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

[franqly]

A simple and effective nerf to bots would be to limit the amount of roboports. Give the roboports and minimum distance they must be placed apart. Less roboports => less throughput, added more bots after a cap wouldn't add sustained throughput because the roboports wouldn't be able to keep up.

[Neovik]

I'm a lover of belt based bases, but I think there is something missing here in the way the devs are viewing the game.

Early Game: Yellow belts, small production runs
Early Mid Game: Red belts, faster production lines, multiple smelting setups: Base is starting to spaghetti around for most players
Mid Game: Main lines are being upgrade with blue belts for increasting thorughput. Bots are being used to move product from one area to another as the spaghetti in early mid keeps you squeezing in yet another set of assemblers. Things are starting to get nuts but the end is in site (1st rocket end). Basic / simple train setups are starting to be used to move raw materials around.
Late mid game; 1st rocket is launched players are now looking at doing the advanced research and seeing the need for big increases in resource comsumption. More trains are beign laid out and now the "belt layout" issues are changing into train layout issues. Bots are preferred as you don't want to spend time optimizing belt lines as you are having to optimize train lines.
End / VERY late game: Mega bases are growing, belts are relegated to very small areas. Bots have take over local production and trains are the new "belts" in the system as most of your time is figuring out how do I increase my train line throughput, do I do extra production in this area with bots or seperate areas connected by train lines. Spreadsheets are made to figure out optimum build configuration etc.

So yes, belts are decreased, limited (dropped in some cases) in late game but the train lines becomes the belt layout issue. Bots handle the local production but your massive product movements are done by Trains(belts) as they are end game evolution of them. To me this is the real factorio progression and you see this all the time in how players play the game. Most players never get to the "end / late" game stage and are done after their first rocket. By nerfing bots, no matter how much I initially thought I liked the idea, is actually something that will negatively affect the progress of the game IMO.

Edit: text cleanup

i completely agree with this

I agree with this too
i play in my bases with both, i like the mix as it is now
and to be honest in end game you use blueprints so you have already your preffered strategy in the blueprint and just put one blueprint beside the other, so i don't see that if you nerfe bots someone would take more time to make spaghettie belts in endgame
and want to add that game play is always about progression, you want to become better so what is the end game about ? you added infinitive research to not only send rockets in space but to only progress further in the Research Tree and there is a infinitive research for Bots but not Belts
why not add something like cable cars or containers on belts and add them also to an infinitive Research to give the Player the choice to buff their way to play in the End game ?

so please don't nerf bots because we player put much effort in our game to get them, but give us new choices like a new transport system for midrange so it would be: bots short range, new system midrange and tains long range
and new possibilities for infinitve researches to buff all of them in endgame

[mavu]

Hi

I registered on this forum after playing factorio every couple of months for 2 years (?) just for this topic. That's how important this feels to me.

tldr first:
Nerf LOGISTICS bots. keep construction bots.
Don''t buff belts. (much)


Long version:
To me, there are several milestones when I start a new factorio base, and as I'm writing this from memory, I might miss some, but then again this could make it more accurate, because a forgotten milestone can't be that important.

1) automated smelting and the start of a main bus.
This is very obvious, but this point is the actual beginning of the game for me.
2) Electricity
Obvious again, but definitely feels like a milestone
3) first automated production of items for player use (make-everything v1)
this hugely speeds up the game/factory expansion.

Now something interesting happens:
These milestones are relatively close together (in time) and the game enteres a different phase.
Now I am off the rails, and have to decide what to work towards
This decision is invariably construction bots which is milestone

4) construction bots and personal roboport
again a huge speedup of the game and factory expansion.
Construction bot enable me to:
- easily try different designs of production lines
- quickly rebuild large structures
- clear terrain of obstacles (trees...)
Basically allow me to focus on the bigger picture, instead of manually building every small detail of my base.

5) solar power/accumulators/laser turrets/walls
this is the next big thing for me. At this point I wall off my factory and expand by walling off large roboport-sized chunks of land and securing it with walls/turrets.
The milestone here is having room to build without worrying too much about defenses and power.

6) logistics robots.
Now, this is a strange point for me.
I have a love/hate relationship with logistics bots.
- I am aware that they are the best way to do things.
- They enable very efficient designs not possible without.
- I love to use them for transporting one-off things. For example consolidating materials to one place after rebuilding a part or parts of your factory.
- Transporting very low volume, high cost items to a central place where you store them for use, or supply them to where they are needed.
- Supplying me through logistics slots in inventory (awesome)
- keeping ammo in turrets

I hate them for large volume continuous production (i.e. magic-air-belts)
- This for me breaks the game and takes all challenge out of it, except for the "How do I build more robots so that I can build more robots" challenge.

Of course I don't have the discipline to completely avoid that, and that is usually where I stop playing for a couple of months, until a new patch brings something new to the game.

