Friday Facts #409 - Diminishing beacons

[guy-732]

Why is alt mode showing the modules "inside" the beacons now?

First image from FFF

That was always a thing, even in the current version.
There is a setting to make alt mode display modules in beacons that way.

[Haighstrom]

The post doesn't really explain how the new beacon maths affects the "old" builds. You've clearly said that single beacon builds work nicely, but have you solved the original goal stated? Is 1 assembler and 12 beacons now less productive than 5 assemblers and 7 beacons? What's the optimal production : beacon ratio now (without quality)?

I'm pretty ambivalent about the change in general, I get the intent, but I'm a bit worried about the lack of transparency with this change (which you noted). It's one more thing (like fluid pipe length) that you just can't understand without a calculator, which I fundamentally dislike as a concept.

[coppercoil]

Who said "using beacons seems restrictive on build variety and tends to feel monotonous and uninteresting"?

screenshots

#1


#2

[kitters]

With just the 2-tile gap between the beacons and the machines, you can only route very few transport belts through. There are a few tricks that help getting up to 2 transport belts in and 1 transport belt out, but once you've discovered these tricks, you just repeat them over and over.

Meanwhile that one mother**cker after spending a week working on design five years ago (me):
*manages to fit 4 belts in and 5 belts out per 10 tiles row of 62 machines

[husnikadam]

Will you increase the pipe troughput? With so many buffs to belts, machines and beacons, pipes now seem to be a heavy bottleneck. Having either parallel pipes or pumps to force higher troughput to many thousands/s is very annoying

In https://factorio.com/blog/post/fff-402 the foundry crafts 260 times per second and I suspect that every craft needs 10 fluid (as max machine input fluid capacity is 20). That makes it require 2600 input fluid per second. For a single machine. Ludicrous indeed

[planetfall]

Okay, I don't want to be a downer, but something in these posts keeps driving me crazy.

These FFF posts keep mentioning the huge benefits you can get from using quality entities in what I can only really describe as a "hype-builder" voice. It's like someone saw the initial ambivalent reaction to the quality mechanic and sent out a memo that for future posts it needs to be hyped up as much as possible. This has taken the form of showing off all the power that high-quality entities bring.

This grates, because I think from what has been shared, the benefits of quality are without question. The problem is that, as I understand it, the solution to all quality production is to just do the exact same recycle loop over and over, for every item in the mall. You can just do the "take what you get without recycling" method but this does not enable you to use quality items at scale, and seems like it would create enormous inventory bloat and cognitive load if you could not standardize on 1-2 quality levels.

As someone who is a fan of how quirky and personal belt malls are - everyone has their own blueprint for how to make all tiers of belts or inserters, for example, compared to something like smelting arrays which are extremely static - I do not like the idea of tearing all that down and replacing it with a boring array of standardized cells, albeit more complicated cells than current bot malls. Yes, I know quality is optional, you don't have to use it, yada yada, but "you don't have to use the boring mechanic" is not really a good defense of putting a boring mechanic in your game, is it? You're really just selling just how much benefit you're leaving on the table by ignoring this mechanic that seems by all rights to be incredibly tedious.

It is especially baffling to see this post go on about how you wanted to redesign beacons so the game is less about stamping down standardized identical cells (An admirable design goal! requester chest nerf when.) and then dangle the benefits of quality beacons in front of us, when quality mall production seems for all the world to be about stamping down standardized identical cells.

I have this disconnect with the majority of the factorio players - or perhaps just the most vocal ones - where I want to convert time I won't get back into time I won't regret spending differently by fiddling with spatial doohickies and solving new puzzles, and the modal factorio player seems to want to convert imaginary rocks into imaginary beakers by reducing the game to a series of repeatable processes and then grinding out those processes forever until their computer breaks. This is probably rude and uncharitable to all the challenge runners, pyanodon masochists, etc, but the hyping of all the tremendous things you can do with quality feel both forced and uncompelling. Look at how much you can make the numbers go up!

