Factorio Forums

LazyLoneLion wrote:

siggboy wrote:What's sad about it is that the space saving does not actually solve a problem. It only looks cool and well engineered.

What problem do you have in mind there?

The non-problem, the one that doesn't exist (lack of space).

siggboy wrote:The non-problem, the one that doesn't exist (lack of space).

Ah, that one. Got it.

Not really agree with you.
Space comes with the cost (quite low in fact, but still...). At least you have to defend your space with more walls and turrets and there are more resources laying on the (longer) belts. And it takes time when you are running from one edge of your factory to another (and your resources have to). And sometimes it takes resources to free some space of biter bases.

But mostly I don't like running through long bases. This game is enough time-consuming already.

And, besides that non-problem, this one module is really well-engineered, well-balanced.

Yes, the module is awesome, I have absolutely no complaints. And the description is really well written, even funny. Whoever made it knows what they're doing.

It's just that from a gameplay perspective, these super-engineered modules (there are also those for Oil and making blue belts and stuff) are not really important. It's more like winning a beauty-contest. This person definitely wins the beauty-contest in the science pack arena.

@LazyLoneLion It looks like your ratios will be off as you research lab efficiency.

Even if you converted everything to Yellow AM's I have you needing...
2 Sci 4
13 Sci 3
7 Sci 2
6 Sci 1

In Blue I have...
3 Sci 4
22 Sci 3
11 Sci 2
9 Sci 1

To feed the 20 Science Labs you have

It's odd to see more Red science than green science assemblers. In all the YouTube videos I've watched, they typically have five Red because it takes five seconds to make a red science and 6 green because it takes 6 seconds per green. Why do you have it with more red than green science?

Shokubai wrote:@LazyLoneLion It looks like your ratios will be off as you research lab efficiency.

First, it's not "my ratios" - it's not my project
Read the reddit description, it's really worth reading (and it's fun)!
What it says about Lab efficiency is mostly that labs are cheap. As assemblers are cheap as well. And science is quite fast without Labs efficiency. It was author's case.
But feel free to use your ratios as you like, of course.

garath wrote:It's odd to see more Red science than green science assemblers. In all the YouTube videos I've watched, they typically have five Red because it takes five seconds to make a red science and 6 green because it takes 6 seconds per green. Why do you have it with more red than green science?

You can read it in the reddit description:

blah blah 5:6 red:green assembler ratio The green modules are built in Assembler 2's, while the red modules are built in Assembler 1's. Since Assembler 2's are 50% faster than Assembler 1's, you get a 5:4 ratio.

Because when you build red science you have no blue assemblers.
Of course, you may replace those assemblers later, if you'd like, but in fact, there is really no need.

LazyLoneLion wrote:Of course, you may replace those assemblers later, if you'd like, but in fact, there is really no need.

...if you like not having enough production once you research lab efficiency

Shokubai wrote:...if you like not having enough production once you research lab efficiency

I suppose you just duplicate the whole project in this case. Using less than 20 labs. You'd have to count the precise number of labs by yourself then.
Good point is that production of Science-3 (as well as Science-1,2) will stay balanced even if it will not be enough for 20-30 labs with high Lab efficiency level.

It will not stay balanced if you upgrade assemblers to green -- there will be the shortage of the incoming resources.

Or you can forget about the precise balance and just make big enough production lines for the green and red circuit, for plastic, batteries etc. and feed them to the production line of Science assemblers as needed. As you, in fact, are proposing here.

LazyLoneLion wrote:

Shokubai wrote:...if you like not having enough production once you research lab efficiency

I suppose you just duplicate the whole project in this case. Using less than 20 labs. You'd have to count the precise number of labs by yourself then.
Good point is that production of Science-3 (as well as Science-1,2) will stay balanced even if it will not be enough for 20-30 labs with high Lab efficiency level.

It will not stay balanced if you upgrade assemblers to green -- there will be the shortage of the incoming resources.

Or you can forget about the precise balance and just make big enough production lines for the green and red circuit, for plastic, batteries etc. and feed them to the production line of Science assemblers as needed. As you, in fact, are proposing here.

