Aren't construction bots available too late?

[dog80]

The best way to verify this is by testing. So I just built a small base with purely red/green science.

I now have around 500 construction and 500 logistics bots (no research beyond green) and TBH they are nearly worthless.

At maxed red/green science level the bots are so slow that it will probably require a couple thousand to do anything at even an acceptable rate.

A personal roboport might make them more useful, but that will be much later.

bots can be super usefull even that early, and with around 500 you are more than set if you dont put storage chests at the other side of the factory and actually give bots the items they need to construct stuff... - a personal robothangar is not *much later because you only have to setup a very minimalistic blue-science setup, which if you already have automated bots is not very hard, because you already *should have miners/engines and red-cuircits automated...
then with the 500 bots you can quickly build more miners / smelting lines and get things going for real... lol
also they are so slow but t1, t2 bot speed are really not a big investment...

the biggest problem at this early stage with bots is the lack of recharge possibilities (roboports) (and sometimes energy)

[zOldBulldog]

The best way to verify this is by testing. So I just built a small base with purely red/green science.

I now have around 500 construction and 500 logistics bots (no research beyond green) and TBH they are nearly worthless.

At maxed red/green science level the bots are so slow that it will probably require a couple thousand to do anything at even an acceptable rate.

A personal roboport might make them more useful, but that will be much later.

bots can be super usefull even that early, and with around 500 you are more than set if you dont put storage chests at the other side of the factory and actually give bots the items they need to construct stuff... - a personal robothangar is not *much later because you only have to setup a very minimalistic blue-science setup, which if you already have automated bots is not very hard, because you already *should have miners/engines and red-cuircits automated...
then with the 500 bots you can quickly build more miners / smelting lines and get things going for real... lol
also they are so slow but t1, t2 bot speed are really not a big investment...

the biggest problem at this early stage with bots is the lack of recharge possibilities (roboports) (and sometimes energy)

Hmm... Good thought. I have a little free space left in my minibase. I will look into adding some minimal additional blue+ science to it.

Roboports and energy were not a problem, my standard layout already planned for it.

[Deadly-Bagel]

I don't necessarily mind the required ingredients, oil isn't hard to get and you only need 1 Refinery to make your first few bots - put the blueprint down and manually place the first one, it's only a little fiddly.

My problem with them is their initial cost vs usability. They're not so bad from static Roboports, they can passively do stuff while you do something else, but they're almost useless in the Personal Roboport, especially with the random nature of how ghosts request items. So I guess just use static Roboports everywhere? Oh yeah that's 45 Advanced Circuits and 45 Steel each. Just use Personal Roboport then, ok now I'm standing here waiting for my robots to come back, because I run 4x faster than them. There is so much waiting involved in getting bots up to scratch, I think this is where the annoyance stems from. If we could get usable bots shortly after oil then I don't think we'd have any problems.

Either reduce the Roboport cost to something like 10 Advanced Circuits and 20 Steel, or triple base movement speed and cut out the first 4 Worker Robot Speed sciences.

[4xel]

I don't know what you're calling "late". Speed runners have their first construction bots at 1h30, and a decent fleet 2 hours in. I just got mine in a new game in 3h30 without try harding.

In my book, that's very early. Yes technically, for speed runners, that's like at 2/5 of their time spent without construction bot, but if you don't like early game, why resarting every time you launch a rocket? Plus the complexity from oil excluded to rocket is much greater than the complexity up to oil included, so for a normal human being we're more looking at 1/5 of their time spent without robots before the first rocket. And for a normal games which you keep playing after your first rocket, more like 1/10. And for mega factories, more like 1/100.

My problem with them is their initial cost vs usability.

