In support of twinsen and open/polite debate

[quadrox]

Trying not to start a flamewar or an instant lock, but for me this is important to be said.

I would prefer to live in a society (or be part of a community) where reading carefully and being open to new/different ideas is the default (instead of always being afraid of saying something wrong), and where it is possible to express even controversial ideas without being harassed. Twinsens text was written very carefully to say that that bots would not be removed, but that this was a hypothetical scenario. Yes he used some clever rhetoric to stirr some emotions (definitely including mine), but he was very careful to never be deceptive or misleading. Go back and read the FFF again if you disagree, you will see I am right.

The point of this post is not to shame those who reacted negatively - and I would like to ask everyone who replies here not to attack either side of the debate. Instead, the point is to ask everyone to be more considerate and open in the future. If somebody writes something that seems wrong or offensive to you, make sure you really have understood their intent before reacting. For example, begin by asking the author to confirm your interpretation of what was written instead of assuming you understood it immediately.

I realize this problem can't ever be truly solved, but if I can convince just one person to be more open and considerate in the future I consider this topic a success.

EDIT: And even if your first impression about a given post is right, responding with attacks or angry rants will never make others agree with you, so just leave it be. Either write a post with polite arguments for your viewpoints or just stay out of it completely (to be clear - this is just a request, not a command, as I have no authority over you).

[Engimage]

That FF was a nice lesson for both sides and I hope well learned

It requires a lot of wisdom not to fall in mental traps and start to spread misleading information and start flames. However every big society has a very low number of really wise people. I dare you see the difference between smart and wise.

However it was Twinsen’s mistake to make unclear post to provoke this discussion and he admitted it which I do respect him for.

No matter do we want it or not our community does not consist of perfectly wise people only so we need to be most wise ourselves to prevent such collisions from happening and all stay positive and constructive.

[quadrox]

However it was Twinsen’s mistake to make unclear post to provoke this discussion [...]

I agree with most of what you say, but I have really hard time agreeing with that statement above. The first sentence says it's a crazy idea, so that should already tell you it might be nothing more than an idea. Then it takes just a tiny bit of patience to read the next few sentences to see that the idea will not be implemented. It was provoking, yes, but absolutely within reasonable limits I think.

But I understand that not everyone will feel the same way.

No matter do we want it or not our community does not consist of perfectly wise people only [...]

This is true unfortunately, I must acknowledge it. However, I wish the way forward to was to try to educate and elighten those who are less wise and less patient, instead of surrendering and making things "idiot-proof". Such an approach will just lead to even more impatience and lack of wisdom, as people come to expect having everything "prepared" for them instead of thinking themselves.

[eradicator]

I still don't get where all those people come from claiming the post was unclear, misleading, "playing with my emotions". Let me quote the original:

Don't worry, logistics bots won't be removed from the game, mostly because it's a feature that has been developed and polished quite a lot.

Like...how much more precise can he even say that? I am baffled. Do people not do thought experiments at school anymore? Or are all of those angry posters so young that those lessons are still ahead of them? The only conclusion i have left is that most of the people who had any strong emotional response to that FFF read only the first few sentences and then immediately stopped and skipped to flame posting. Which saddens me greatly (and there goes my plan to not have any sort of emotional reaction to this whole #situation).

I can only hope that this does not lead to wube replacing FFF content by real human developers with some soulless content-managed-to-death and written by community-mangement-proxy-guy shit to avoid directing emotions towards the devs. Though i could totally understand if they did that so devs can concentrate on making an awesome game and not waste time having to apologize for things they never said.

[golfmiketango]

That pertains to removing bots but there is serious discussion of debuffing bots which is what most of the hyperbole has pertained to. So, we do in fact have some skin in the game; it's not just a total misreading of the FFF. Those who read the FFF as stating that bots might be removed from the game ... well different people have very different levels of reading comprehension, especially as a function of the length of the text in question. I am not surprised, and I doubt the authors were either, to see this mistake.

I do agree with OP's sentiment however. I know I can handle it and would prefer not to be mollycoddled all the time. I personally very much like and appreciate Wube's style in this regard.

