How is "SPM" measured?

Post all other topics which do not belong to any other category.
User avatar
T-A-R
Fast Inserter
Fast Inserter
Posts: 156
Joined: Tue May 22, 2018 4:20 pm
Contact:

Re: How is "SPM" measured?

Post by T-A-R »

I would like a standardized catagorisation, like how it is done with speedruns, to specify certain standard requirements.
It shouldn't dictate a single way of playing, since Factorio is a sandbox game, which is enjoyed in many ways.
World settings are noteworthy for their respective credit, SPM however is a isolated factor.
the importance lies in the ratio of producing science packs. I see that as following:

Production statistics are read from the production graph, production should meet consumption, in a continious (input = output) matter. Both with, or without military science. All research packs (including space science and military) should be able to be consumed in the correct ratio, buffering science packs should have no use in SPM stats. (due to the ever growing amount of required packs). Imo the wiki is too vague here.

Cheating is fine, Factorio is a sandbox game pur sang. Examples: infini chests, ore patch creation, water creation, power interfaces.
But not the same class pure vanilla run including pollution and expanding biters.
Above 10K spm i think the main challenge is bringing in full ore trains, and sending empty ones back without clogging your rail network. But i would not want to force other to build a rail network just to legitimize your "official" SPM specification.

If any base has a certain SPM you should not need to build in a certain way to meet the classification. the SPM classification is based on the lab consumption. Since the labs are the important factor that all self respecting factories are sharing.

Edit:
A base with infinity chests etc. could also be phrased as a incomplete factory (a sub system, a proof of concept). Some parts are simulated.
This simulating may result in "unrealistic" situations which can't be reproduced in a regular vanilla way of playing (when you have to feed it "real" ore).
It is a matter of opinion (is it?) to call the design flawed/unrealistic if it can't. (personally i'm quite strict, but that is personal and i don't wanna force others here, each set their own goals here).
Last edited by T-A-R on Fri Apr 02, 2021 6:16 pm, edited 1 time in total.

pichutarius
Fast Inserter
Fast Inserter
Posts: 154
Joined: Mon Feb 27, 2017 12:47 am
Contact:

Re: How is "SPM" measured?

Post by pichutarius »

how is spm measured? press P (production tab) and read xxx/min.

no matter 1 big factory or several many identical blocks, every SPM is as real as every others.

if we think otherwise, there will be grey area like putting blocks very close to each other and swap some factories to make them not identical and "1 big factory".

it seems like you dont like same block copy pasted and claim high SPM, you feel like that doesnt count. thats fair, if so we probably want another metric to measure blockyness vs big factory. there probably is a clever way to do this but i dont know.

a stupid attempt: maybe find the area of bounding boxes (https://en.wikipedia.org/wiki/Minimum_bounding_box) of labs, furnaces, assembly. calculate SPM ÷ total area of bounding boxes. blocky has spread structure, has big bounding box, and big factory tend to clump same structure, and small bounding box. it works but a headache to calculate.

https://imgur.com/z1Y2FcT
the left side is 4 block copy pasted, the right side is 4 outpost producing different item, both use same area, producing same rate.

Trific
Fast Inserter
Fast Inserter
Posts: 118
Joined: Thu Dec 31, 2020 7:57 pm
Contact:

Re: How is "SPM" measured?

Post by Trific »

pichutarius wrote:
Fri Apr 02, 2021 12:12 pm
it seems like you dont like same block copy pasted and claim high SPM, you feel like that doesnt count.
Sure, it counts, but it's generally uninteresting. It might be interesting if someone were doing the copy pasting technique in order to reduce or eliminate variability in testing UPS impact of various builds, and then discussing their findings could be interesting. But IMO, "here's my 50K spm base of 300 copy pastes of the same module" is uninteresting.

User avatar
ptx0
Filter Inserter
Filter Inserter
Posts: 988
Joined: Wed Jan 01, 2020 7:16 pm
Contact:

Re: How is "SPM" measured?