Initial cost being high is not really an issue, and the usability is tremendously high. The same thing about prohibitive entry cost can be said about solar panels, MK3 modules and beacons. But I think we have to consider that :

• This is a long term investement. A robot port can never become useless. At worse, you mine them to place them elsewhere as they finish being useful. Oh, and by long term, I mean short term, construction bots are (a must) used in speed runs, the same can't be said about above mentionned solar panels and MK3 modules.
• By the time you got robot port, even much earlier actually (1/4 to 1 hour), you can have practically infinite ressources, that is, more ressources that you're able to spend on new miners and oil wells, which brings the next point :
• Soon enough (arguably right from the beginning), the most precious ressource in the game is your time, and by far. Construction bots are the only way to cut through the (then huge) construction time, and are a no brainer as soon as you get them, even when they look expensive. That's also the reason why I discard using personal robot port, or advocating for increasing its use and availability, as mostly stupid, since you gain a lot more time with real robot ports. Personal robot port have niche uses at best.

I now have around 500 construction and 500 logistics bots (no research beyond green) and TBH they are nearly worthless.

You are judging them by endgame standards. By these standards, you are worthless too. In fact, we (a player) are orders of magnitudes more worthless. The question you should be asking is :

who are the fastest at mass construction, one Human player, or 50 unupgraded construction bots?


I'd choose the construction bots any day of the week. Even if the human player is a speed runner (oh and did I mention that speed runners do use construction bots?). My money would still be on 20 bots. Maybe even on 10 or 5, if the player is inefficient and slacking like most humans are.

And that is disingenuously ignoring the fact that by the time you got construction bot mass produced, G and R research, red chips and engines are automatized, so there is very little to do to unlock the first two speed bonuses (75% boost).


Finally, one can use simple lua commands to skip over specific part of the early game that one might find tedious, by giving one self items or technology.
https://wiki.factorio.com/Console#Inven ... on_scripts
SImilar goals can be achieved without coding through the ingame editor, and I guess mods.
(It does not replace a bot rebalance if it ends up being judged as needed, but these tools are worth noticing, since they enable a degree of personalisation no balance patch could replace).

[Jap2.0]

Wait, speedrunners use bots nowadays?

[4xel]

Wait, speedrunners use bots nowadays?

Since about 3 month ago, but they used to use bots in 0.15 too. I've always thought construction bots were a must for speedrun, but that's an unsubstantiated opinion. Fortunately, someone better than me, rain9441 I think, proved this point and changed the Meta :

https://www.youtube.com/watch?v=4N1ALULDy3k

https://www.speedrun.com/Factorio

https://www.speedrun.com/Factorio/run/y43972nz

[Deadly-Bagel]

Initial cost being high is not really an issue, and the usability is tremendously high. The same thing about prohibitive entry cost can be said about solar panels, MK3 modules and beacons.

I tried Solar Panels post-nuclear update, they are also quite painful to build since science now requires a LOT more Steel. Why bother? You're better off investing the Steel in rails and solidifying your coal plant, which provides over 20x the power for 1/2 the cost, and that's not even including the Accumulator costs.

Besides, you're missing the point. Bots are supposed to automate the setup of your factory (to a degree), but you actually need a reasonably large factory to use them in the first place. What are we, a bank? "We won't give you a loan until you can prove you don't need it."

By the time you got robot port, even much earlier actually (1/4 to 1 hour), you can have practically infinite ressources

As you said, the most valuable resource is time. Sure, I could go set up 100 Furnaces and either their coal feeds or additional power plant (depending on if Steel or Electric Furnaces) but it would be by hand. Now, I don't necessarily mind the early pre-bot game, it's the repetitive tasks that I can't stand, and for bots you need quite a lot of Steel and Circuits, which means a lot of extra Mining Drills, Furnaces, and Electronic Circuit Assemblers. Hopefully you don't need any rails any time soon because they're a real pain to set up by hand, and slow with low-research bots. Oh yeah, it also requires the Personal Roboport so good luck with that.

The question you should be asking is :

who are the fastest at mass construction, one Human player, or 50 unupgraded construction bots?

If it's straight rails or Mining Drills we're building then the player shouldn't have too much trouble being faster, especially if you initially skip non-junction signals. In fact I'd say trying to use bots with just the Modular Armour is pretty painful in general, but the Power Armour still requires Processing Units as does the Fusion Reactor.

[zOldBulldog]

As you said, the most valuable resource is time. ... Now, I don't necessarily mind the early pre-bot game, it's the repetitive tasks that I can't stand.

