of my bases, very low throughput. When I have some assembler that require material in very little quantity I use bot to send them material. Belt everywhere is good, but I don't think having a single belt for 1 item every 10 minute is really mandatory.

I could live in universe two like I don't care that mine is universe one. I play my own way, other people play theirs. Some people are indeed having fun in bot bases.

One solution I could see is having Three type of bot, construction like now, ligthweight logistic, fast, only able to provide and take from players with max item per robot set to 2. And bulk logistic for the base, very VERY slow but can transport 5-10 items. I know slow will mean more and player will do that without doubt. But I do think it could be interesting.

We may not live in unverse two but we could implement their bot system as a scenario ?

[Kaufcraft]

Hello,

I would like to comment on that although I am not a pro gamer and have no significant experience (meaning no idea) about the late game (I got a rocket once in an earlier version).

But I always thought that those logistic bots feel kind of a late game replacement of belts. And I mostly used bots just for building (construction bots) and for refilling myself. And also for doing some odds and ends (dont know if you get the meaning, Im not native english, I mean "Kleinkram" (german)).

Thus, I kind of like the suggestions made in the FFF. But maybe one could consider making bots fermionic (maybe you have a physicist in the team). At the moment the robots can stack apparently infinitiveliy, you can have arbitrarily many robots in a certain area. If you change it such that only one robot can be on a certain amount of area (a tile) maybe this would restrict the use of logistic bots to use cases where you want to transport only a low amount of items like transporting some rocket parts. But of course I cannot estimate performance implications or other problems. For example player refilling should stay fast. But maybe that suggestion has already been discussed and rejected.

[BLuehasia]

belts just can not compete with the throughput bots can do. bots just can bring items faster (with 100s of bots)

if i were to build a belt only base i would want means to transport items faster and more item types to locations. also insertions would need to be faster