[vampiricdust]

What was not clear is whether bots are getting nerfed and how bad will the nerf be. The thing that upsets me is that Twinsen gave the reason to nerf logistic bots is because "they are not fun". His subjective opinion is the only reason the nerf talk is given any more time. He trashed bot bases by calling our work lazy, boring, and not interesting to justify forcing us to use his ideal fun way to play because he thinks it is better that way.

He used no objective measurements and even Kovarex is using misleading and deceptive tests at an attempt to prove bots were better by sheer brute force. They completely ignore the cost, research, and effort that goes into setting up a logistic network so they can trick people into agreeing bots need nerfs. If they nerf bots, it isnt because the game is not balanced, but because some devs dont like logistic bots and dont want players to want to use them.

So yeah, I'm heated. The devs and a lot of people are putting down my way of playing and want to nerf it so it is no longer any fun at all and I dont use them because they have become too weak to justify the cost and time.

[eradicator]

So yeah, I'm heated. The devs and a lot of people are putting down my way of playing and want to nerf it so it is no longer any fun at all and I dont use them because they have become too weak to justify the cost and time.

And why would that be bad? Games change. Let me start with an anecdote if you can spare the patience to read a longish post.

Back in school (arg, that sounds awful) i used to play a lot of Diablo 2, an Action-"RPG" game made my a somewhat popular company named Blizzard at the time. A game where you click on monsters and make them die. (Just for the people who really don't know, yea?). So, i like playing magic casters so of course i played a sorceress, and my favourite skill was firewall. Because, wohoo, i can put the whole screen on fire! That was in the beginning. Thereafter came the patches. In every patch (you can look up the changelogs if you care) firewall was either majorly nerfed, or boosted. So, every few weeks my character randomly became stronger or weaker. Eventually they added a cast delay to higher level spells to prevent spamming. Which made firewall a bad skill - for a while. They have motives besides balancing of course. Those frequent changes (and firewall was by far not the only skill to recive the up/down treatment) were supposed to keep the players enganged. Make them try old previously unused skills, make new characters, try new classes. And it worked. Of course every time my current favourite skill was nerfed i went "argh. not again". But my next reaction was to figure out what the new best skill would be and try that. And in addition to keeping players enganged they managed to slowly, with baby steps, approach a good overall balancing in which probably several hundred spells all have some sort of use, even in high level games. And they did the exact same thing for Diablo 3.

So why am i telling you this? Because Blizzard is often by their fans called "a master of balancing", mostly for Starcraft 1/2. So even a master, a multi-million dollar company with hundrets of employees, and a huge experience base still requires lots and lots and lots of time to get balancing right. Because balancing is ridiculously difficult. Because balancing is the attempt to please everybody, which is in real worls scenarios usually considered impossible.

So...back to factorio. You're accusing twinsen of having a "subjective opinion" and "not using objective measurements" ... which is a weird thing to say. There are no objective opinions. There are no objective measurements of fun or balancing. They do not exist. Don't fool yourself. You are in the extremely unique position - like all of us - to read a developers thoughts while he develops a game. This does not usually happen, not on small indie games, and especially not on large AAA+ titles. So the only option for wube to avoid your criticism would be to cease posting FFF, and keep everything a secret until the next patch/version is delivered. This would be a shame.

Let's face the facts: You like using bots. Great! But why do you expect them to never change? Belts have changed over time. Inserters got stacks. Science was completely overhauled and includes rocket launches now. Trains got fluid wagons. But bots haven't changed in a while, they're next in line. But maybe the better question is: Why do you really prefer bots over belts? Because they automatically balance all in/outputs? Because you can built extremely compact full-beaconed setups with 20 input items and still have it look neat? Or maybe you like their sound? Is there anything at all that you think belts do better than bots? What if - without any changes to bots - belts were boosted to be 10 times faster than bots, would you still prefer bots? If you're not using bots only because they give you bigger numbers than please, by all means, tell us all what is so much more fun about bots for you! Nobody can relate to your heated feelings if you don't explain them.

I think belt vs bots is ultimately about control. When you use belts, you control the exact path any item flows from source to destination. When you use bots you control no path, and no bots. You tell the network a source and target point and it automatically dispatches the closets possible robot (automatic optimization) to fetch the resource from the closest possible source (automatic optimization!) and deliver it to the target.

