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

[Avezo]

While limiting bot access to chests might sound good when comapred to limited access chest have with belts, it will create even worse monstrocities like these below (just ideas, multiply it by 100x when smart people math them out):




Once again - there is no issue with bots being too strong. There is issue with belts being too weak. In this case, instead of nerfing bot access to chests, find a way to buff belt access to them (psst, pssssst... Loaders...).

[Marconos]

I have already replied once and the more and more I think about this I really think nerfing bots is the wrong thing to do. Either buff belts or leave things alone. There isn't a problem to fix IMO. I have been playing since version 0.9 and was always pushing against bots, heck my last play through I finished last night for the Lazy B achievement I never made a single bot. You are running a HUGE risk with your fane base by altering something this integral to the system this late in the game.

I don't know, but whatever you do, be careful. I would really hate to see Factorio with it's overwhelming praise and fan support hurt itself this close to actual release.

[Avezo]

Other matters aside, I hope devs addres the idea of beacon removal in upcoming FFF.

[Syriusz]

Splitter configuration might be useful, but I already use something better- loaders (and especially with merging chests mod). With them you can do all of this, on little bigger space, but without another gui and you can see it clearly. I fear that with those splitters, you will find yourself questioning: why does those belts/ blueprints work funny? With loaders and merged chest there is no such problem since you can clearly see what's going on.

I usually play on maraton, so I need huge throughput but I use bots mostly for tedious tasks, not on everything. They are too expensive to use them in maraton on huge scale, at least until it's very late game. But I have huge problems with belts throughput and loading to machines (even with loaders!). Biggest problem is with electronic circuits- I use shared chest (about 20 tiles high) that feeds copper and takes coils from the assemblers. Then I transfer those coils into assemblers that make circuits. And this is most difficult, since I need several belts with coils to feed one machine. And this is the problem- with robots and only inserters you would always be better, cleaner and etc. With loaders and merging chests you have to work hard, but you can find solutions that are economically better than robots. But you cannot do this in vanilla.
And I always felt like loaders should be base for loading, and inserters should be used only in specialised places, they should be more expensive than loaders (but be configurable like in bobs mods).

So, implement loaders and make them stack items on belts? And maybe merging chests?

[Ahry]

I'm not sure how would you make belts more space efficient other than totally getting rid of inserters in belt setups and embedding a direct loading from the belts into containers and assemblers.

The devs proposed upgrading all belts with one research into stack belts as the ir final option that they like the most.

This would affect throughput. Space efficiency is how much space you need to build a production setup. Now the more complex the recipe the more complex and large the belt setup is (unlike the bot based setup which is always 2 chests, two inserters, and a power pole per an assembler). Beacons make this even worse.

I posted this on Space Efficiency underneath that:

Space Efficiency? You should count the neccessary space and resources to build needed Solar Fields and Accumulators or Nuclear Mining&Refining&PowerGeneration into the such an equation too. That is easy to change too. One or multiple of these options: Increase power demand. Nerf the Kovarex Process. Make Solar panels much bigger and requiring more space per panel. Or you could also decrease the roboport logistic radius like you proposed (which hamper Beacon Set Ups). Or increase the size of the robotport.

This is wrong metrics then. Because no one take care in late game about space in his huge mega base. There are plenty of space where to put solar panels, nuclear reactors and so on. So you can not calculate that. What matter is if i can build assemblers with theirs need easily in that place. You can put solar almost everywhere... but when you want to build green circuit you want it on place, where you use it, or in only one place where you will puck up it.


But i do not know why belts and bots can not have same space effeciency? i think it is ok, if belt have much higher throughput (for ex. with loader or packed items).

[Hammerchief]

Those new splitter options looks amazing, I always wanted something like that. Thank you!
I never liked the logistic robots, it seems to me that they change whole logistic game of getting reagents in and products out of assemblers to very boring "make enough bots and they will take care of it". But then again, I can very well play without them. Just maybe option when starting the new game to allow or disable logistic network would be nice.

[unobtanium]

Concerning the splitter buff: What happens if i want to sort iron plate to the right side, but the right belt is backed up. Do the iron plate go down the left side then?

[Lastmerlin]

This is one of the few FFF threads ever, that leaves me downright angry. Hence the comment, although I somewhat doubt if will get many reads so far down the thread...

I have played about 20 serious maps with vastly different styles and (self-chosen) objectives. Only one of these maps was pure-bot at the end (4 rpm megabase in 0.14) and 5-7 other used bots to a different extents (including a 1rpm mainly belt-based, 1.5k spm, mainly bot-based), the rest was pure belt. My personal conclusion concerning the *overpoweredness* of bots is: They are only better in each regard when coming to the point, where construction costs and power are irrelevant and you have quite a lot of the upgrade technologies. In other words: When reaching (or at least approaching) megabase status. For most other objectives, the fastest and most direct path is to skip them entirely (e.g. no speedrunner uses bots).