Logistics bots as belt replacements break the game:
- you don't have to plan ahead.
bots fly, so you don;t need to make room for belts or plan the layout of your base in any but the most basic way.
- expansion is trivial.
need more X? plop down your blueprint an done.
no need for input/output planing, resource routing etc. bots do it.

in short, the game becomes boring and I stop playing.

In my opinion, the perfect solution would be something that enables logistics bots to fulfill the small volume transport and player supply role. but nothing else.

Unfortunately I have no idea how one would achieve that, but that's your job

In closing:
I have no idea if or how much my style of playing / opinions align with the majority of players, but this one factorio player would very much appreciate a big fat nerf to logistics bots.

[mavu]

I believe the real problem with bots, their game breaking ability, is their inventory access speed. Think about it, if I have a chest which needs to be filled with resources, there is a hard limit on how quickly I can supply items to it using belts. Intuitively that limit is the number of inserters I can place around it times the inserters throughput. Bots do not have any such limit. No matter how slowly bots move, or how few items they carry they will always be able to transport items faster than belts given that I have enough bots.

I think chests should only be accessible by one bot at a time, and this access should take a fixed amount of time. This would cause bots to queue up at chests, and put a limit on their throughput. This kind of limitation "makes sense" to me, and wouldn't feel annoying like lengthening bot charging times would.
Bots would still be incredibly handy for the low volume tasks and "housework" which can be anoying to do using belts, but it would create a compelling reason to invest the time into developing a really impressive belt base.

I also think it would be important for bots to still be able to access a players trash slots, or probably their entire inventory without queuing up.

This is a very good observation. that could be a good solution.

[Aflixion]

These last 2 FFFs have shaken my faith in Wube as developers. It seems you went into this with the subjective opinion that belts are more fun than bots and with the decision already made that you want to limit bots in some way. Then you got the wall of feedback to last week's FFF. You listed the suggestions from that thread in today's FFF and explained why each of them wouldn't work, then presented your plan to nerf bots anyway, effectively ignoring most of the feedback you got that bots are fine as they are.

I don't see why there's even a "versus" in this issue. Bots are the logical next step in terms of intra-factory logistics (for a single module at least). You start with slow belts, upgrade to faster belts in the mid game, then upgrade to bots in the late game and keep improving them through research. Lots of other people here have pointed out how much investment is required to unlock bots and get them to a usable state, and how much continued cost they require through power consumption. Any change you make now to limit bot throughput will only have the effect of reducing the possible size of megabases in the future. To address the specific changes mentioned in the conclusion: removing cargo size research will quadruple the required number of bots to maintain throughput (lost UPS) and increasing charging time by several times will increase the required number of roboports to maintain throughput (more lost UPS). Neither change does anything other than reducing the production capacity of a 60 UPS megabase, and for no reason other than you don't find that playstyle fun. Please listen to the community and to some of your own devs like Rseding91 and leave the bots where they are.

I personally feel that if you push through with nerfing bots like you seem set to do, then I'll probably have no further interest in playing Factorio. Not because of the specific change you propose, but because you seem to be making it by framing your personal opinion as objective fact. That makes me hesitant to trust your ability to make good decisions for the game in the future.

[lyallp]

Just a note, it's been a little while since I played, so maybe my comments, below, are no longer relevant.

I think that bots should be nerfed such that only a limited number of bots can access a chest or construction item at one time.
It's completely unrealistic that 100+ bots can all get stuff from a chest or perform construction in the same instant.
This not only makes them more realistic but brings things back in line with belts, you can't get dozens of inserters processing against a single chest.
Limit bots to the number of inserters could do, when it comes to pickup/dropoff/construction. Make them queue up and take a few seconds to load/unload/construct.

[ShermyPermy]

I think that it would be a good idea to leave the bots as they are and leave the belts as they are also.
To balance the bots/belts issue i would say that your are best off to give the fabricators a boost.

What i mean with that is make the fabricators use wooden crates/palets to create a stack of say 10 items before the inserter picks it out off the fabricator.
this way you just gave the belts a 10 times boost in trueput.
Use wood as a base product to create these crates/palets en use bots to deliver them to the fabricator without the need off a insterter or chest.
So the bot is inserting the created crate/palet right into the fabricator.
the fabricator uses these crates/palets to increase the output of 1 fabricator by almost 10 in amount.
If there should be a speed increase im not sure. i guess that is in need off testing.