I am proposing that you are maybe 1/4 supplied for 20 Labs. You might as well have just built 5 -10 labs and saved the space.

Shokubai wrote:I am proposing that you are maybe 1/4 supplied for 20 Labs. You might as well have just built 5 -10 labs and saved the space.

...OR I can build 30 labs, which are cheaper than learning all those Lab Efficiency
In the end all that counts is the quantity of science packs and production speed of those. Because any tech you learn takes particular count of science packs and not the specific time. Will you use 5 labs for 100 science packs or 30 for the same 100 packs -- there is no real difference while you can produce those packs and consume them. It's production of science packs that counts, not the performance of a single lab.

LazyLoneLion wrote:

Shokubai wrote:I am proposing that you are maybe 1/4 supplied for 20 Labs. You might as well have just built 5 -10 labs and saved the space.

...OR I can build 30 labs, which are cheaper than learning all those Lab Efficiency
In the end all that counts is the quantity of science packs and production speed of those. Because any tech you learn takes particular count of science packs and not the specific time. Will you use 5 labs for 100 science packs or 30 for the same 100 packs -- there is no real difference while you can produce those packs and consume them. It's production of science packs that counts, not the performance of a single lab.

THAT is a matter of opinion. On a scale of efficiency, assuming you never research Lab Efficiency, this works but will be a significant limiter for any game type where time is a scale you wish to shorten. Myself, even when not trying to speedrun, I want to get to specific techs ASAP. Doing so is faster with Lab Efficiency and more Sci Packs.

Not bashing the build...but its a niche build for the slow roll game type.

Shokubai wrote:Doing so is faster with Lab Efficiency and more Sci Packs.

Definitely NO.
It will take same amount of time or probably even less to get ANY tech with 30 Labs and Lab efficiency 0 then with 5 Labs with Lab efficiency 5. AFAIK Lab efficiency 5 gives ~250% speed boost, it means you'll learn a tech like with 13 Labs with Lab efficiency 0. AND you will have a surplus of science packs. AND you will need to get Lab efficiency 5 first (and it takes quite a lot of time and science packs itself).

You will NOT get tech much faster than you produce science packs (except if you learn it in one cycle). And it will take the same amount of science packs produced.

Instead you may build 20-30 Labs. It takes 36 Iron and 14 copper per lab. It's a laugh. Even Lab efficiency 1 takes (200 iron+100 copper) for scincepack-1 and (550 iron + 150 copper) for scincepack-2. It's enough for 18 Labs. Not speaking about Lab efficiency 2, 3 or 5...

https://wiki.factorio.com/index.php?tit ... (research)

LazyLoneLion wrote: Definitely NO.
It will take same amount of time or probably even less to get ANY tech with 30 Labs and Lab efficiency 0 then with 5 Labs with Lab efficiency 5. AFAIK Lab efficiency 5 gives ~250% speed boost, it means you'll learn a tech like with 13 Labs with Lab efficiency 0. AND you will have a surplus of science packs. AND you will need to get Lab efficiency 5 first (and it takes quite a lot of time and science packs itself).

You will NOT get tech much faster than you produce science packs (except if you learn it in one cycle). And it will take the same amount of science packs produced.

Instead you may build 20-30 Labs. It takes 36 Iron and 14 copper per lab. It's a laugh. Even Lab efficiency 1 takes (200 iron+100 copper) for scincepack-1 and (550 iron + 150 copper) for scincepack-2. It's enough for 18 Labs. Not speaking about Lab efficiency 2, 3 or 5...

https://wiki.factorio.com/index.php?tit ... (research)

This is my favorite part of this game...Throwing around numbers! My turn.

First, I will reiterate. I like the build. It fits for space savings and a slower game pace. I may use it myself on a minimalist game.

But...(you knew this part was coming)

Lets just pretend we want to research everything. Lets call that the victory goal.
*Edit* My friend here pointed out below that all research is not 30 seconds. I failed to mention that for the sake of simpler math(laziness?) I made everything 30 seconds. This would effect actual times but not the overall result since LE4 will reduce all times by half.