Post by ptx0 »

Trific wrote:
Fri Apr 02, 2021 1:06 pm
pichutarius wrote:
Fri Apr 02, 2021 12:12 pm
it seems like you dont like same block copy pasted and claim high SPM, you feel like that doesnt count.
Sure, it counts, but it's generally uninteresting. It might be interesting if someone were doing the copy pasting technique in order to reduce or eliminate variability in testing UPS impact of various builds, and then discussing their findings could be interesting. But IMO, "here's my 50K spm base of 300 copy pastes of the same module" is uninteresting.
if that's all they claim there is to it, sure. their story is what should make it interesting. why did they design the module the way they've done it?

it seems like you completely disregarded everything i wrote in my earlier comment.
My Mods - Fish Per Minute base size metric
• • •
Base: Bob's @ 1 Million SPM
• • •
Linear search and overflows are indicative of sloppy coding practices.

Trific
Fast Inserter
Fast Inserter
Posts: 118
Joined: Thu Dec 31, 2020 7:57 pm
Contact:

Re: How is "SPM" measured?

Post by Trific »

ptx0 wrote:
Fri Apr 02, 2021 5:26 pm
if that's all they claim there is to it, sure. their story is what should make it interesting. why did they design the module the way they've done it?

it seems like you completely disregarded everything i wrote in my earlier comment.
The module might be interesting (I remember seeing one that was very small and successively iterated using Recursive Blueprints to eventually launch a rocket, and I found that very interesting), but it's usually about the MaSsiVe SPM!

I don't have time to read every comment. If I didn't quote you or address you, then I'm not responding to something you said.

User avatar
ptx0
Filter Inserter
Filter Inserter
Posts: 988
Joined: Wed Jan 01, 2020 7:16 pm
Contact:

Re: How is "SPM" measured?

Post by ptx0 »

Trific wrote:
Fri Apr 02, 2021 11:14 pm
I don't have time to read every comment. If I didn't quote you or address you, then I'm not responding to something you said.
what a fantastic way to chat on a forum. just come in halfway down the thread and start talking over everyone else. 8-)
My Mods - Fish Per Minute base size metric
• • •
Base: Bob's @ 1 Million SPM
• • •
Linear search and overflows are indicative of sloppy coding practices.

User avatar
eradicator
Smart Inserter
Smart Inserter
Posts: 4863
Joined: Tue Jul 12, 2016 9:03 am
Contact:

Re: How is "SPM" measured?

Post by eradicator »

pichutarius wrote:
Fri Apr 02, 2021 12:12 pm
it seems like you dont like same block copy pasted and claim high SPM, you feel like that doesnt count. thats fair
While I agree with @Trific that I don't find blocks very interesting I don't think the root cause of that uninterestingness is nessecarily the "blockiness". Every large factory contains some sort of repetition. But I've never seen a copy-/paste block based factory that uses naturally generated ore patches. And thus they don't have to deal with the random distribution/location of ore. It would be interesting to see how much blockiness is possible at all with natural ore without resorting to full-sized city-blocks. For example I can imagine a medium-sized (say 500SPM) block being spread over large parts of the map whereever a fitting constellation of ore-patches is near enough - but even that would likely completely break any alignment to other blocks.
T-A-R wrote:
Fri Apr 02, 2021 12:05 pm
I would like a standardized catagorisation
I think for me the main "point of concern" is the ore-patch layout. And the design choices enforced by that layout. But there's other concerns like: "Do you have to start the factory from scratch and/or research every technology?". Because if tech-cheating is allowed one can simple start at 50 mining productivity which would significantly reduce the amount of miners required. In minecraft "survival" and "creative" made a good distinction between the two I think.

(A) Survival: Choose whatever map-settings the GUI allows. No cheats, no spawning of ore.
Possibly allowed a limited amount of cheats to speed things up a bit? 1) Unlock all non-infinite tech. 2) Instant ghost revival (still needs to cost items though). 3) Change game speed.

(B) Creative: Anything goes. Build to your hearts content.
Author of: Belt Planner, Hand Crank Generator, Screenshot Maker, /sudo and more.
Mod support languages: 日本語, Deutsch, English
My code in the post above is dedicated to the public domain under CC0.

User avatar
T-A-R
Fast Inserter
Fast Inserter
Posts: 156
Joined: Tue May 22, 2018 4:20 pm
Contact:

Re: How is "SPM" measured?