As a final point i would suggest that every fabricator can use a single item or a crated/paletet item without the need to unpack them seperate in a different fabricator.
in a perfect world it should even be posible to use single items and crated/paletet items on the same belt without the intake fabricator careing about the item.
so it doesnt matter if the fabricator picks a single item or a crate/palet off the item it can use it in both instances with the only difference that the crate/palet is 10 times the single item.
(extra point to consider is to destroy the palet/crate after use or to reuse them a few times)

I think that in this way you are making the belts the high output so that way end play bases have a need to use them.
plus your connecting the bots and belts play together witch should lead to a more differce base all together.

added bonus is that you can get that pesky wood a reason to be used in late game builds.

[Confuzz]

A bit off the wall, bear with me here - but what if logistic bots projected a small field around them that disables or slows down certain surrounding mechanisms similar to how an efficiency module does?

For example, disable all functions of stack inserters and slow down assembler speed whilst in the field of a flying bot. Toggling the field based upon the load of the bot would allow finer tuning should the fields be too powerful with permanent uptime. Belt builds would be immune to this debuff (seeing as there are no slowbots flying around!), shifting a bit of power away from the bots and incentivizing belt use more. Thoughts?

[Hakisak]

I'm all for the belt-rebellion but when you nerf the bots with charge time it might deter newer players going into bots when they first start a small bot network, another suggestion would be to scale the nerf, like, the more bots you have the more power each roboport needs to keep them going in its area (the roboport area).

[scali]

So in order to make belts better than bots, you may need to limit bots. Bots and belts fill the same niche at the moment, so what if you found a new niche for logistic bots? By limiting what they can carry and how efficient they are, you can effectively limit their usefulness in a way that enforces that belts need to be used for cutting-edge content, but you can then follow it up with bots afterward.

For example, if you added a Basic Roboport, which supports a small number of extremely limited Ground Logistic Bots (think like a roomba with an inserter and a chest on top) that can move a limited type of good from several gameplay progression steps ago. When you unlock them, they are slow, clumsy, and can only carry raw resources (like non-uranium ore, wood, and coal). Around when you unlock Science Pack 3, perhaps they can carry early intermediate goods like metal plates. Additional techs to allow delicate goods (circuits), bulky goods (belts, trains), petroleum materials (plastic, sulfur), petroleum-gated goods (electric engines, red/blue circuits), uranium ore, radioactive goods would allow players to need to improve bots over time in order to harness their usefulness.

This would also need to be combined with the idea that where the current tech tree has "Logistic Robots" is where you'd unlock the ability to make flying logistic robots (out of a flying robot frame, a ground logistic bot, and advanced circuits), which is its own payoff, but with this framework, a player would only be able to use them to move ore and plates by that point until they unlocked the ability to move more things.

[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

[GotLag]

Any bot nerf can be un-nerfed with mods if you're that concerned.

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

You can do that with mods already, but it would be nice to see in vanilla.

[Sephty]

he big difference between Bots and the Death spell is, that in Baldur's gate, Death spell only works on the low/middle level monsters

A Baldur's Gate way of doing thing would be adding "weight" to items in the game and adding a max "weight" capacity to the bots ?
- bot will not be abble to pick of item that overload them.
- more weight, less move speed, more energy consumption.
- late game research to increase that bot limit and let player that want it use the bot like before ?

But maybe the issue about the bot are not their speed or item movement capacity but the fact that they are too smart ?
Train are very powerfull, but this is not a problem because you have to think ahead to use them properly and to solve a lot of issue too. Bot do all of that by themself, just drop a robotport and a bunch of robot and voilà !
Find a way to force the player do the "logistics network" stuff !
- 10 robot should take things that are in that chest but only if copper > 50k, they should move that there.
- Ok now I have a "bitter" issue, I should define a zone where I send construction robot (or you ICanDoEveryThingNowBitch Robot) to repair the turrer.
- NIce that work, but the Robot are focus by the "bitter", I have to send 10 logistic Robot to feed that robot port with construction robot to have at least 10 in that logistics network...
- etc.

[rldml]

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:

[Blazku]

Please just leave the bots alone, make belts stronger if you really have to. I think you (the devs) are trying to fix a "problem" that doesn't need to be fixed and is not a real problem at all.

[js1]

I believe the real problem with bots, their game breaking ability, is their inventory access speed.

This is a very good observation. that could be a good solution.

I agree with this, and it wouldn't be difficult to create a queue per each logistic chest for bot access. Then each bot would be penalized by some time delay. They could add this to game with the time delay set to 0, and players would see no visible change, and you could then have mods that would modify the delay (and balance belts to better than bots). Or vice versa, they could add a delay and if people really wanted a bot megabase, they could modify the delay by setting it 0, getting the old behavior.

[Duny]

I don't usually post since I mostly always agree with the devs, but I don't want this thread to leave the impression that only a few people agree with them just because the people who disagree are much more likely to express their disapproval.

Bots are too powerful for the amount of thought put into using them. I can't fathom how anyone could ever honestly disagree with that statement. They're also simply too powerful compared to everything else in the game. All logistics challenges are instantly eliminated as soon as you get bots and either a decent amount of solars or a nuclear reactor. Logistics challenges are what makes factorio a real game instead of a clicker game in which the player just watches their production increase with minimal input from them.