Costs...LE itself will add 14,875 copper and 20,000 iron to max. A very expensive price to pay for speed.
In total the LE full research tree will be
676,835 Copper
888,040 Iron Plate

In Total the non LE research will be
661,960 Copper
868,040 Iron

A difference of, this can't be right, ~2.2%. That seems...ABSOLUTELY CORRECT

Now you may argue materials to set this up but they are not as different as you think. For my tests we both researched Assy2 +Fast inserters right off the bat. Now I could have futs'd with numbers on your Assy1 Red packs since you don't need Assy 2 for those but we are only 1.5 minutes into research at this point so i called it fudge room.

The essential differences in our builds are this
Mine 5/6/11/10
Yours 5(assy1)/4/8/20
so we trade 5 Assy2 + 5 Assy1 for my extra setup cost + some mining/smelting for throughput.

Our output is
Mine 45/45/41 per minute
Yours 30/30/30 per minute

Times
It's going to cost me an additional ~15:20 to research to LE 4 This accounts for the steps in speed at each level. Meanwhile you're moving on...and were off to the races.
Now since the rest of the tree is static and you are researching with 20 labs at a speed of 2 cycles each per minute(remember you are 10 SciPacks/Minute short of supplying 20 Labs) and I am researching with 10 labs at a speed of 4 cycles per minute...lets math. Drumb roll please.
With a grand total of ~23 hours played....your build....is not first
That prize goes to the ole standby 5/6/11/10 with a total of ~17 hours

So tell me again how this makes you laugh.


OK, so here are some explanations why. As I said before you are under supplied. With a maximum output of 30 Science Packs per minute you are 10 short of supplying your 20 labs before any research is ever done. This more than anything else is the issue with your statements about LE on this build. It's interesting to me that LE is moot in terms of research speed when properly supplied. Assuming you fixed your supply problem to reach 40 SciPacks/Minute there is no difference in research time. Which is a bit disappointing really...You could fix pretty easily by adding 2 Greens, Turn Reds to Assy 2 and add 3 blues.

Shokubai wrote:The essential differences in our builds are this
Mine 5/6/11/10
Yours 5(assy1)/4/8/20
so we trade 5 Assy2 + 5 Assy1 for my extra setup cost + some mining/smelting for throughput.

Our output is
Mine 45/45/41 per minute
Yours 30/30/30 per minute

So we start from the point, where you have a bigger production line of the scince packs. Point to MY argument -- because I say that production of the sciencepacks is all that counts, not the LE

But that's not all.

Shokubai wrote:Times
It's going to cost me an additional ~15:20 to research to LE 4 This accounts for the steps in speed at each level. Meanwhile you're moving on...and were off to the races.

Meanwhile I can duplicate my project and have not 30/30/30 production, but 60/60/60, or 90/90/90, or more. Maybe each with less labs than in the original project (15 probably), if I don't need them, but labs are cheap anyway, so I probably just don't care.

Shokubai wrote:Now since the rest of the tree is static and you are researching with 20 labs at a speed of 2 cycles each per minute(remember you are 10 SciPacks/Minute short of supplying 20 Labs) and I am researching with 10 labs at a speed of 4 cycles per minute...lets math.

Well, if we're talking about tech with 30-second cycle (not all are 30-seconds), then production of 30 per minute is enough for 15 labs, not for 20, it's obviuos. For 60-second cycle it's 30 Labs. For starting techs like "Advanced electronics" it's 15-second cycle and we need only 8 Labs. For "Turrets" it's 10-sec cycle and 5 Labs. It's all in the reddit post.

If we're talking about sciencepack production speed (not consumption), then I may have 60 or 90 SP/min. And sufficient quantity of labs. And it won't take additional 15 minutes for me. And it will consume the same amount of input materials. Less, if you count SP needed for LE-4.