I KNOW you can make the numbers go up by selling out to The Same Recycler Loop Over And Over, that's exactly the problem. Quality looks like it's going to be requester chests v2.0 where it de-dimensionalizes a major aspect of the game and you have to pick between Big Number and the game still being enjoyable, with a community that is always crowing about how Big Number is the only thing that truly matters. You keep trying to sell quality as powerful and completely failing to sell quality as interesting.

[husnikadam]

I predict that the next blog post is the third planet

[Edriel]

I'm not an expert on calculations and probably I didn't understand the post or didn't get the point but this change basically changes nothing.

What do I mean?
I mean that even if getting less/more beacons affecting a machine is more/less effective than vanilla, in the end I'll always want a 12/8 (16 with furnaces) beacons setup, so what's the point of the first paragraph where you say "We aren't so sure the resulting layouts are all that interesting - something both us and a lot of players have said many times before."?

The layouts are gonna be the same anyway, so again what's the point of this change?

To me, the Earendel's solution in his mod with the "wide-area beacon" made much more sense if your purpose was to make the player redo the factory layouts and in a "interesting way".

If I'm missing the point can someone explain me why this is a good change and mostly why this was "necessary" and not just a waste of time?

P.S. I get that in "early stages" this is a nice buff, but you all know that the important part is the end-game progression.

[Terrahertz]

I predict that the next blog post is the third planet

Likely, but we only have two reference points

[morsk]

This is a fake solution imo. It's still 12-beacon endgame.

I realize anti-nerf culture makes removing 12-beacon impossible. I expected devs to do nothing. This is better than nothing, so I should be happy. My best ideas were some change to layouts that made beacons smaller, or let belts overlap them. I wouldn't dare take 12-beacon away from people, only ask developers make it less ugly. In hindsight, now that I've seen this non-change, I would. I would use quality to buff 4-beacon past the current 12-beacon, make 4 beacons the hard cap, and say it doesn't count as a nerf. Oh well.

But this will probably make me use some 4-beacon builds in real games, so it is improving things for me.

[Mooncat]

I'm not an expert on calculations and probably I didn't understand the post or didn't get the point but this change basically changes nothing.

What do I mean?
I mean that even if getting less/more beacons affecting a machine is more/less effective than vanilla, in the end I'll always want a 12/8 (16 with furnaces) beacons setup, so what's the point of the first paragraph where you say "We aren't so sure the resulting layouts are all that interesting - something both us and a lot of players have said many times before."?

The layouts are gonna be the same anyway, so again what's the point of this change?

To me, the Earendel's solution in his mod with the "wide-area beacon" made much more sense if your purpose was to make the player redo the factory layouts and in a "interesting way".

If I'm missing the point can someone explain me why this is a good change and mostly why this was "necessary" and not just a waste of time?

P.S. I get that in "early stages" this is a nice buff, but you all know that the important part is the end-game progression.

I think I'm on the same page with you.

Even with diminishing beacons, the transmission effect is still getting better when more beacons are affecting the same machines, and therefore ultimately everyone will still keep the same layout for the best UPS.

I like the idea of Beacon Overload, but also agree that placing the beacons is a bit tricky with that.

Instead of making the effect transmission to be the square root of the beacon count, why not just make it to be the quotient (not sure if it is the right word) of the beacon count, so that
1 beacon = 100% effect
2 beacons = 50% effect from each beacon, i.e. the same as placing 1 beacon
3 beacons = 33.333% effect from each beacon, i.e. still the same as placing 1 beacon
Beacon Overload, but without the problem of making the machines completely inactive.

Edit:
And with Quality Scaling, the equation can be something like

Code: Select all

transmission_effect = (100% / (beacon_count <= 1 ? 1 : (beacon_count / beacon_quality))

so,
Q2 allows 2 beacons working together without any lose, and
Q3 allows 3 beacons working together without any lose, etc.