Finally let me say that i agree with twinsen in the point that i find belts more interesting because they make for a much better visual feedback of what my factory is doing right now. Huge bot swarms at high speeds are imho not something you can look at and understand what's going on right now, and they block the view of the underlying factory. As i play mainly singleplayer i couldn't care less about if other people use bots or not. But this is a luxuary that any wube dev simple does not have. They must care, because they are the only ones who can improve the game. They must think about what parts of the game they want people to enjoy most. If they think that encouraging players to think about logistics (belts) is more fun than using a black-box logistic network that automatically handles logistics without you doing much for it, then why do you not have any counter-arguments? You list "cost (of research)" and "effort". I don't see much of an argument there. Making infrastructure more expensive is not "more fun", and "effort" is something of course also required for belt setups.

Personally as a "belter" i find it extremely annoying that in the end-game i feel like the game is trying to forced me to use bots. Because belts simply do not offer adequate speed or versatility for large scale production. As a "belter" i have to actively decide against the method that has higher output. In a game all about building a larger factory that is a devestating flaw. But nerfing bots down to the speed of belts wouldn't improve that situation at all. It would simply mean that there would be no adequate solution at all anymore.

So in conclusion i'd say everybody who expected bots to never change again needs to think again. The game is still in development. Balancing is an ongoing process even - or especially - for games which were released long ago. I personally do not expect bots to become much worse than they are now, but they might become different to use. And launching personal attacks on a developer for sharing his thoughts is like facepunching the guy who just treated you to a good meal but told you that next time you'll have to pay yourself.

PS: And to prevent any misunderstandings: @vampiricdust some parts of this post are directed at you, but others (especially the last paragraphs) are general thoughts that i hope you don't interpret as personal attacks.

[vampiricdust]

Wow, that was text wall I dont habe time to pick through. I accuse him of usimg his subjective opinion to say thay bots are objectively better. He uses no measurable numbers to back up his claim that bots are better, but the FF points multiple times to his subjective view. There are different things and you're conflating the two. There is Twinsen's opinion and then a measurable objective that bots are better than belts. They are separate and neither justifies the other, but he wrote as if they did.

I never said I dont expect the bots to change, you're pinning you're own assumptions onto me with zero evidence. Please stop assuming things about my views. Bots have changed numerous times and it was always to put bots in line with their ability. I feel bots are good where they are, even if I think it is perhaps a bit underpowered for their cost. So if any change, I think they need a buff or with a nerf, their costs need to go down to match. Change is gine, nerfing them because someone doesnt think they are fun is another matter.

I like bota because I like watching them swarm around, like that they consume power to move items, and that I dont have to constantly solve the same belt problems for the hundredth time to increase production.

Belts are all around better than bots except for beacon builds (which I hate beacons) and burst dispersal at train stations. Outside those 2 uses, bots are too expensive when belts do the same for tons cheaper and provide way more consistent throughput without having to build massive extra power production.

Logistic bots are not as simple as people think. Often having more bots is worse tham having fewer as too many bots clog up roboports to charge, sending other bots to farther roboports to charge which causes them to take longer to get back into the flow later. Because you dont find them interesting doesnt mean there are no challenges or tricks and methods to optimizing them (because they are not self optimizing, a bad build is exponential worse than a bad belt build).

So in conclusion, you need to stop assuming you know more about what people think and use as reasoning because you're completely wrong. You're assuming your likes and interests are relevant to what I like, they are meaningless. The devs can change whatever they want, but unbalancing the game because they dont like something they put in and improved for years without making better is on them. I actually like what they did with bots until this talk about more nerfs. Now we're getting to the point if they nerf them anymore, they might as well just remove them because they wont be worth anything more than wasting resources for worse performance then speghetti belts.

[oracleofepirus]

Part of understanding which side of a debate is more or less correct is understanding both sides.

Every single pro-bot player uses the same arguments. Quantity of bots per roboport, research level, power usage, initial investment, ups, fun. I don't see a single such person conceding a single point. In fact, pro-bot players like to point out many of its advantages, such as inherent sorting and buffer space, no collision detection, or infinite research, as weak or obligatory.