And my setup (let's call it "my") is balanced, consumes one standard belt of iron, one belt of copper, petroleum and a little of coal. So I can duplicate it and know that another "belt of iron" will be just enough. Yours -- I haven't seen yours. Is it compact? At least it definitely takes more than one belt of iron with your production. And less than two. And it will probably take more space (just because I haven't seen more compact design than this one).

Shokubai wrote:So tell me again how this makes you laugh.

I've tried.

Shokubai wrote:As I said before you are under supplied. This more than anything else is the issue with your statements about LE on this build.

No, this means that my project is of lesser scale than yours and more compact. And this means I maybe need fewer labs per design, not more assemblers. Even fewer if I get LE>0.

Shokubai wrote:It's interesting to me that LE is moot in terms of research speed when properly supplied.

It's easy, read the wiki.
LE gives speed of consumption, not an economy of the sciencepacks.

Shokubai wrote:This more than anything else is the issue with your statements about LE on this build. ... Assuming you fixed your supply problem to reach 40 SciPacks/Minute there is no difference in research time. Which is a bit disappointing really...You could fix pretty easily by adding 2 Greens, Turn Reds to Assy 2 and add 3 blues.

I hope I've already explained my point.
If I "add 2 Greens, etc." I will have to rebalance everything or I will NOT have bigger production (and therefore research speed) because of input shortage. For example, I'll need to add some parallel belts or replace the regular belt with a fast belt. Either way, it will not be as compact as it was. And if I really care about speed, then I can just duplicate the whole design. Usually, I don't care about science speed that much.

And finally I'd like to quote the original reddit post:

...
Plan to build 5-10 labs - 5 labs will saturate 10-second research, 10 labs will saturate 20-second research.
...
Plan to upgrade to 15 labs - most research from now on is 30-second.
...
Some projects are now 60-second - if you want to run those at full speed, extend to 30 labs, either via splitters and going north, or via two underground belts and going south.
...
Rocket Defense, and many of its prerequisites, are 60-second - it is definitely time to upgrade to 30 labs. Yes, this beast will keep 30 labs fully fed. Unless you research Lab Efficiency in which case I think you end up needing only 13 labs? Man, I don't know. Screw Lab Efficiency, labs are cheap, science is expensive, math is hard. I'm watching Die Hard right now and Bruce Willis just tossed an improvised bomb down an elevator shaft. It was **** metal. Just FYI.

LazyLoneLion wrote: So we start from the point, where you have a bigger production line of the scince packs. Point to MY argument -- because I say that production of the sciencepacks is all that counts, not the LE

...and not time...Which is perfectly fine for some folks

LazyLoneLion wrote: Meanwhile I can duplicate my project and have not 30/30/30 production, but 60/60/60, or 90/90/90, or more. Maybe each with less labs than in the original project (15 probably), if I don't need them, but labs are cheap anyway, so I probably just don't care.

Couldn't we both though. Kinda defeats the purpose of testing anything if your argument is "double it til it works" The ratios remain the same and still comes out (percentage wise) to the same end.

LazyLoneLion wrote:Well, if we're talking about tech with 30-second cycle (not all are 30-seconds), then production of 30 per minute is enough for 15 labs, not for 20, it's obviuos. For 60-second cycle it's 30 Labs. For starting techs like "Advanced electronics" it's 15-second cycle and we need only 8 Labs. For "Turrets" it's 10-sec cycle and 5 Labs. It's all in the reddit post.

Reducing your labs to match your production would still cost you time because you still research twice ass slow.

LazyLoneLion wrote: If we're talking about sciencepack production speed (not consumption), then I may have 60 or 90 SP/min. And sufficient quantity of labs. And it won't take additional 15 minutes for me. And it will consume the same amount of input materials. Less, if you count SP needed for LE-4.

I actually said this specifically. 20 Labs is equivalent to 10 @ LE4 given proper supply. Selective Reading maybe?

LazyLoneLion wrote:And my setup (let's call it "my") is balanced, consumes one standard belt of iron, one belt of copper, petroleum and a little of coal. So I can duplicate it and know that another "belt of iron" will be just enough. Yours -- I haven't seen yours. Is it compact? At least it definitely takes more than one belt of iron with your production. And less than two. And it will probably take more space (just because I haven't seen more compact design than this one).