Another edit:
Ah ok, on second read, I see

In 2.0, maximizing beacon count will still generally be a good practice to have things produce quickly and save on computer performance. However, there will now be a lot more room for creativity, and with quality return on investment comes into question more.

Now I know the devs have no intention on completely removing the maxed beacon layouts. I think it is fine. Not like the game is unplayable. Just felt weird why the equation was like that if they didn't want to see those layout again.

[Terrahertz]

I'm not an expert on calculations and probably I didn't understand the post or didn't get the point but this change basically changes nothing.

What do I mean?
I mean that even if getting less/more beacons affecting a machine is more/less effective than vanilla, in the end I'll always want a 12/8 (16 with furnaces) beacons setup, so what's the point of the first paragraph where you say "We aren't so sure the resulting layouts are all that interesting - something both us and a lot of players have said many times before."?

The layouts are gonna be the same anyway, so again what's the point of this change?

To me, the Earendel's solution in his mod with the "wide-area beacon" made much more sense if your purpose was to make the player redo the factory layouts and in a "interesting way".

If I'm missing the point can someone explain me why this is a good change and mostly why this was "necessary" and not just a waste of time?

P.S. I get that in "early stages" this is a nice buff, but you all know that the important part is the end-game progression.

It might have something to do with UPS, because the design with the most beacons might no longer get you the most production per UPS.

Using as much beacons as possible is always at odds with using direct feeding as much as possible.

In the end this seems to be easily modable so if you don't like it, change it.

[Alsadius]

I KNOW you can make the numbers go up by selling out to The Same Recycler Loop Over And Over, that's exactly the problem. Quality looks like it's going to be requester chests v2.0 where it de-dimensionalizes a major aspect of the game and you have to pick between Big Number and the game still being enjoyable, with a community that is always crowing about how Big Number is the only thing that truly matters. You keep trying to sell quality as powerful and completely failing to sell quality as interesting.

I can see this argument, but I don't agree. I'm fairly new to the game, as these things go (~100 hours), but I can't say I've ever felt much pressure to use requester chests, and the descriptions of quality make it sound the same. Both seem like the sorts of things you want to use for low-quantity work, with anything you need in bulk going by train/belt and not chest, or going for lower quality and not higher. It's like tier 3 modules - until you're up at megabase levels, you pick and choose where to use those, because you don't have enough to go around. And this isn't a laziness thing or a roleplay thing. They're just too dang expensive to use in any real quantity without crippling your base. From the way they've described it, quality will be the same.

Also, I'm surprised to hear the community reaction to quality was lukewarm. It sounded pretty cool to me, and I'm rather excited for it.

I'm not an expert on calculations and probably I didn't understand the post or didn't get the point but this change basically changes nothing.

What do I mean?
I mean that even if getting less/more beacons affecting a machine is more/less effective than vanilla, in the end I'll always want a 12/8 (16 with furnaces) beacons setup, so what's the point of the first paragraph where you say "We aren't so sure the resulting layouts are all that interesting - something both us and a lot of players have said many times before."?

The layouts are gonna be the same anyway, so again what's the point of this change?

To me, the Earendel's solution in his mod with the "wide-area beacon" made much more sense if your purpose was to make the player redo the factory layouts and in a "interesting way".

If I'm missing the point can someone explain me why this is a good change and mostly why this was "necessary" and not just a waste of time?

P.S. I get that in "early stages" this is a nice buff, but you all know that the important part is the end-game progression.

Cost efficiency matters. If the 12th beacon is doing 30% as much as the 1st, but it still costs 100% as much resources and power, I'm a lot more likely to switch to 8-beacon builds, or maybe even less. Maybe I take the 8-beacon builds and remove every other row of beacons, so it's only 4 now. That's almost as good as an 8-beacon build in 1.1, but for way less investment, and I can pack it in tighter (which is sometimes pretty handy). For things where speed isn't *that* important, that's fine.