This is the only reason I regard the pro-bot side as wrong. Your defense is petty. You literally do not concede any points without requiring an equal opposing concession. You fail to understand why belts are disadvantaged. I don't actually care whether the arguments you're making are right or wrong. If you were any proper defender, you'd be pointing out which parts of bots and roboports should change or stay the same. Not making stupid kiddie arguments like "You don't know me."

[quadrox]

-
-
All of you, please remain calm and civil!
-
-

vampiricdust:

Bots are better than belts because:

1) They have virtually no running cost (power is infinite with solar, and nuclear is not much better).
2) Although they have higher material cost, once you have automated their construction, they are practically free, because resources are also practially infinite. It is expensive to get to that point, yes, but once you are there, there is no limit to the number of bots you could have. None whatsoever (apart from UPS concerns). You don't even have to do anything - your factory just keeps on pumping out bots.
3) They are more UPS friendly than belts
4) They are much much more flexible than belts - they can get into places that are hard to reach with belts, all it takes is a single requester chest.

There is virtually no downside to bots over belts - if you can't see that I don't know what we can say to convince you. The things that make them "more expensive" than belts don't really count, because once you have automated their construction that cost doesn't require anything from you any longer - you have already set it up. (Technically you have to expand your power production once in a while and maybe set up more roboports, but that's just a couple of clicks with a blueprint as well, so I consider it negligable).

[Koub]

All of you, please remain calm and civil!

Hey, that's my job
But yeah I'm trying to keep a close eye on all the currrent threads running on that highly contrevorsial topic, but I'm afraid I'm on the verge of losing the battle . I highly dislike having to censor someone when I think he's gone too far, which I am happy to say does not happen pften despite the large community.
I wanted to point out that the initial intent of the OP was not to discuss the balance between bots and belts, but the way the last two FFF were made, and the way the community reacted.
Fot pure belts vs bots arguments, please discuss that in the FFF main thread.

OK so that's enough for the moderator-me, now my opinion of the community member-me on the OP.
I've seen many :
- Everybody knows
- We all know
- It's obvious that
- ...
bots are better than belts (but this truly applies to any balancing in a game)

These arguments forget that you can't prove a theory right. You can only prove it wrong by providing an exemple disproving it.
Most people who ask for bot nerf (but once again, this can be transposed to any XXX vs YYY balancing question) use the very same arguments :
- bots are too powerful
- bots are too easy
- bots are too flexible
- bots are too quick

Vampiricdust brings more data to account to determine if you really get too much bang for your buck with bots :
- When you get access to them
- how much it costs to get there
- how much it then costs to actually build the infrastructure
- how much bot infrastructure upkeep costs

And the answer he gets is "cost is meaningless, only result is revelant".

Well I disagree. This reminds me very much on the way people thought that heavier things freefall more quickly that lighter things, because "we can all see that". It took around 2k years for someone to actually account for the air resistance, and disprove the initial theory of Aristotle. Do you imagine his results dismissed by a "we don't care your arguments, we all see what's happening, and what you tell is not what we see, so your arguments are irrevelant". I'm glad they weren't .

Balance is something particularly subtle, but has different goals depending on the game. Diablo 2, which was used as an example, had many classes that had to be balanced between themselves. It's not the case for Factorio. I've been a long time MMORPG player, and I've seen many times people create a reroll of "that particularly OP class of the moment" to follow the nerfs and buffs trying to balance the different classes. I'm deeply convinced perfectly balancing classes in a game is impossible, unless they are basically identical.

Factorio, on the opposite, is not such a game. You don't have to balance that finely belts and bots (or whatever vs whatever) so they are perfectly equal, because if you do so, then ... you'll end up with two exactly identical systems. Which, in turn, will definitely not be fun (I think).

To conclude, there are many things for which people have called for balancing (solar power, bots, turrets (and turret creep), shotgun, ...), and I think that for a game like Factorio, just them being loosely balanced should be OK.

[vampiricdust]

EDIT: By the way Koub, thank you for not just dismissing my arguments completely without a moment of consideration. I really appreciate you <3

1) They have virtually no running cost (power is infinite with solar, and nuclear is not much better).

Did you know it takes 96 solar panels to charge just 4 bots? Those same resources could have been used to make 125.8 blue belts that cost 0 power.