Correct on all counts. Maybe you're starting to get it?

LazyLoneLion wrote:

Shokubai wrote:So tell me again how this makes you laugh.

I've tried.

Failing MISERABLY.

LazyLoneLion wrote:

Shokubai wrote:As I said before you are under supplied. This more than anything else is the issue with your statements about LE on this build.

No, this means that my project is of lesser scale than yours and more compact. And this means I maybe need fewer labs per design, not more assemblers. Even fewer if I get LE>0.

OH HERE is that quote. We go back to the point of my argument that I can get places more quickly with LE4. Less of anything in your build widens the time gap. I think the selective listening part of your brain can't get passed this point.

LazyLoneLion wrote:

Shokubai wrote:It's interesting to me that LE is moot in terms of research speed when properly supplied.

It's easy, read the wiki.
LE gives speed of consumption, not an economy of the sciencepacks.

You keep saying this but it is not in debate...Yes...it costs something to research LE...and yes it doesn't make anything cheaper.

LazyLoneLion wrote:I hope I've already explained my point.
If I "add 2 Greens, etc." I will have to rebalance everything or I will NOT have bigger production (and therefore research speed) because of input shortage. For example, I'll need to add some parallel belts or replace the regular belt with a fast belt. Either way, it will not be as compact as it was. And if I really care about speed, then I can just duplicate the whole design. Usually, I don't care about science speed that much.

Your poorly defended point hasn't really ever been at issue. This is a nice design. It is well thought out and planned for balance. It fits nicely in a compact base with a strict set of resource throughput. I believe I've said this before.

My point that you keep missing is that it is inherently limited by it's balance and will, in an efficiency measure of time, be a limiting factor to those individuals who wish to build on a larger scale and spend time on a smaller one.

LazyLoneLion wrote:And finally I'd like to quote the original reddit post:

...
Plan to build 5-10 labs - 5 labs will saturate 10-second research, 10 labs will saturate 20-second research.
...
Plan to upgrade to 15 labs - most research from now on is 30-second.
...
Some projects are now 60-second - if you want to run those at full speed, extend to 30 labs, either via splitters and going north, or via two underground belts and going south.
...
Rocket Defense, and many of its prerequisites, are 60-second - it is definitely time to upgrade to 30 labs. Yes, this beast will keep 30 labs fully fed. Unless you research Lab Efficiency in which case I think you end up needing only 13 labs? Man, I don't know. Screw Lab Efficiency, labs are cheap, science is expensive, math is hard. I'm watching Die Hard right now and Bruce Willis just tossed an improvised bomb down an elevator shaft. It was **** metal. Just FYI.

What you miss here is that LE4 makes those times 5, 15, and 30 seconds.

Look. Your grasp of this is tenuous at best and I appreciate your defense of the build(which i very much like) but your failure to grasp it's limitations brings me to leaving this thread with you. You will no doubt post on and quote lots of stuff from this post and unless it adds any new information or you actually prove some of these points wrong I think I've said my peace.

One last thing
Efficiency
noun, plural efficiencies.
1.
the state or quality of being efficient, or able to accomplish something with the least waste of time and effort; competency in performance.
2.
accomplishment of or ability to accomplish a job with a minimum expenditure of time and effort:
The assembly line increased industry's efficiency.
3.
the ratio of the work done or energy developed by a machine, engine, etc., to the energy supplied to it, usually expressed as a percentage.

I apply this definition to our situation like this.
This build - Expends time to accomplish a job with minimal raw materials & minimal space.
Standard 5/6/11/10 - Expends materials (and space?) to accomplish a job with minimal time

Til next time.
~Shokubai

Shokubai wrote:I apply this definition to our situation like this.
This build - Expends time to accomplish a job with minimal raw materials & minimal space.
Standard 5/6/11/10 - Expends materials (and space?) to accomplish a job with minimal time

"...to accomplish a job with minimal time" ...implying that researching LE4 is faster than just building more assemblers and labs.
Yes, this design has excessive labs for the 30-second cycle (if you have no backlogged scincepacks). No, it has NO difference with your design in terms of research speed for all techs. No, LE has nothing related to research speed AT ALL. It ONLY lets you have fewer labs, which is no problem anyway.

