#What are the use cases of red belts from a purely strategic standpoint?

1 messages · Page 1 of 1 (latest)

brazen lynx
#

When we first research red belts they're great. They solve throughput issues without having to worry about the actual issues, and belt weaving is much less complicated with only one belt. But eventually, we get past that, and at that point, at a high level of play, I wonder what use cases red belts have. Obviously, you would eventually want to use them in a big enough base, but let's assume reasonably small bases (pre-rocket).

Yellow belts are more than twice as cheap, and space is not a huge issue, which would make it seem like red belts are just the clear loser since, well, there's not much else going for a belt. You could argue that cost isn't a big issue either, but it turns out it is, as much as people would like to ignore it. For example, in my deathworld marathon playthrough (currently at utility science), yellow belts account for like 5% of all iron ever produced. Red belts are 5x more expensive for the same amount of throughput... I think the implications are clear.

However, there are cases where red belts are actually cheaper as pointed out by Putin's Chest here on the discord. For example, using one red belt for steel is cheaper than two yellow belts since the steel that has to pass over the belt is simply so expensive. Having more belts mean a bigger buffer, which ultimately results in more frontend expense. So it turns out there are use cases for red belts. And in resource-rich worlds maybe using them in some places and not others is a pretty feasible strategy as well. So what are the use cases of red belts?

tacit hornet
#

They ratio well with yellow belts. Two yellows can perfectly sideload (if you have full compression). Plus, as expensive as they are, blue belts are much worse. You can use them in medium-throughput areas that yellow belts aren’t sufficient for. Plus yellow inserters can miss items on blue belts if they’re sparce enough, but red is slow enough that they’re always caught.

wind raptor
#

but red is slow enough that they’re always caught.
except far corner on turn

#

you mentioned belt weaving: and for that it is useful to have multiple belt types

#

in some spaghetti cases it may be needed to split blue belt into red/yellow ones to weave them through some tight spot among other blue belts there

obsidian tree
#

In the case of bus-based base design they help keep the number of lanes of the bus reasonable. More lanes means more undergrounds needed to cross lanes which itself is a cost.

Otherwise they mostly fall into the case of improving something that already exists to go faster. Full belt of circuits produced with grey assemblers? Upgrade to red belts and blue assemblers and potentially have access to more circuits!

#

As the game progresses, the cost of red and even blue belts eventually becomes trivial. This is one of the reasons why most mega bases end up spamming blue belts everywhere.

pastel acorn
#

From my playthrough, as soon as there are trains moving basic resources like iron, copper plates and green circuits, a simple mall will supply more than enough red belts and stack inserters

obsidian tree
#

It is more an issue in deathworld situations where the additional ore used to make red belts means additional pollution which means more and tougher biters. Although you should be unlocking flamethrower turrets around that time, which handle all that efficiently.

quick ivy
#

deathworld is weird because space is also a premium

fair lantern
#

in my opinion sometimes the complexity saved by using red belts far outweighs the material costs. Also, in the lategame of factorio, you're going to be bottlenecked on copper anyways, so may as well use that extra iron.

#

and when you're doing beaconed designs, saving space by using blue belts winds up being cheaper than giving up precious beacon space to run belts.

brazen lynx
#

I was wondering before rocket, so while I agree with a lot of the post-rocket discussion, I don't think it's as relevant. Would "upgrading previous designs" be able to be seen as just a lack of foresight?

#

I suppose the extra 20% resource cost is more negligible on default settings too.

abstract flame
#

when you upgrade your smelting arrays from stone to steel, you can plop in a red belt to maintain ratios

glass dock
#

It's also an availability thing. If you have a lot of red belts thanks to automation, they're convenient to use. Why not use something you have in abundance. If you don't have red belts in abundance, yellow is definitely the answer.

Blue should be reserved for best optimization builds

brazen lynx
#

but I guess that is somewhere which isn't a lack of oversight

waxen thorn
#

Yeah, as j5 said. There are a lot of ways in the mid game (pre-rocket launch) to upgrade existing early game builds (stone furnaces -> steel, assembler 1s -> 2 or 3, adding modules, etc). When doing these upgrades, its fairly common that they will produce or consume more than the original yellow belt can move so red belt is an easy and worthwhile upgrade to make. Sure you could rebuild or place twice as many belts but that takes time that you could be using elsewhere.
I normally rush to get red belt and use it as soon as I have an abundance of iron production. And keep using reds primarily until very late game. (when all my builds are using beacons and mod3s)

brazen lynx
#

I've never had throughput issues, other than possibly smelting arrays, but even then getting the full iron ore throughput would be a huge problem

waxen thorn
#

If thats your plan then why use any better equipment at all... only use assembler 1s, normal inserters and yellow belt for everything. I mean you have enough space right?