2) Although they have higher material cost, once you have automated their construction, they are practically free, because resources are also practially infinite. It is expensive to get to that point, yes, but once you are there, there is no limit to the number of bots you could have. None whatsoever (apart from UPS concerns). You don't even have to do anything - your factory just keeps on pumping out bots.

Material cost is the #1 most common means of balancing items power/ability/usefulness to another item. I value resource efficiency, for you to just shrug off 1.48 million resources of initial research (1.28 million more than getting blue belts, which means not researching capacity 4, 240% speed bots turns into ~25k blue belts). By the way, you do have have to do things. You have to build roboports to cover the areas you're supplying, you have to place them as best as possible to reduce wasted flight times, you have to increase your power by 1mw per second of every bot charging, and if you don't not only do bots stop working, your whole factory stops working. I could place 25k blue belts and never build a single power pole, solar panel, or anything ever again because they have magic infinite power.

3) They are more UPS friendly than belts

Actually, from what I have been hearing, belts are now either equal or marginally better UPS than bots. Xterminator had a stream they were talking about it and I think the wiki admin said belts were now better UPS wise as of 16.16

4) They are much much more flexible than belts - they can get into places that are hard to reach with belts, all it takes is a single requester chest.

Yes they get into places hard to reach with belts, but there are only 2 instances this happens: Train stations and beacon heavy builds. Every other time if you cannot fit belts, you're doing it to yourself. If you spread out your assemblers to fit the belts, they would almost always be better than bots (use yellow belts for low throughput items and for every bot used to move those items, you could make 58.1 yellow belts).

There is virtually no downside to bots over belts - if you can't see that I don't know what we can say to convince you. The things that make them "more expensive" than belts don't really count, because once you have automated their construction that cost doesn't require anything from you any longer - you have already set it up. (Technically you have to expand your power production once in a while and maybe set up more roboports, but that's just a couple of clicks with a blueprint as well, so I consider it negligable).

Bots require a larger power production system, bots can cause blackouts if you don't have adequate production. Just because you ignore the costs doesn't mean the costs don't matter. 1.28 million in resource is mining out a couple starter patch or most of further out patch. It takes time to get that far and I think logistic bots are a fair reward for the effort and time it takes to scale that big. You can wave away and ignore all the flaws, costs, and effort required to get them, but that doesn't mean they don't matter.

[oracleofepirus]

I'm deeply convinced perfectly balancing classes in a game is impossible, unless they are basically identical.

Please see the current state of Diablo 3.

Factorio, on the opposite, is not such a game. You don't have to balance that finely belts and bots (or whatever vs whatever) so they are perfectly equal, because if you do so, then ... you'll end up with two exactly identical systems. Which, in turn, will definitely not be fun (I think).

This particular outcome is impossible. It is impossible to have bots and belts end up as identical. There are a number of properties that each has that cannot be replicated in the other under any circumstances.

To conclude, there are many things for which people have called for balancing (solar power, bots, turrets (and turret creep), shotgun, ...), and I think that for a game like Factorio, just them being loosely balanced should be OK.

As a person that views the boardstate, deckbuilding, and politics of multiplayer Magic: The Gathering easy, I'm telling you this from experience: Loosely balanced is not ok. It may be a different game, but that does not mean game theory cannot be applied.

Here's an example: Primeval Titan puts any two lands from your library onto the battlefield. What's the two best legal lands in Commander? One is Gaea's Cradle, which makes tremendous quantities of mana, and the other is Cabal Coffers, which makes tremendous quantities of mana. If, for some reason, Primeval Titan has haste, that makes four of the best legal lands in Commander, the next two being Urborg, Tomb of Yawgmoth, to facilitate Cabal Coffers, and a Vesuva, to make another Cabal Coffers. And then you get another two next turn. There are a handful of cards that get any land from your library. Most search only one. Most have significant disadvantages. None can triple or quadruple your mana output on the same turn.

It took over two years to ban Primeval Titan. Most of my playgroup was tired of it after a month. The point at which players decide to use specific cards just to make use of a single card that they don't even play, is way past fun. Literally every player that didn't either abuse Primeval Titan or steal an opponent's Primeval Titan got run over. By that time, quantity of Commander players had been cut in half. Now, because the banlist is even worse (ie, the same), the quantity of Commander players is now a tenth compared to its peak popularity, because no one wants to play against such cards, and those that do play, choose not to use them. (Does this sound familiar?)