Having 20 labs instead of balanced for the 30-second cycle 15 labs doesn't slow me down (compared to 15 labs). And it doesn't slow me down comparing to 6,25 Labs with LE4 (thus 140% speed bonus, 240% in total, equal to 15 Labs). However, it is slower comparing to fully supplied 100 Labs. Will it be more compact than this design for 15 labs? Definitely no. Can I beat 100-labs design with this design? Easily - just duplicate as needed. Is this design "LE-ready"? Totally. You don't even have to remove excessive labs because excessive labs don't slow you down, they will just idle without supplies. Can I add more assemblers here (to have 5/6/11 for example)? I can. Will it help? No, because compactness will be lost. Can I duplicate it? Yes, totally. Will it increase the production and help to end the game sooner? Yes.

You can use your "standard build 5/6/11" without LE with more labs and have the same winning time as with LE4 and fewer labs. It will even take LESS time (for 15 minutes less, as you say). Just because you don't have to research LE4. Building twice as many labs takes seconds instead of minutes and a small amount of basic resources.

But you have NO "standard 5/6/11" design as compact and well-engineered as this one.

Your "Standard 5/6/11 design" is worse than abstract "8/9/17" design by your (!) arguments. Why do you stop at 5/6/11 then? Why not 10/12/22? Why not compare 100/120/220 with this compact 6/5/8 design? Because it works both ways: if you choose "5/6/11" to compare with this "6/5/8" -- I can compare your "standard 5/6/11" to my "double 12/5/16" and win the argument. If just "bigger is better" then there is no limit with this design too, and it will still win because it's more compact. And it doesn't depend either on Labs count (if it's sufficient -- and it is!) or on LE level.

There is only one option against this design regarding LE (and no one besides LE). If and when you learn LE4 you can build only one column of labs instead of two (7 instead of 15, or 13 instead of 30 for the 60-sec cycle). And so you'll have more compact design again. Why would you want to research LE4 -- I don't know. LE doesn't speed up the game, it only takes 15 minutes of time.

Tell me -- why, really?

Shokubai wrote:Couldn't we both though. Kinda defeats the purpose of testing anything if your argument is "double it til it works" The ratios remain the same and still comes out (percentage wise) to the same end.

You see, in my eyes it's you who does "double it til it works". You just use more production, it's less compact, but you say "it's better". Well, if we can use less compact, then 2x"6/5/8" is better then your "standard 5/6/11". So, what' the argument?

Shokubai wrote:Reducing your labs to match your production would still cost you time because you still research twice as slow.

Not twice, but 41/30. It's like 1.5 time slower. And I'm not reducing labs count for speed. I'm saying you can build fewer labs from the start until your tech needs 60-sec cycle. For speed I'm saying you build more assemblers, and it's better to build more assemblers duplicating this design, than just using something imbalanced kinda your "15/17/3337" design (which will still be faster than this "6/5/8").

Shokubai wrote:I actually said this specifically. 20 Labs is equivalent to 10 @ LE4 given proper supply. Selective Reading maybe?

If it's equivalent, then what's the point in researching LE4 if you can just use 20 labs and save 15 minutes and thousands of materials?

Shokubai wrote:

LazyLoneLion wrote:I haven't seen yours. Is it compact? At least it definitely takes more than one belt of iron with your production. And less than two. And it will probably take more space (just because I haven't seen more compact design than this one).

Correct on all counts. Maybe you're starting to get it?

Yours takes more input, it's less compact, and you admit it. What's the point in speaking about different ratio and LE4 than? If you have more space to build your less compact design, then why not build my more compact design twice? You will still save 15 minutes and lots of materials. AND space.