This hits the issue squarely on the head, and determines the replayability of the game.

Doing the repetitive tasks when you are new is good, it helps you learn the game. By the second time you start, it is boring. By the 3rd or 4th you start pulling your hair.

Mods are a viable alternative to the repetitive drudgework... unless - like many players - you want to complete the achievements before you start using mods.

I can see two possible solutions:

- Give the player a map generation option (that becomes enabled only after having launched a rocket in a previous game) to retain all past bot research levels.

Or,

- Remove the restriction that disables achievements if you use mods. Why? Because achievements are supposed to be a personal thing, not something you use to compare with others. And from that perspective... nobody should care whether you did them with mods, or as many do now... in multiplayer with the assistance of others. Heck, I would make most achievements personal, and only leave a few "competitive" achievements in the Steam public view.

[mrvn]

I'm always playing with the bluebuild mod (similar: nanobots). This automates building ghosts or deconstructing things that are within the players reach. That basically gives you a mini personal roboport from the start. So I really don't feel it when it takes a long time to research construction bots. Just means I have to walk to where I blueprint.

[zOldBulldog]

I'm always playing with the bluebuild mod (similar: nanobots). This automates building ghosts or deconstructing things that are within the players reach. That basically gives you a mini personal roboport from the start. So I really don't feel it when it takes a long time to research construction bots. Just means I have to walk to where I blueprint.

I am looking forward to it. I just have a couple of achievements to complete before I can start using mods and get rid of the drudgework.

[mrvn]

Don't worry, with mods you get to do all the achievments all over again.

[Deadly-Bagel]

Yeah but they don't come up in Steam. Mainly the 20mill Electronic Circuit one is painful, it's difficult to actively try to get. I mean you can, but really what's the difference between that and just building a big factory? Everything needs circuits.

Perhaps a flag could be included in a mod that says "this is non-achievement-breaking" - could be entirely up to the discretion of the mod author, perhaps put a flag on the Mod Portal. Use it for those little QoL mods like a fast-start or earlier bots or w/e. Who cares if a mod comes up to get all achievs lulz, you could already download a savefile to quickly unlock them, not sure if base game prototypes could be edited as well?

If any mods are included without the flag, it just clears all achievements from the game. They can add their own right? I mean it seems kind of pointless running vanilla achievements with something like Bob's Mods, they're mostly incompatible.

That way you could play the game a few times vanilla, then add a few QoL mods like better early bots, furnaces and armour build into each other, recycler, whatever, and still enjoy the satisfaction of Steam reporting you've 100%'d the game.

[Jap2.0]

Wait, speedrunners use bots nowadays?

Since about 3 month ago, but they used to use bots in 0.15 too. I've always thought construction bots were a must for speedrun, but that's an unsubstantiated opinion. Fortunately, someone better than me, rain9441 I think, proved this point and changed the Meta :

https://www.youtube.com/watch?v=4N1ALULDy3k

https://www.speedrun.com/Factorio

https://www.speedrun.com/Factorio/run/y43972nz

Okay, that's interesting. I haven't paid much attention to speedrunning since AntiElitz sub-2hr one in 0.14.

[4xel]

Okay, that's interesting. I haven't paid much attention to speedrunning since AntiElitz sub-2hr one in 0.14.

Science revamp is what made the game long enough for bots to be useful in speedruns, so no wonder you have not seen them in 0.14.

Initial cost being high is not really an issue, and the usability is tremendously high. The same thing about prohibitive entry cost can be said about solar panels, MK3 modules and beacons.

<Solar Panels are useless>

Sure, not everybody agree, but you're entilted to an opinion. How does that invalidate my claim about similar prohibitive entry cost?

Besides, you're missing the point.

NO U.

My problem with them is their initial cost vs usability.

I think I pretty much addressed this point by saying the cost is not that high (comparable to commonly used SP and modules) and later explained why the usability is "tremendously high".

Bots are supposed to automate the setup of your factory (to a degree), but you actually need a reasonably large factory to use them in the first place.

