The red circuits are buffered and used to craft the roboports and provider chests. There are also the two prod module 1 assemblers that use some of the red circuits.

For the any% duo botless plan, we iterated on size of builds, timing of builds, build arrangements, handcrafting queue, resource distribution between players and more. Usually, it's about finding the bottleneck/problem and fixing it. But I want to give a concrete example, using the saves linked here so that it's easier to follow: first complete run with the plan, latest complete run with the plan.

In the first run, you can see that we have some extra mini-smelting used to provide the pipes and gears for engines for chemical science. This smelting is removed in the latest run and the engines are fed from the mall iron. This was done because we found that the mall was overproducing and the mini-smelting was complicated to build and messed up the arrangment of the whole lower drill + smelting area. So, the iron and gears for the engines were moved up into the mall and build to buffer much earlier than in the lower area, so that we could still use them for handcrafting. This change had a rather big impact because it allowed us to rearrange the entire lower half of the base to be much easier to build (alignment, amount of complicated spaghetti) and reduce some of the hand-feeding for electric engines.

Another bottleneck in the first runs was that the rocket control unit build needed a ton of time to be built while we already had the ingredients for the RCUs, meaning it was late. So we switched to handfed RCUs with its lower setup time but requirement to be babied after setup. Practice had shown that I had the time to do the handfeeding, but the initial setup was too late, so this solved that bottleneck. Similar thing with switching the northern green circuit smelting + steel expansion to be handfed coal instead of belted coal - faster setup for a build that was always late with the tradeoff of some extra time spent later when it was available.

Often, these bottlenecks or problems only matter once build times are consistent(ly fast) and mistakes are reduced to a minimum. Before that, the mistakes/placement time variations are the bottlenecks and small build adjustments not worth worrying about. I think this is very well illustrated by the any% runs that are based on Nefrums guide (like mine): I started off with a 3:47 time, heavily limited by my knowledge of the plan (=mistakes and bad placement speed). So the fact that the plan is not optimal per se does not matter yet. My latest run has a 2:43 time, which is partially limited by blue science being low in the end. So the non-optimal plan partially limited that run.

It is not clear to me how the techniques that I described are not speedrun design techniques. They are techniques I've used to plan speedruns, so they are speedrun design techniques. I do have some more info on techniques for planning rough guidelines/general structure of speedruns though:

You can find the gist of how the main MP team does speedrun planning here: https://factorio.com/blog/post/fff-344. From what I can see from a quick glance, Nefrums follows a similar "target time, subtarget times, assembler calc, initial plan, (run, iterate)" technique in this (older) speedrun guide, Nefrums' target time calculations are shown in the spreadsheet linked in the video description. The "any% cheatsheet" by macros that is linked in the resources section shows similar target time calculations for science packs. This technique is also what we used for the botless any% duo plan, modified to have much earlier and more frequent iterations - we already started running and iterating the plan when red&green science builds were set up instead of waiting until we got to the rocket with the planning save layout. I think this is a very individual modification to make, but it worked well with our personal planning strengths.

I think that beyond that, it's mostly experience. (Factorio) speedrunning is a lot like learning a language: You can learn grammar theory and vocabulary all you want, to get a feeling for the language and get really good you have to practice it. So, speedrunning is a lot of practice and iteration of plans. As an example for how much a group of people can iterate a plan, the any% mp run linked in the FFF above was in April. We practiced that run and iterated the plan over and over to get sub 1h in in July.

To sum up, in my personal opinion, the theoretical time estimation techniques presented by you will never get better results than human experts trying and iterating a plan in practice. Theory-focussed time estimation when refined may be useful to improve the initial plan that is iterated on, however I am already satisfied with the quality of my initial plans.

It is also unclear to me why you say you "will give me a pass" on potentially lacking run submissions, considering that you haven't submitted a single speedrun. The outside perspective is not a good position from which to judge speedrun activity. But if you must, you can find my run submissions on my profile as usual.

I don't think it is useful to tear apart wording of individual sentences. Once you have read my post completely and followed the links, perhaps compare my description of tools and listing of alternatives to your own post listing only one tool for the job and giving no advantages/disadvantages of the tools over others, like I suggest in my first paragraph.

How people plan and what tools they use is as individual and subjective as everything in Factorio. I find it very disrespectful to other techniques and even misleading that you present these tools as "the (only) one/the best" solution, including advertising your own. I'd appreciate it if you could at least give the advantages of the tools you list over other tools. I'll do so for what I use below, while giving even more tools that are available.

I personally vastly prefer the map editor and the built-in infinite/creative entities over creative mod due to its better intergration into the vanilla game and improved performance over anything mods can do. Another planning option is the Editor Extensions mod or the vanilla sandbox scenario, and various other creative-mode style mods.

More to the map editor advantages over "creative mod": Tick-perfect time forwarding also via keybinds, instant research and unreasearching techs all within the tech tree, expanded zoom without console commands.

Another option for incremental builds is using the surfaces. The map editor provides keyboard shortcuts to allow to easily switch between them, and you don't have to tax your perfomance by zooming out beyond normal/map editor limits.

Regarding saves for segmented practice, something that is a bit hidden is that you can save in replays - so you can take your PB run, replay it to where you want to start to practice, make a save and play that save from there. For making sure you always have replays on: ctrl + alt + click on "settings" -> "the rest" -> ctrl + F "replay" -> check the "check-enable-replay-checkbox" option (1.1+ only), this will check the enable replay checkbox by default. For mods that aid segmented practice, give this search a go: https://mods.factorio.com/query/speedrun?version=any

For tech calculations, an easy way is to use one of the spreadsheets for techs that is flying around, e.g. the any% cheatsheet linked in https://www.speedrun.com/Factorio/resources. The linked resources also link to other resources, e.g. the "cheat sheet" links to multiple other calculators, including science/lab calculators. There is no need to reinvent the wheel when others already provide very accessible tools :)

For blueprint resource/item calculations I use BlueprintBot (should be familiar from Discord, message it !help to see all the useful commands beyond blueprint rendering) because I can rely on it always being up to date. It breaks down items to the last handcraftable component (e.g. steel or engines) instead of completely down to ore, this can be seen as both an advantage and a disadvantage. It also has some general counting of buildings function which can be useful to compare different setups, like different smelting layouts, by item count.

Some people may no longer consider kirkmcdonald the gold standard calculator (partially due to still being on 1.0 right now), so I want to link https://factoriolab.github.io/ as a possible replacement. Like I said above, more links to caculators can be found in the cheat sheet linked in the resources section. Choice of calculator usually depends on invidiual requirements and uses, so I really recommend to also take a look at the other calculators linked in the cheat sheet and also mods like max rate calculator, factory planner, helmod, and more.

@blazespinnaker, please edit your posts/post all your thoughts at once instead of making double posts. Every new posts generates a notification (if that is turned on, which it is for me), so you are often generating two notifications for posts that follow each other. I'd rather not have to turn off the notifications for this entire forum just because they get clogged by your double posts. Thank you.

Picture of the recent notifications with your posts highlighted, to make it clear what I mean:

Allowing all game options would also allow custom map strings that go beyond the in-game map gen settings and all debug options. Both of these things are legitimate parts of the game, but the current any% category does not allow them. What is your opinion on allowing them in any%?

You can play a multiplayer game with only one player. This is what most runs do. Usually done through multiplayer -> host new game and then simply playing in the same way as a singleplayer game.