Post by T-A-R »

The catagorisation for all different options in editor usage isn't practical i guess. The best we can do is describing the game settings and used editor functions if you want inform your audience in a complete matter when presenting a build, if it's relevant to the content (is the builder showing a state or a process).

SPM says something about the destination (a state), editor- and map setting tell something about the journey (a process) to reach that destination.

SPM doesn't have to anything with describing the effort of the builder, nor whether a build is "realistic" in a survival run. SPM is imo just the specification about the item flow required to sustain the lab consumption of beakers. It thereby give a idea of the scale of the build.

If one want to show a science pack assembly line the mining productivity doesn't really matter, while it does if one showcases a complete survival base. You simply can't credit a build properly without additional info around the SPM facts. That is up to builder. Else you can only guess what the builder cares for. Without further description viewers have to guess what the challenge was for the presenting builder.

Same with a deathworld build: it doesn't have to do much with spm, but credit could be given for the effort of handling all addition measures in the mean time of building the project.

And Factorio is a game where indeed the journey can be as interesting as the destination. But i am not sure if that matters for the subject for this topic (if i am correct the topic is about measuring a state, not describing a process).

To measure SPM, the production statistics should be "stable" to rule out the effect of buffers:
The throughput requirements need to be met (enough belts/inserters) to reach the stability between input and output.
A build including military science at ratio has higher requirements then a build where military is scaled down.

I can't awnser what the joy is of reaching the UPS limit by editor-copying magic-chest-fed builds. I don't consider them valid bases by my terms, where editor builds should be realistic reproducible in survival run.
I could imagine it compares by checking how much wagons your model train loco can pull, just for the sake of it. (by the lack of a less silly example). I guess its a form of challenging the Factorio engine.

User avatar
eradicator
Smart Inserter
Smart Inserter
Posts: 4863
Joined: Tue Jul 12, 2016 9:03 am
Contact:

Re: How is "SPM" measured?

Post by eradicator »

T-A-R wrote:
Sun Apr 11, 2021 6:24 pm
And Factorio is a game where indeed the journey can be as interesting as the destination. But i am not sure if that matters for the subject for this topic (if i am correct the topic is about measuring a state, not describing a process).
Welp, the topic title is more of a clumsy attempt to capture my fuzzy understanding of why the word "SPM" seems to be so important that it is in every topic about large factories even though - in my view - it doesn't seem to have a commonly agreed on definition or relevance. But along the course of this thread I've come to realize that it at least gives a very rough indication of the size/complexity of the factory in a very short format. And other details/constraints can be explained in the topic itself (also disappointingly they are often not explained).

I've noticed that now there seems to be at least a small trend to write it as "20kSPM (20x1000)". This might already be a sufficent classification for me as it gives me seperate meaures for complexity (1000) and scale (20k). In this example I can disregard the scale as simply being the author also testing how performant a build is on their system.
Author of: Belt Planner, Hand Crank Generator, Screenshot Maker, /sudo and more.
Mod support languages: 日本語, Deutsch, English
My code in the post above is dedicated to the public domain under CC0.

User avatar
T-A-R
Fast Inserter
Fast Inserter
Posts: 156
Joined: Tue May 22, 2018 4:20 pm
Contact:

Re: How is "SPM" measured?

Post by T-A-R »

eradicator wrote:
Mon Apr 05, 2021 5:39 pm
In minecraft "survival" and "creative" made a good distinction between the two I think.
"Deathworld" is a good Factorio description which describes a set of map parameters. I like the distinction between survival, and creative. the A x B SPM seem to be a good way of describing a copy/paste build.
A speedrun base can be expected to be optimized for building. A megabase often can be expected to be built with an eye on performance. But here the terminology is already getting broader/ more vague.
I guess thats what i ment with catagorisation earlier, we may lack some common terminology about describing a format/sets of parameters efficiently. We can describe each parameter exactly, but one could easily forget to add one or more relevant parameters.

User avatar
NotRexButCaesar
Filter Inserter
Filter Inserter
Posts: 926
Joined: Sun Feb 16, 2020 12:47 am
Contact:

Re: How is "SPM" measured?

Post by NotRexButCaesar »

Freeplay and Editor 8-)
Sarcasm and insults are generally neither helpful nor appreciated.
You never know what type of day someone else is having.

Post Reply

Return to “General discussion”