Not everyone is a UPS-limited gigabase. (And if you're building at that scale, samey builds are just fine, because you're just stamping down blueprints a million times anyway - you're not focused on designing every little assembler individually, it's all about bulk automation.)

[thermomug]

1 beacon = 100% effect
2 beacons = 50% effect from each beacon, i.e. the same as placing 1 beacon
3 beacons = 33.333% effect from each beacon, i.e. still the same as placing 1 beacon
Beacon Overload, but without the problem of making the machines completely inactive.

This does not make sense, what you mean maybe is using the infamous Inverse-Square function:
1 beacon -> 100%
2 beacons -> 100% / sqrt(2) = 0.707...
3 beacons -> 100% / sqrt(2) = 0.577...
4 beacons -> 100% / sqrt(4) = 0.5

[Brathahn]

a thing about beacons i usually had trouble with: when you dont use them they still consume power. so the obvious solution is to use power switches. but with that you get multiple screens worth of blinking "power not connected" warning signs on the beacons/assemblers. it would be really nice to have the game be smart enough, when a power grid is disconnected by a power switch to not show that blinking warning sign.

[Alsadius]

1 beacon = 100% effect
2 beacons = 50% effect from each beacon, i.e. the same as placing 1 beacon
3 beacons = 33.333% effect from each beacon, i.e. still the same as placing 1 beacon
Beacon Overload, but without the problem of making the machines completely inactive.

This does not make sense, what you mean maybe is using the infamous Inverse-Square function:
1 beacon -> 100%
2 beacons -> 100% / sqrt(2) = 0.707...
3 beacons -> 100% / sqrt(2) = 0.577...
4 beacons -> 100% / sqrt(4) = 0.5

The one you suggest is what'll be in game, but the one that Mooncat suggested is their preferred alternative. The suggestion they were making was to make it so that you could only have one beacon's worth of effects apply to any given building, but without the Space Exploration "it shuts down with 2+ beacons" thing, which feels pretty punitive.

[Mooncat]

1 beacon = 100% effect
2 beacons = 50% effect from each beacon, i.e. the same as placing 1 beacon
3 beacons = 33.333% effect from each beacon, i.e. still the same as placing 1 beacon
Beacon Overload, but without the problem of making the machines completely inactive.

This does not make sense, what you mean maybe is using the infamous Inverse-Square function:
1 beacon -> 100%
2 beacons -> 100% / sqrt(2) = 0.707...
3 beacons -> 100% / sqrt(2) = 0.577...
4 beacons -> 100% / sqrt(4) = 0.5

But it is just the same as what the FFF have said?

With this math,
2 beacons = 41% better than 1 beacon
3 beacons = 73% better than 1 beacon
4 beacons = 100% better than 1 beacon
n beacons = m% better than 1 beacon, where m = (1/sqrt(n)*n-1)*100
Yes, when n keeps growing, the growth of m will be almost 0 eventually. But with the vanilla beacon coverage, I don't think we can reach that point.

So, the question is, why should I stop using multiple beacons to boost the same machines?

[FasterJump]

I like to build my end-game design to optimize the energy cost per craft (assuming maximum productivity modules). In 1.1, that means rows of beacons and machines (1:8 and 8:1), which is boring.

The diminishing ratios sounds great. But I'm afraid that the most efficient designs will still be rows of beacons in endgame.

But there is a way to prevent that! All that needs to be done is increase the base beacon energy consumption. Double it! (maybe make it scale with the number of machines serviced). By doubling or tripling the base beacon energy consumption, the optimal count of beacon source per machine will become closer to 5:x / 7:x, rather than 8:8.

[Zlutz]

What if we put 12 beacons and load 6 of them with speed and 6 of them with efficiency, do we get 2.45 of each effect or 3.46 of total?

[y.petremann]

Agree 100% with this. You boys realize that you have now effectively required the use of online calculators for mere mortals to get accurate throughput numbers? Since calculators are now a requirement, are you including one in-game?

Or use Helmod ...