So the whole discussion secretly revolves around megascale production. When asking *is this mass-bot playstyle fun* you are effectively asking, whether megabase-building is fun. When discussing nerfs, you are essentially discussing how much to screw with megabases.

In this threads dozens of people propose different ways to nerf bots. I ask myself, how many of them ever built a megabase themself. Somehow I suspect, this thread is about many players who do not need a feature to its full extent discussing how to make life harder for those who need it. This is what makes me angy. I am quite sure, the majority of the players have no idea about the topic of mega-base construction. Thats ok. But then they should stay silent about screwing with this specific playstyle.

As I did build some larger basis, I feel confident giving my opinion/experience: First of all, building these larger bases was fun. It took more time than the other maps (so much for trivial) and gave me lots of new problems to solve. The problems that bots allowed to skip, were those I had already solved several times and was not keen on solving again. The single modular production cells were rather boring to build. However, they would be equally boring on belt-basis. For me, they are just black boxes with train input/outputs on this level of play. You mainly design the choice, placement, number, interaction of these black boxes. I am not even super keen on using bots inside them. If you can find an equally UPS efficient way with belts, it would be cool. But just making these black boxes worse in terms of throughput/UPS-efficiency without replacement actually reduces the type of challenge you are seeking here.

Last comment: The new splitters are really cool

[Ancalagon]

How about attaching an operational cost (other than electricity) to using bots instead?
I didn't fully work the idea out yet, just throwing it out there.
Perhaps some system of inserting fresh batteries in "used up" robots? Maybe the roboport needs to be fed with batteries in order to recharge the bots or something.

This keeps the usability of robots the same, but requires resources to operate. Whereas belts have the advantage of being free once you place them.

[voddan]

With bots, there is no reason to think about other types of transport (Cable cars or automated vehicles for example)

Well I think we should do just that!
With small modifications logistic bots can be transformed into Cable Cars, saving their good parts while fixing the game balance.
I wrote it up here Improving Logistic Bots - Cable Roads
What do you think?

[sathill]

For me bots, belts and trains are just fine as they are.
Now im trying no logistic bot achievements and its pain to produce some items.

Its not true for me "more bots - solve everything".
I make some very nice vanilla megabase in 0.15 and in every single free space i was placing roboport, becase of limited charge ports. There was 35k+ bots flying but most of them just wait for charge.

Best thing is always to mix bolts/belts/trains.
Balance its fine. Just dont add unlimited research "add +1 charge spot to roboport" without adding unlimited research "belt moving speed +%"

Btw. Great game, insane amount of dev activity (updates).
You guys should thing about adding some way of donate.

[Avezo]

I may be chiming other people now, but no matter the game, no matter the balance, the very top/last/endgame in any game will be dominated by a single choice of all possible choices, even if it's better just by 0.0000001% than other choices. Therefore, balancing around such engame is bad for majority of the rest of playerbase who don't participate in it.

Balance midgame instead, where most of us are.

[blueblue]

Thanks Shados for the superb summary post. I agree that the issue would be solved by assigning belts and bots their own niches where they excell. Instead of mid-distance and low-distance transport I suggest high-throughput and low-throughput as those niches. I am very much in favor of the options S11, S13 and S15 in the post summary. Each of them impacts a different part of the game and together they might balance out bots in mid and late game.

S11(Increase bot total power consumption). One of the disadvantages of bots even now compared to belts is their enormous power cost. If you want to build a bot-only base even now, you will need a good nuclear reactor, which is a lot further down the line than the logistic chests research itself. Alternatively an enormous amount of solar panels or steam engines would work, both of which would push your evolution up enormously, in fact you might be very close to behemoths when you finally have a bot-only base, depending on map settings. With this in mind, increasing the electricity cost just a bit more and teaching the player that bots have a large electricity cost would deter from using logistic bots excessively, so players would use them only for low-throughput items in the mid-game. As an example, in my recent default options game I got logistic chests in five hours, but only had a fully complete logistic bot base after some 20 hours, which was way above my estimate, because I couldnt provide the 800 MW the base wanted. I ended up building a 4x and a 10x reactor and by the time I had no more belts in my base, my evolution was at 83% (with steam power only or solar only this number would be higher still).

S15 (Closed high-speed belt). This suggestion would add interesting gameplay options, it might help belt performance and it would certainly help the incredible number of iron belts that belt bases tend to have in the late game. The issue I see here is that we already have three belt tiers, adding a fourth one would be too much. This could be offset by removing one of the other belt tiers. This has its own issues since all three existing tiers have a place in the game as it is, but then again this might be offset by clever rebalancing. For example give yellow a 20/s throughput instead, red 40/s and make blue the closed version with 120/s. The game does what it can to avoid multi-tiered entities where the only thing that changes are numbers, for example each inserter type has a unique feature, and most weapons have some unique shooting mechanic. A closed belt tier would fit this paradigm better than three different tiers of belts which only differ by speed. The other issue I see with this is that new player might never have the need to learn how to manage more than a single belt of iron. Which means that this should be gated behind space-science. This change would hopefully make bots unable to compete with belts for high throughput items.