Which boils down to "bots come to late", while I initially thought you were trying to make a slightly different point. And I've already ansered that.

On that, let's agree to disagree. In my opinion, a factory with barely more than one line of iron and less than one line of copper and steel, with 5 to 20 research lab and one or two refineries, is what I call a bootstrap base, not a reasonably large factory. And you can build it and research and automate robots within 1,5 to 4 hours, which is what I call early game.

I'll edit this post later with a save game to prove my claim of 4h, 2 metal lines and <20 research labs.

What are we, a bank? "We won't give you a loan until you can prove you don't need it."

There's no bank in factorio, you have to build anything that you want to be built, with your own ressource, that's worse than any optional investement system involving a bank you can imagine. If you played any economics simulator, you'd know that, most of the time, having any money right now is good thing, even when you don't strictly need it. We're not talking about consumer loan, we're talking about investement loan in a video game when you're near 100% sure to have higher ROI than the interest rate.

And you always need bots. They are the only things which do what they do. Unless you stop playing after your first rocket (and even then) or sooner. In which case I'm not sure you'rre after the full unmodded experience.

By the time you got robot port, even much earlier actually (1/4 to 1 hour), you can have practically infinite ressources

As you said, the most valuable resource is time.

As I said later. I was preparing my point which was, no, bots are not too expensive for their purpose. My line of reasoning is "very soon, building/having bots is the only meaningfull way to spend ressource".

Does bot come too late? that's an other question, to which my stance is, as you already know, 4h is not late.

Hopefully you don't need any rails any time soon because they're a real pain to set up by hand, and slow with low-research bots. Oh yeah, it also requires the Personal Roboport so good luck with that.

That's just plain wrong. On the surface, building rail with personal bots sounds like a good idea, and it is indeed a decent one, but in no way is it the only nor the best solution.

Solution 0 : Building by hand.
Solution 1 : going on site, using personal roboport, running around, babysitting construction.
Solution 2 : going on site, dropping a temporary roboport, babysitting construction, dismantling roboport and repeat.
Solution 3 : open radar view and drop blueprints containing rails, radars and roboport.

• Solution 0 is really not as bad as you think, maybe 2 to 3 times slower than other solutions to build new double railways.
• Of course solution 2 are much lower tech than solution 1, but that's part of what you want to change, so I won't touch further on this.
• A mobile roboport like solution 1 benefits from the thin long characteristic of the structure, since you can cut robot wandering time by running around.
• For solution 1, you're limited by the power of your armor, and by the construction bot count. That's not too bad for thin long structures, but it does mitigate the advantages.
• Solution 2 can be refined by dropping the whole railway at once, going from roboport to roboport with a car, for robot to build the whole railway in parallel. It goes from significantly slower than solution 1 for a single track to siginficantly faster. (In a sense, it makes the power and robocount advantage matter, despite the structure being long and thin).
• Solution 3 is probably the slowest, in the sense construction time is the biggest, but it also is the one requiring the less player time, by a huge margin. You only have to click the blueprints, and you can order dozens of other bluprints during the construction you'd have to babysit otherwise.
• One disadvantage of solution 3 and refined 2 is the roboports (and chest) you have to leave along the way. It can be mitigated by removing them during a one pass after the track is built. For that particular use, I'd recommend personal roboport if available, but they're not a gamechanger.

Which is why I say personal roboport are mostly useless, reguardless of their cost and technology requirement.

The only way I see to make them temporarilly useful would be to make them available before normal robot port, as well as the necessary infrastructure, which effectively means making power armor available before or at the same time as construction bot, which I don't think is desirable. (I think we should be able to slowly -15mins to 1h- reload modular batteries from the main grid, but that's for another topic)

The question you should be asking is :

who are the fastest at mass construction, one Human player, or 50 unupgraded construction bots?

If it's straight rails or Mining Drills we're building then the player shouldn't have too much trouble being faster, especially if you initially skip non-junction signals.

I'm ready to take bets. I'll probably do the test tonight. I'll do semi-dense mining area rather than straight line, because you're choosing the one particular case robots struggle with, and a single straight miner line don't make sense (semi dense mining still favors human an awfull lot, compared to sparse mining and dense mining with underground belts to make room for poles).