The easy way out should never be the optimal way out.

Personally I only use requesters to deliver fuel, repair packs, or ammo. I force myself to not use them for other things because they are too ridiculously strong and most importantly using them removes any need to think. If you're not producing enough of something you just plop down more assemblers with a requester and a provider and you're done. Maybe add more bots and plop down another solar cell/nuclear reactor. If you need something sorted, well bots will just do it naturally. Also makes for uninteresting factories of copy/pasted assembler setups, as said in the FFF. There is nothing impressive or interesting about a bot-based base, no matter how big since making it bigger is always just the same simple process.

I think logistics bots should be much more limited in some way (only one bot being able to access a given inventory at a time would be a start and logical, and/or making it so bots can only carry one or two items at a time; however an increase to charge time would require smarter bots, they are extremely dumb when it comes to choosing where to recharge and it's already infuriating as it is when using construction bots to build/deconstruct something large).
Using bots for punctual deliveries like train fuel, or player resupply, is perfectly fine but using them for everything should not be stronger than every other means of transport even though they are the simplest to use by far.

Just make it so bots can be modded back to being their original ridiculous selves and that's it.
I don't see why balance should be decided by the people who like to use the imbalanced feature, and not the developers, when said people are free to use a simple mod to make it imbalanced again anyway.
Just like people used a mod when the logistics research was made harder to obtain to get the easy one back, they can do it again if bots are made balanced.
When I read this thread, I have the impression everyone who complains is assuming logistics bots are going to be completely removed and made impossible to mod back in. That's obviously not true.
This is not a MMO, a balance patch doesn't mean your OP class is ruined because it's all server-side, you can cheat all you want in factorio just by downloading a zip.


In short, the last two FFFs have strengthened my faith in Wube as developers. Too many games have fallen into the trap of letting the community persuade them out of fixing imbalances, and that really shouldn't happen in a game where any individual can mod the imbalance back in very easily if they so choose. Don't fall for it. And I say all that as someone who was pissed about fluid wagons becoming better than barrels (though in that case, complexity was nerfed in favor of simplicity, so it's the opposite of the logistics bots situation).
Keep up the good work, you're the best devs I can think of and it will take much more than reasonable opinions about the balance of your own game to change my mind. Remember there are lurkers who don't bother posting because they either agree or don't care that much.

Now fix mapgen, I want to be able to use bigger biomes god dammit they're too tiny and chaotic.
Allowing for multiple starting areas and separate research without being enemies/neutral would also be nice. There is no way to play with separate research and still see each other/be able to use each others doors and trains (or even trade without using some clunky mod which might not be very ups-friendly). And no simple way to have multiple starting areas (I'm aware RSO has a feature like that but it's clunky as fuck).

[gerren]

Hi there!

This post made me think:

• Let the bots handle only subset of products - eg. let them handle only advanced stuff, not the basic materials (ore, plates, basic circuits, gears, and so on)
• Another subset could be based on production statistics of average rocket launching base - everything with items/minute > X would be banned, only things "not belt-worthy", as in "it makes no sense for 6 items/minute reserve a belt" -> just robot it
• Make two tiers of logistic robots - one "light" for ordinary small stuff, one "heavy" for bulk shipments (ores, plates,...), that would be much slower, possibly with bigger capacity, but overall smaller, just to nerf things right
• Ban logistic robots to haul machines - assemblers, pumpjacks, engines - generally things too heavy for a teeny tiny robot to carry - this would make things weird, as it feels the robots were made just for these cases

[equitime77]

Right so as a player since 0.11ish this is my thought. Bots are extremely useful for short term transport, ie getting stuff off a train and onto various belts. Also for sorting a few belts into many. (Yes there is a proper term for this but atm I have womanflu lol). Ie in places where to split belts would leave a huge mess and take up so much unneeded space. Belts just cannot keep up with bots as it is.

You say that you dont want to have stacked items on a belt for certain reasons but my response to this is that your thoughts are rubbish. You are taking away the creativeness/options away from the player. What if a player wants more from their belts but has to use bots because of throughput. I seriously suggest that you think again.

Reading through the first few pages there are some great ideas given that you should listen to and implement. Especially loaders and that mid range transport system a few players have suggested.

Factorio is a great game because there is many ways of getting from A to B. Dont limit that because of how you think players should be playing. Let them be happy how they are playing but give them more options of how to game....

[gerren]

AND the other option of course is: make a bot-nerfed option as a new standard, make it a challenge, and let the mega-super-duper-boosted-bots as an option or as a legacy mode in a starting option, and say that the players just do not have it to play it the proper way - they'll bite, believe me