Shokubai wrote:Less of anything in your build widens the time gap.

No.
15 labs instead of 20 labs does not widen the time gap. Not until you hit 60-second cycle, and then you'll need 30 labs even in this project (as it is directly said).
So I'm saying "MAYBE you NEED fewer labs [sometimes]". It's about short research cycles. You don't *need* 20 labs *for that*, 15 will be enough. But even in that case 20 instead of 15 don't slow you down anyway. 20 labs are not a problem either with or without LE. As any sufficient (or excessive) quantity is not a problem as well.

Shokubai wrote:My point that you keep missing is that it is inherently limited by it's balance and will, in an efficiency measure of time, be a limiting factor to those individuals who wish to build on a larger scale and spend time on a smaller one.

Don't see how it is so.
If "larger scale" assumes "more space and input belts" possible, then there is the obvious and effective way to increase research speed with this design. And not so obvious with your abstraction "5/6/11".
And you are saying that using some design (probably with LE4) it's possible to be more effective, than using only this design without LE? Because the word "design" means you can repeat it in any game any number of times.
And the word "compact" means that it gives more production per space unit than yours one.
If the goal was "to make as small research design as possible" it would be definitely "1/1/1/1" design (better than your "5/6/11/10").
If the goal was "to make the fastest research possible" it would be kind of "1000 000/1000 000/1000 000/1000 000" design (again better than your "5/6/11/10"). Your 5/6/11 is "a limiting factor" there, so what's the point in using it? BTW that "1000 000/..." design might happen to be just duplicated "6/5/8" design, but I doubt it will be "5/6/11".

This "6/5/8" design is smaller, could be built faster and will give more research then your 5/6/11 on any given space and time big enough to fit them both. Where is the point where your design becomes "the best"?

Shokubai wrote:What you miss here is that LE4 makes those times 5, 15, and 30 seconds.

So, what? You don't really need LE in case of speed run. You don't really need it in any case. At least it's VERY hard to imagine such a case.
Because it's easier and more time-saving to NOT research LE4 and just build more labs instead.

I see three arguments from you against this design:
1) there is not enough production for 20 labs.
2) there is even less enough production in case of LE4 researched
3) 5/6/11 produces more than 6/5/8

Correct me if I'm wrong, please.

And here are my (obvious!) answers:
1) this design is not supposed to be balanced for 20 labs. They are sufficient (but not necessary) for 30-sec cycle (and for 10-15 sec cycle) and just fit into the rectangle. And you'll need more than 20 labs for 60-sec cycle and labs are cheap, so you will even build 10 more later.

2) see answer 1), and LE is inefficient -- it "speeds up research" but you will lose more time and materials to research it, comparing with just adding more labs (in which case you'll easily catch up with any LE tech, not losing a second, but *saving* time).

3) ...and 6/7/12 produces more than 5/6/11, so what? Where is The Line? And what is the point?

Yours takes more input, it's less compact, and you admit it. What's the point in speaking about different ratio and LE4 than? If you have more space to build your less compact design, then why not build my more compact design twice? You will still save 15 minutes and lots of materials. AND space.

To answer your question I'll remind you that I said you might have well have just built 5-10 labs (turns out its 15. Yay math). This was so your usage matched your production. Your reply was to just build 30...because being able to use twice as much of the thing your not making makes sense. Then we got off on all these tangents about LE or not LE and double this or halve that. So i decided why not test THIS Build vs the standardized accepted balanced ratio build for LE4 and somehow none of it is relevant to you. So...

Just for fun I built it. Obviously used Red belt and Fast Inserters throughout for throughput. I probably broke something somewhere in there but testing it on Follower Count 17 it seemed to be holding nicely until my Oil run out because i forgot to turn off my main bases Plastic and Sulfur.. I was happy enough with it and I feel like the ratios are PDC. *Insert Disclaimer of how thoroughly untested this is here* Space differences Overlay

Leon wrote:this is the best automated 4x science pack setup

4x science pack

Can we get a blueprint string for this? I have been having so many issues coming up with efficient setup.