Most difficult ResearchNet level

[vo3435]

Which is it, in your opinion?

[Leylite]

It's pretty hard to choose because there are several different "categories" to choose from, which test different SpaceChem skills. Here's some contenders for puzzles that are the hardest to solve at all:

RESEARCH

Volume 4, Issue 4 - "Getting Pumped" by cake>pie

Probably the most infamous. Requires a ton of re-bonding, space re-usage, and special cases. The sensor helps a bit but the paths really have to be crammed in. Back when the SolutionNet online leaderboards were up, only about 12 people (including the author) beat this one.

Volume 6, Issue 3 - "Catalyst II" by zaq1xsw2tktk

GuavaMoment chose this as a final puzzle for the original 2012 tournament for a good reason - the fusion and fission required to do this puzzle legitimately (without exploiting a bug) are unparalleled, and once again require an incredible amount of space re-usage and special cases.

Volume 3, Issue 10 - "Squaric Acid" by Kanddak

This puzzle seems like it should be relatively straightforward, but it's pretty hard since there's a lot of sub-tasks involved. This ended up being one of the last ResearchNet assignments I solved back when there were only 4 volumes.

PRODUCTION

Volume 8, Issue 13 - "A New Beginning" by Lanky

The math in this one ends up being pretty obnoxious, even if you ignore the random inputs and just use known quantities for everything.

Volume 4, Issue 4 - "Inorganic Pigments" by cake>pie

The outputs here are unbalanced compared to the inputs, so avoiding circular deadlocks or inefficiencies are quite difficult.

Volume 3 Issue 8 - "Nobility" by GuavaMoment

Probably not that hard by today's standards, but it was the first assignment to introduce "sorting by bonds". That paired with tricky nuclear math means getting a good score isn't easy.

HONORABLE MENTION

Volume 8 Issue 3 - "Lies" by Blueeyedrat

Aaargh!

[vo3435]

Wow, I did not expect such a detailed response. Thanks a lot

[Counterpoint]

I still have a ways to go on ResearchNet, but I've beaten all of Volumes 1-3, all but two on 4, and about half the ones after that. That being said, I skipped Getting Pumped, because it just looks like its on the verge of what is possible to solve within a SpaceChem reactor. I think that Squaric Acid is probably the most difficult one that I've beaten...it looks pretty simple when you start it up, but it ended up taking 3 or 4 hours to get a big mess of a working solution. Kinda repeating some of what Leylite said here, but I thought I'd give my two cents anyways.

[GuavaMoment]

I think I was the 6th or 7th person to solve Getting Pumped. It's just a mess no matter what you do.

I need to throw in some support for Vol 1 Issue 8 "Chlorination". It looks easy! It's 2 out of 3 difficulty somehow! The math works pretty nice! Then you realize that you can't just split that chlorine down and the limiting factor is that the hydrogen in the hydrochloric acid has to be a hydrogen from the borane. UGH

Also need to show some support for Vol 6 Issue 2 "Deadly Microwaves". I read a story of someone who thought that microwaving plastic wrap would produce dioxins, and once I realized I could fit dioxin in a large output, the design of the challenge was easy. I spent months trying to see if it was possible to solve at all. It felt so possible since the math works out so nicely, but doing it was so beyond me for a long time. I added the beta input just to make life easier for everyone else, and because I was trying to prove it was solvable it since I needed it for the finals of my tournament. I'm honestly surprised both finalists solved in in the two weeks (I think?) I gave them.

In my opinion, based solely on how much time I spent solving it, the three hardest challenges in increasing difficulty are Getting Pumped, Catalyst II, and Deadly Microwaves.

[vo3435]

Getting Pumped was hard... but not too hard. It took something like five hours.

[vo3435]

Squaric Acid looked dangerous, but the solution I found was neat and almost simple, and it didn't took so much time.

[Berahlen]

I'm pretty proud of myself for finishing Getting Pumped. It was ugly as hell, but I did it.

[vo3435]

"Deadly Microwaves" was a real fun. I think there is a solution without second input, but cannot prove it right now.

[vo3435]

Originally posted by Leylite:

Volume 6, Issue 3 - "Catalyst II" by zaq1xsw2tktk

GuavaMoment chose this as a final puzzle for the original 2012 tournament for a good reason - the fusion and fission required to do this puzzle legitimately (without exploiting a bug) are unparalleled, and once again require an incredible amount of space re-usage and special cases.

Wait, what exploit are you talking about? I don't see a way how bugs could help in that case.

[Leylite]

GuavaMoment covers it in the 2012 tournament final results video:

https://youtu.be/al1F1b5XFZA?t=624

(spoilers on how to solve Catalyst II, of course)

Basically, the "input fusion" bug allows you to input a molecule on top of a single atom, then fuse that atom into the just-inputted molecule on the same cycle, resulting in no collision at the end of the turn (and allowing fusion into atoms that would normally be completely surrounded).

[vo3435]

Wow. I knew about fuse-in trick, but just didn't realize how to apply it in that case. That was nice.

BTW, I've just solve that puzzle without cheating. Amazing experience)