I can trace this specific pathway of decline in many games, including Yugi-Oh, and Dominion.

Let me be specific. When I say a game needs to be balanced, I mean that there cannot be a player that chooses to not use a specific mechanic because it is too good. There are no players that don't use belts because they think belts are too good. There are plenty of players that don't use bots because they think bots are too good.

Whether the difference is changed by way of buffs, nerfs, or adding content is irrelevant to me.

[vampiricdust]

As a person that views the boardstate, deckbuilding, and politics of multiplayer Magic: The Gathering easy, I'm telling you this from experience: Loosely balanced is not ok. It may be a different game, but that does not mean game theory cannot be applied.

Here's an example: Primeval Titan puts any two lands from your library onto the battlefield. What's the two best legal lands in Commander? One is Gaea's Cradle, which makes tremendous quantities of mana, and the other is Cabal Coffers, which makes tremendous quantities of mana. If, for some reason, Primeval Titan has haste, that makes four of the best legal lands in Commander, the next two being Urborg, Tomb of Yawgmoth, to facilitate Cabal Coffers, and a Vesuva, to make another Cabal Coffers. And then you get another two next turn. There are a handful of cards that get any land from your library. Most search only one. Most have significant disadvantages. None can triple or quadruple your mana output on the same turn.

It took over two years to ban Primeval Titan. Most of my playgroup was tired of it after a month. The point at which players decide to use specific cards just to make use of a single card that they don't even play, is way past fun. Literally every player that didn't either abuse Primeval Titan or steal an opponent's Primeval Titan got run over. By that time, quantity of Commander players had been cut in half. Now, because the banlist is even worse (ie, the same), the quantity of Commander players is now a tenth compared to its peak popularity, because no one wants to play against such cards, and those that do play, choose not to use them. (Does this sound familiar?)

I can trace this specific pathway of decline in many games, including Yugi-Oh, and Dominion.

Let me be specific. When I say a game needs to be balanced, I mean that there cannot be a player that chooses to not use a specific mechanic because it is too good. There are no players that don't use belts because they think belts are too good. There are plenty of players that don't use bots because they think bots are too good.

Whether the difference is changed by way of buffs, nerfs, or adding content is irrelevant to me.

Your comparing it to MTG is not even remotely applicable because we all have the same deck, which the only way to change the deck is to mod it. If you really don't like bots that much, THEN MAKE A MOD and don't force people you'll never play with to play by your ideal set of rules. I already plan to make a mod to balance bots and remove belts, you're perfectly welcome to make a mod that takes out the requester chest.

[Koub]

You're comparing games whose focus is PvP to a mostly PVE game. To my knowledge, even though there is some PVP possibility with custom scenarios on Factorio, only a minor proportion of the player base actually does PVP, it is mainly PVE and cooperative.
When people play on a server, the one using belts will not lose if a bot user connects. There is the same "imbalance" between trains and belts, but nobody points it out. Anyone has measured the maximum throughput achievable on a 1k tile between a 2x blue belt vs a nuclear fuelled train ?

It reminds me at my job (I'm in IT support, in a quite big company), when a project would cause much user impact, they usually pay for some reinforcement of the IT team to mitigate the effect of the project on our run. Recently a project told us they planned to make a quite heavy upgrade that would most probably end up with on average one call by migrated user. The issue is that the deployment is planned to be 3 months long (the userbase is a few tens of thousands). They had the credit card in hand to pay for the IT support reinforcment, but when we added up things, the reinforcement would have been ... just twice our current staff in volume. Money was the least of our problems : where do I fit 200 more prople in my actual premises ?

When you watch the the links provided in the last FFF, it is said that everything should be accounted for when you try to balance a game, even the player's time. Vampiricdust has provided some numbers to account for the player's time needed for the bots to actually outclass belts. My time is valuable to me, because I have so little of it. I have almost no time to play Factorio, because 90% of my time goes into :
- Work
- Forum moderation
- Eating/Sleeping
So if I want maximum flexibility, when I play Factorio, for how I choose to spend the few hours I have left for myself.