Or If I do rails, I'll use refined solution 2 explained above and double straight rails and power poles.

In fact I'd say trying to use bots with just the Modular Armour is pretty painful in general.

The correct way to use them with just the modular armor is to mine them as they try to recharge. Painful indeed, but you should be using real roboports instead. Even for things you think personal ports are necessary.

[BlakeMW]

The correct way to use them with just the modular armor is to mine them as they try to recharge. Painful indeed, but you should be using real roboports instead. Even for things you think personal ports are necessary.

Personally I do fine with Modular Armor setups w/ something like 2 roboports, 2 batteries and 11 solar panels. Of course it's not possible to build all the time with such a setup, but most the time I'm using the grid roboports anyway and when I actually go on an excursion the batteries and roboport are all fully charged.

Also, if bots go on short excursions they don't bother charging meaning their trip is free. An example of how this can be useful is if you're using the Tank offensively and using the powerful "drive slowly" technique, bots pop out, repair, and pop back in, most the time using no charge whatsoever. Another example if doing offensive turreting you can drop clusters of turrets on top of yourself and the distance out and back in is so short they use no charge. This means for the major offensive uses of the portal roboport charge is almost never an issue and you can keep going for as long as you like.

The main limitation of the portables is when placing rail by bot, trip times have a tendency to be maximized by the bot AI so you don't get the charging waived. Of course, you can still place a lot of rail by hand and just use the bots for the deconstruction of obstacles.

Anyway overall it's pretty rare I have to resort to hoovering up the bots, and if I'm not going rail heavy it's entirely possible I wouldn't need to do it once in a game (and I always get Modular Armor early, and use it for a long time).

[Rizzano]

I've used the mod Nanobots for earlier personal bots. You don't get them right off the jump, but you do get them significantly earlier than vanilla. If construction bots were unlocked earlier, I wouldn't turn to mod to do it for me. I do agree with someone else that said logistics bots should be later than construction bots.

[Deadly-Bagel]

What are we, a bank? "We won't give you a loan until you can prove you don't need it."

There's no bank in factorio...

While I don't care to try to convince you that arduous repetitive tasks past the 15th playthrough are not fun in any way, I do have to point out that you COMPLETELY missed the point of this analogy. A bank is usually a bit funny about your first loan and seem to want you to prove you don't need it before they'll give it to you, similarly in Factorio you don't get these awesome bots to automate arduous repetitive tasks until you do a lot of arduous repetitive tasks.

The correct way to use them with just the modular armor is to mine them as they try to recharge. Painful indeed, but you should be using real roboports instead. Even for things you think personal ports are necessary.

Though I do have to lol at this, "The correct way to use them is to abuse the game's mechanics" xD

[zOldBulldog]

[quote="Deadly-Bagel"]arduous repetitive tasks past the 15th playthrough are not fun in any way,/quote]
Completely correct. Although the game misses that point completely. Thank God for mods and mechanics that can be abused.

Some games solve this issue by giving you critical advanced tools from the beginning once you beat the game once.

For example in Factorio, once you launch a rocket you could crash ontonthe next planet with a few chests, and among other things they could contain your full army of bots, roboports, armor, armor contents (including personal roboports) from past playthroughs...hey! Cool! They survived the crash! Of course, you better protect them because if you loose them you are going to have to research them again. But in the meantime they retain their old abilities, even that awesome speed you researched through space science.

Mmmm... I wonder if someone could make a mod like that.

[darkfrei]

once you launch a rocket you could crash ontonthe next planet

The main character is a very unlucky guy. Landing, landing is much better than crash into it.

[zOldBulldog]

once you launch a rocket you could crash ontonthe next planet

The main character is a very unlucky guy. Landing, landing is much better than crash into it.

Good point.

For the backstory maybe the origininal crash planet lacked an ore (or ores) needed for long-range travel, so he needs to exploit additional planets. That could open the door to significant replayability and even expansion into previously untouched tech research.