#

Yes I know I took that to the extreme but that does feel like the argument you’re making.
If you look at the most common/popular designs for early game. Most of them are set up so they can be upgraded mid game to use better machines and belt. If you aren’t getting full iron ore throughput then it sounds like you need more iron outposts to me rather than think I wasted resources upgrading my base

#

That said, this is the beauty of the game. Play the way you want to play as there is no "right" way. Anyway I'm guessing this is probably just an agree to disagree topic. Have a nice day

glass dock
#

You can absolutely beat the game on mk1 assemblers and yellow belts. Speed runners do. Only time they tech up is to quickly produce those last few rocket parts

#

For speed runners it's just a matter of quick return for materials

#

Aesthetically though, I go MK2 everything until it's megabase time

brazen lynx
#

But I should go back just to check

brazen lynx
# waxen thorn Yes I know I took that to the extreme but that does feel like the argument you’r...

Sounds like you're really not following my argument. It's about cost versus benefit. Saying cost of red belts versus benefit doesn't seem worth isn't really the same as, say, doing the same for steel furnaces, which have pretty immediate payoffs. Like, those are wildly different ballparks. Red belts are more of a post-rocket thing, whereas steel furnaces are immediate, despite being unlocked at the same time.

brazen lynx
waxen thorn
# brazen lynx Sounds like you're really not following my argument. It's about cost versus bene...

I do believe I understand the cost argument. Red belts are twice speed as yellow but cost more than double to produce and with simply more space, you can produce the same products with yellow belts as you can red. (If I have this wrong please correct me).
My argument is that there is a missing factor... time. Larger builds take longer to place, more space means you need to clear out more area (biters, trees, etc) and just having a larger build requires you to travel farther (taking more time). Its very much the same reason for the new feature from the latest FFF. It enables scaling vertically rather than horizontally. The default gameplay of factory is a race against the biters so my argument is that time should absolutely be considered in the conversation.
I don't get what your argument is about upgrading to steel furnaces being dramatically different. Stone furnace (5 stone) to steel (20 stone, 30 iron ore) is arguable a more expensive upgrade compared to yellow (1.5 iron ore) to red belt (11.5 iron ore). Both upgrades double their respective abilities and both cost significaly more than double in resources. So by the cost argument I listed at the start of the post, why use steel furnaces when you can just build double the stone furnaces for much cheaper.
Final note: You are right. You absolutely can develop a base using yellow belts that takes you all the way to rocket. Speed runners can/do (though it depends on the category, see 100%) with only limited use of red. You were asking for the strategic standpoint of red belts, and thats what I was trying to provide with my first answer.

drifting minnow
#

@brazen lynx the red belt is "cheaper" if you consider the items on top of it
in the event that something is backing up on two yellow belts vs one red belt, the yellow belts can store twice as much, and depending on the scenario it can actually be better to use a red belt because of it's reduced "storage"

#

but if you're that concerned about buffer then you should use bots instead so you can control it directly, of course a massive generalization that doesn't apply to every scenario

#

just something to keep in mind

brazen lynx
drifting minnow
#

rarely

brazen lynx
#

ah okay

drifting minnow
#

you need to have the need for between one and two yellow belts of throughput exactly

#

and have a sufficiently expensive item that buffering too much of it would have a noticeable impact on how long it takes for that to happen

brazen lynx
#

Thanks for the response. It all comes down to benefit vs. cost. I suppose space is a hard thing to measure due to its other impacts, but generally I find myself rarely needing >1 yellow belt of throughput to an area pre-rocket. In a bus base, I suppose that might be useful, but bus bases are inefficient compared to starting spaghetti. The space argument is definitely something to be considered though, since there are still some places in any base that need a red of throughput.

Also, while the cost proportion is similar, it's more about the difference, plus needing to factor in the benefit. You get basically 15 stone and 30 iron ore for what when it comes to steel furnaces? It's for halved smelting costs, not to mention half the space. Red belts only give the latter, and only in specific situations. Not to mention the space savings is technically at least a 2x2, and even a 2x6 in most base designs due to added inserters/belts. So going purely by space, the cost/benefit tradeoff is 12x more on space, and just under 5x more on cost. And space isn't even the reason I go with steel smelters; it's the fuel. In speedruns at least, this allows me to completely skip a second coal patch; that's how good it is! If I were to put a number value on it, I'd say the fuel savings is like 3x more important than the space savings, but that's a debatable part for sure. So we have "36x more benefit for 5x more cost" (obviously a lot of simplifications were made, hence the quotes). That's like 7x more benefit, but wait! Red belts only save on space when you need exactly 2x more throughput. So... I think rounding up to 10x at least is reasonable. So... steel smelters are literally an order of magnitude better.

Of course, there does come a point where the payoff times necessitate better logistical items, but I'd say this is a much farther off time for red belts, even to be after launching the rocket.