[oracleofepirus]

I see my point hasn't been understood yet. Application of game theory is fundamental to both design and use of systems. There is nothing that cannot be viewed in terms of game theory, as math is the fundamental language of the universe. At the point at which one can effectively evaluate the state of a multiple-input non-deterministic game like Magic: The Gathering, a single-input deterministic game like Factorio becomes rather easy.

What is the difference between an equation and an expression? An equation has a specific calculable value, whereas an expression doesn't. That's the fundamental difference between sandbox games and non-sandbox games. A non-sandbox game wants you to use game mechanics to reach a specific point, whereas a sandbox game lets you decide which point you're going to. The ability to understand the math behind them is the same either way.

The absolute farthest one can reach inside a sandbox computer game is limited only by the game's code, and the hardware it runs on. There is no question about the effects of either belts or bots on UPS, and no question that belts have literally reached that point sooner than bots. That's the entire reason behind the belt code rewrite, and these splitter changes. Now, this rewrite is rather unstable, and obviously has some more issues ahead, so we may not have the ability to retest belt ups for a while, but here's the point. Using v0.15.40 or before, on the same hardware, a belt-heavy factory cannot reach the same size as a bot-heavy factory because of ups issues. So without using bots, a person literally has less game to play, because the code and hardware literally provide a smaller sandbox environment before breaking down.

You say that player time is valuable, and that is completely true in all games, and in fact, the most valuable resource in all games. Here's the problem in different words. You can either spend 40 hours (example value) and do everything you want using belts. After that, there is no more game to play, because you have literally done everything belt related and the rest isn't interesting. Or, you can switch to bots, and you can wring another 40 hours (example value) out of it. This time investment that vampiricdust has accounted for is literally the issue. A belt-heavy player literally does not make this time investment, because the code and hardware literally stop them (as of v0.15.40).

Would you, as a developer, want a player to spend 40 hours in your game before stopping, or 80 hours in your game before stopping? Which is preferable, one playstyle with 40 possible hours and one with 80 possible hours, or two playstyles both with 60 possible hours? Or perhaps two playstyles, both with 80 hours possible?

[Trudel]

What was not clear is whether bots are getting nerfed and how bad will the nerf be. The thing that upsets me is that Twinsen gave the reason to nerf logistic bots is because "they are not fun". His subjective opinion is the only reason the nerf talk is given any more time. He trashed bot bases by calling our work lazy, boring, and not interesting to justify forcing us to use his ideal fun way to play because he thinks it is better that way.

He used no objective measurements and even Kovarex is using misleading and deceptive tests at an attempt to prove bots were better by sheer brute force. They completely ignore the cost, research, and effort that goes into setting up a logistic network so they can trick people into agreeing bots need nerfs. If they nerf bots, it isnt because the game is not balanced, but because some devs dont like logistic bots and dont want players to want to use them.

So yeah, I'm heated. The devs and a lot of people are putting down my way of playing and want to nerf it so it is no longer any fun at all and I dont use them because they have become too weak to justify the cost and time.

This was the only thing that irked me about his post as well. I don't really care if logistics bots are nerfed or even removed, but I allready like the new splitter mechanics and would like to see belts improved in meaningful ways if possible. Nerfing something because it's "not fun" is the worst possible reason you can find. It reminds me of the constant community backlash against Blizzard and especially their former game director Jay Wilson running up to and after the release of Diablo 3. Granted the Diablo fanbase is "not fun" to please, but Blizzard contiously using this reasoning was fanning the flames needlessly.

Thankfully Twinsen seems to have realized this:

What is most fun is of course an subjective and extremely complex topic that dives deep into game design, human psychology and player motivation.

[Koub]

You say that player time is valuable, and that is completely true in all games, and in fact, the most valuable resource in all games. Here's the problem in different words. You can either spend 40 hours (example value) and do everything you want using belts. After that, there is no more game to play, because you have literally done everything belt related and the rest isn't interesting. Or, you can switch to bots, and you can wring another 40 hours (example value) out of it. This time investment that vampiricdust has accounted for is literally the issue. A belt-heavy player literally does not make this time investment, because the code and hardware literally stop them (as of v0.15.40).

Would you, as a developer, want a player to spend 40 hours in your game before stopping, or 80 hours in your game before stopping? Which is preferable, one playstyle with 40 possible hours and one with 80 possible hours, or two playstyles both with 60 possible hours? Or perhaps two playstyles, both with 80 hours possible?