S13 (Nerf power production). In the very late game, concerning mega bases, solar power being UPS-free is an issue. A related issue is that solar power is trivial to set up, you just stamp down a blueprint, you dont even need to supply any ingredients after it is set up. Considering how late you get them (they are not that deep in the tech tree but they also cost a large amount of resources compared to the power you get so you'll get them late), and that these are literally two completely passive entities, this is an issue. This could be fixed by reworking the way solar panels or accumulators work, for example by replacing accumulators with electric boilers or adding some maintenance cost to solar panels or accumulators. This should fix bots for mega base builders and fix a weird imbalance in the game.

In general I'm on board with many of the QOL-improvements for belts that have been suggested, like automatic compression (which was all but confirmed in the fff). One of the reasons why bots seem better currently is that doing anything with belts instead of bots is so much more effort, like routing belts, unloading trains on belts, balancing belts. I would like to see improvements to this in the form of QOL-features.

I'm not going into changes to bot mechanics, for my taste they are complicated enough as is and every addition needs to be taught to new players so it should really be worth it. That is not the same as saying bot mechanic changes won't work but it's hard to find one that can be taught intuitively in addition to all the complexity bots already have, and that would push bots into a niche compared to belts.

Here are some more comments.

S2 (limited speed of bot un/loading): This sounds like a good idea, but as has been posted above, this would lead to ugly setups where players simply build lots and lots of logistic chests. It appears more tedious than challenging.

S7 (belt and inserter speed research): Increasing the speed of something too much tends to have negative effects on graphics, gameplay and performance. In this case, additionally, the research would take some time to kick in, so bots would still be good until then, and mega bases would probably still use bots, because performance. Plus I'm not sure this would make the game more fun (beyond the initial "omg so fast!").

S8 (make beaconed belt builds easier): This is an interesting point. Most items can be made with the way beacons work currently, see for example blue circuits. It does however require considerably more effort than using bot designs and for no payoff except the decreased power cost. Again, if power were less trivial to obtain, then this might even be fine. I like that beacons currently add a bit more spaghetti to the game and I'd like to see this preserved but this issue needs to be adressed.

[diederTheBeater]

I think a switcher that flips left and right part of the belt would be very usefull in many convooluted cases. That would make the belts less tedious.

[anarcobra]

A lot of text

I was almost entertaining the thought of writing a summary. Thankfully you did it before I had a chance to be too lazy not to do it.

[wren6991]

Re: Belt buffing.

Multi-level belts, with a flip piece?
Functions like a braided belt, except it has n times higher throughput than an express belt, where n is the number of layers, and does not require repeated undergrounds getting in the way of your layout. Flip piece is still 2 tiles long like undergrounds, to not make layout too easy, or maybe even longer/wider. Velocity would be the same as express belt, so no need to buff inserters.

Each flip piece cycles to the next layer of the belt, which is a requirement for all flip pieces being identical.. You can put different produce on different belt layers to save space for complex recipes, like with current braiding, except legit.

Costs n express belts, some low density structure to support the layers, and some more lube? Idk it seems like it would be a fun idea, and doesn't suffer from the problems kovarex mentioned in FFF. Could perhaps triple or quadruple belt throughput.

[SirLANsalot]

Bots have always been good for short range transport, Ergo, your base. They suck at med to long range for transport, as such, even in the late game Belts are still good (and modded belts stay relevant vs bots even (aka Bobs or 5dim's belts)). Bots are good for inside the base, but as the base grows even they start to not be as efficient, like when trying to get something from one side to the other in a large base.

Belts are good for moving stuff TO the base that is at at a medium distance, like an ore field just outside of your walls. Trains are perfect for long distance as we all know, like that ore field half way across the explored map.

I would be for a belt buff, but not a bot nerf. Bots are powerful, but not perfect for everything in a late game base.

[ZombieMooose]

Bots should replace belts. They should be the evolution of logistics. Beyond our current techonlogy into the realm of sci-fi and overall better setups.

They made several combat options OP for a sense of power and triviality over the native populations, nuclear is vastly superior to every other option in terms of power. Why not a logistics version of this power?

[Avezo]

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

I feel weird throwing my somewhat random thoughts into these threads like that.

[Venatos]

bots can absolutely handle a few nerfs here and there. not realy sure about the chargetime, might get annoying with personal robots. maybe cut down the range, making em suck at longrange transporting would give belts the edge in that category.
i would prefer a balanced nerfing, cut some range, increase chargetime somewhat, only have 1 cargosize upgrade(mid-lategame). should bring em more in line while keeping em somewhat overpowered.