#

I know a lot of the judgements above are wishy washy, but they seem the most accurate in my 4500 hours of play, and if it came down to it, I'd be able to back any of the individual points up with more evidence.

sacred bobcat
#

When you are in the late game and resources are functionally infinite, blue belts are the only ones worth using I think. Might be some setups where it's actually better to use a lower tier of belt but I feel like that'd be pretty jank

#

In the midgame, red belts are good because they're double a yellow belt and they don't cost oil which blue belts do. They only cost iron. You can just go to a massive iron patch and make it dedicated red belt production fairly easily. It's very low effort to getting your belt throughput doubled.

digital egret
#

If you are using productivity modules and beacons in back to back arrays you can only use 2 lanes of belts. Often you need to braid a red and blue belt on top of each other using undergroundys.

arctic laurel
#

long time no see! I'm still trying to improve my deathworld marathon speedrun and still sticking to yellow belts as well.

In my case, up to building blue science & roboports, red is soo expensive so yellow belt is the clear winner. When building purple/yellow, the cost of red belts is actually manageable. Besides cost, having different belt types makes it possible to over produce one type and run out of the other.

I didn't find enough upsides for red belts. A big chunk of the P/Y factory is actually like smelters, gc, rc, plastics... where I would just copy paste from the existing factory so yellow belts.

Although there are admittedly a couple places where red belts can save me some build time like connecting ores from train stop to smelters.

#

steel furnace is worth it because they save a lot of time - placing down more belts/inserters/poles for new smelting lines... but red belts don't save that much time

brazen lynx
#

Ye thanks for the input!

lilac merlin
#

Very sorry for bumping an answered question, and I have seen multiple answers. I would also agree for being in the yellow belt gang, red belts eat up SO MUCH GEAR I keep handfeeding my one blue assembler with gears instead of joining it to the main bus just so that it doesn't eat up and starve my entire gear belt. We'll all assume the needs of red belts vs yellow belts in this discussion is for pre-rocket silo i.e. starter base.

Use cases for red belts would include:

  1. Assuming the pre-rocket base does not use trains, having an expansion about 600 (this value is just an example, not a rule) or more belt lengths apart from the base would benefit a lot from the red belt's transport speed and capacity. 60 mining drills feeding directly to saturate one red belt will reach much faster to the base than two fully saturated yellow belts. We can split to the yellow belts in the base.

  2. Assuming space constraints due to spaghetti, for expanding the main bus, instead of adding another yellow belt, upgrading to red belt is a reasonable option. Case example would be using a single belt line of copper to (in order) green circuits, red circuits, then LDS. With yellow belts, it would be very likely that red circuits starve the copper line before LDS, so upgrading this one line to red belts would be beneficial.

brazen lynx
#

I’m not sure if I put the answered tag on just to be done with it, but I suspect it may have been a mod.

#

I guess it’s better to have the only two good answers (yours and Kevin’s) both arrive at the end since that’s what I’ll see when I come back here.

lilac merlin
# brazen lynx Tbh I didn’t consider the question answered until this response. (Any more than ...

One more which is usually avoidable with better designing is specifically using the extra range of the red underground belt for quick belting solutions where a yellow one can’t help, even if the belt line we’re fixing is a yellow one.

Having at least one or two pairs of them can be handy if I want to, let’s say, path a certain belt above water when I have no landfill, or above a thick cliff when I have no cliff explosives. Sometimes more resource and time efficient than pathing the belt away from the obstacles.

dull heart
#

If all those little intermediates that are usually on separate belts in your science and mall area have a shared sushi belt, well, you’ll be happy with red belt

#

Also the cost comparison for just basic iron plates at 30/second would be 19.5 iron for red belt vs 19 iron for 2 yellow belts

#

Pretty much As soon as you put prod modules in gears that makes red belt have the edge

#

I think it’s very important in these conversations to make sure you are discussing the properties of a useful belt (one that’s moving items) and not a crafted belt (moving no items).

#

After all, we want to move items with our belts

abstract flame
#

Why does putting a prod module in a gear make a difference?

#

ohhh in terms of gears to make the belt

arctic laurel
#

actually yellow belt costs 2.5 iron, red belts costs 22.5 iron. With prod 1 red belts is still over 8 times more expensive so not close in terms of cost

dull heart
#

not sure where you get those numbers from

dull heart
#

2 yellow belts moving plate: (1 gear + 1 iron plate + 16 iron plate buffer) = 19 iron. 1 red belt moving plate = (5.5 gears + 0.5 plate + 8 plate buffer) = 19.5 iron

arctic laurel
#

oh it was numbers for marathon. So for default, 2 yellow belt vs 1 red belt costs 3 vs 11.5. I see what you mean now\

lilac merlin
dull heart
#

Only as an example

#

No where did it say: only discuss expensive / marathon

dull heart
abstract flame
#

Putin is including the cost of items on the belt. Two yellow belts hold 16 iron plates. One red belt holds 8 iron plates. Putin is arguing that you should include that cost when considering the cost of the belt, which is imo reasonable.