Actually, both examples don't really compare, because making a belt-only (or mostly) factory doesn't require you to invest your time in the same things as a bot-only (or mostly) factory.

Let me take another example :
According to Steam I have actually around 900 hours played on Skyrim SE, out of which I spent maybe 50 hours actually having fun playing, around 50 hours playing without, retrospectively, having fun (pure skill exp grind, inventory management, ..., but tht's my own fault, I'm a maxer, I physically need to max out everything, collect everything, ... and I just lose interest somewhere on the path), and finally around 800 hours afk skill exp farming.

That's 5.5% fun time, 5.5% unfun time, and 89% afk grind. Or 50% fun play time, 50% unfun play time if I remove the AFK part.

According to Steam, I have short of 1200 hours played on Terraria, out of which maybe 100 hours (maximum) was afk farming for gold/items with an automated farm of one sort, 50 hours maximum (but probably less) having fun farming items/ressources actively (bosses/events), and that's probably around 1050 hours playing and having fun.

That's 87.5% fun time, 4.2% fun grind time, and 8.3% afk farming. Or 95.5% fun play time, 4.5% fun grind if I remove the AFK part.

The time both games have been played is in the same order of magnitude, however, the time ratio during which I did tedious things over total time (or even total time not AFK) is radically different.

The logistic network (I mean the full logistic network, with requester/provider chests) is a quite endgame mechanism. Once you have access to researching it, you're best going directly for the rocket it your goal is to launch the rocket. It's a bit like the endgame stuff you'll get in Terraria. Yes, it's OP, and with it, endgame content is less of a challenge. But that's normal : it's the endgame content. Logibots and logistic network are endgame content. If they were less powerful than the belts or on par with the belts, it would be unfair.

[Zavian]

Yes, it's OP, and with it, endgame content is less of a challenge. But that's normal : it's the endgame content. Logibots and logistic network are endgame content. If they were less powerful than the belts or on par with the belts, it would be unfair.

Whilst I agree with the main thrust of your points, I disagree with this wording. If bots were inferior to belts in all respect I'd say "What's the point of the game having bots. They need to be re-balanced. In order to justify their existence they should be superior to belts in at least one respect".

With Factorio as it stands, even if the devs nerfed bots and added stacked belts, and faster inserters (to help with high throughput belt to assembler builds), bots would still be superior in at least one aspect. Player convenience. Once you have requester chests, low volume builds for things like locomotives will still be simpler and more convenient to automate with bots than with belts, so bots will still have a niche. I don't expect the devs to nerf bots to the extent that belts are better, but I hope they can be convinced to add some variety of stacked belts, for high throughput items. (eg iron and copper plates, green circuits. A stack size of even 4, for just these items, would make belts more powerful in 8x8 late game beaconed builds).

Since bots are a late game item, they should to be superior to early and mid-game items (belts and trains) in at least some aspects. They don't need to be superior to mid game items for every use case. If new late-game belt mechanics (eg stacked belts) were added, then bots just need to be better for at least some use cases (simple low volume builds, player convenience) to continue to justify their existence.

[Koub]

Well I wanted to try and keep on topic with the OP, aka not remake the debate belt vs bot, but discuss more generally on the way the devs and the community have exchanged around that so sensitive and polemic balance issue. That's why I try to stay generic, but the bot vs belt keeps sliping even in my own posts .

I can at least count one aspect on which belts are infinitely superior to bots. I mean a real infinity, not just "a lot" : Electric consumption. The whole belt system of a base as big as you like use exactly 0 power. A single bot consumes in one second infinitely more power than all the belts of Factorio, since the game's first belt.

And the second thing belts are highly better than bots is cost. For a given throughput, logistic network costs a LOT more than belts (not even factoring the research cost, I just mean the infrastructure and the bots). Belts are extremely more cost effective.

Given this, I think it justified that bots can superior in performance and versatility : you pay the price for it. The is no obvious imbalance between both, they all have their pros and cons.

By the way, belts have another thing bots don't : it's fairly easy to finish the game without using a single logistic bot (even without using a single construction bot). I dare you launch a satellite without laying a single belt