Opus Magnum Transmute-All

MightyMooquack, I was hanging in there with you until it came to optimizing for cycles on Airship Fuel. After spending more time than I care to admit, the best I can manage is 27c.

I looked up a 20c solution and... yeah, I don't think I would have come up with it. However, ...

Spoiler:

I did surmise that all three inputs had to be fully utilized, with two elements getting "buffered". I just couldn't work it out.

Ted wrote:

MightyMooquack, I was hanging in there with you until it came to optimizing for cycles on Airship Fuel. After spending more time than I care to admit, the best I can manage is 27c.

I looked up a 20c solution and... yeah, I don't think I would have come up with it. However, ...

Spoiler:

I did surmise that all three inputs had to be fully utilized, with two elements getting "buffered". I just couldn't work it out.

On Airship Fuel and others like it:

Spoiler:

This puzzle is an example of one in which the atoms required for its output are not evenly divisible by its inputs. Rocket Propellant and Explosive Phial are others, and there are plenty of others still.

The naive way of solving any puzzle is for the overall loop of the solution to work like this:

  1. Pull inputs off their sources.
  2. Assemble them.
  3. Output the result.

But for this class of puzzle, this is not optimal. The bottleneck for nearly all puzzles is how quickly you can pull reagents off their sources. To use Airship Fuel as a specific example, with its three input atoms to four atoms required in the output, this means you would spend the first two cycles grabbing the first three atoms, and then the next two cycles grabbing only one additional atom. The other two reagents would go to waste for that slice of time, which means any such solution will not be as fast as it could be.

The LCM of 3 (our number of reagents) and 4 (the number of atoms in the output) is 12. Thus, we will need to handle 12 atoms before our machine can loop, which translates to three distinct outputs, across four rounds of retrieving inputs (or eight cycles).

Actually designing a solution for this is fairly complicated, but here is how I got started with it. First, let's identify all 12 of these atoms individually, and see if we can find a reasonable design for where and when each one will come from, and where it will go.

Let's call the three output items A, B, and C. Let's also number each atom in the output from 1 to 4. This allows us to identify each of our 12 atoms by combining the two: A1, A2, and so on up to C4.

Let's also identify our three inputs as left, middle, and right (L, M, R).

Finally, let's number the cycles on which we will be able to retrieve inputs, from 1 to 4. Putting this all together, we can assemble a table like the following:

_ L M R 1 A1 A2 A4 2 B1 A3 B4 3 B2 B3 C4 4 C1 C2 C3

There are a lot of ways of arranging this table. This is an arrangement very similar to what I used in my solution.

This tells us, on each cycle, where each atom will come from, and where it needs to go. The challenge of the solution is actually implementing this table, which I leave as an exercise to the reader.

This may be the most interesting puzzle in this class. Rocket Propellant (for example), with its 2 inputs and 5 outputs, is considerably simpler. Its table might look like this:

_ L R 1 A1 A2 2 A3 A4 3 A5 B5 4 B1 B2 5 B3 B4

The loop only involves two outputs, and a single round of inputs that gets split between the two of them.

Keep an eye on the exact numbers of inputs and outputs you have. If they don't divide evenly, you're in for a complicated time.

Thanks the detailed explanation. After sleeping on it I realized my whole problem was keeping myself stuck in the mental box reinforced by my approach to previous puzzles.

I'll give it another go in a day or so with this approach:

Spoiler:

create two distinct fabrication streams: the first using two inputs to create the product and the other using only one input to create an entire product, contributing when it can.

Into part IV now, some of my solutions are starting to get messy!

Spoilers for Stain Remover:

Spoiler:

The area score for this wasn't actually that bad, but it took a large amount of cycles.
IMAGE(https://imgur.com/O7yXahP.gif)

I'm trying so hard not to just brute force stain remover^^

Yesterday's patch made two score-altering changes, and they will be resetting the leaderboards later this week as a result:

Today's update contains two fixes that will essentially invalidate the current scores in the leaderboards and histograms:

1. The cost of the piston arm has been increased from 30G to 40G.

2. Solutions that loop fewer than two full times before producing all outputs will now trigger an additional "verification check" to verify your solution can loop without breaking. Cycle and area scores should not be affected.

If you see anything that looks like a bug in the scoring system, please email us ([email protected]) so that we can investigate it right away. I'm hoping to reset all histograms and leaderboards in the next few days, and hoping to not have to do it again after that.

ALSO: I recognize the irony in how we just implemented a fast-forward feature that is completely unusable to the high-instruction-count optimizers who want it most. We are working a solution for this and hope to have it out soon!

Point 1 is a significant change to low-cost solutions, and I'm going to need to reevaluate my "cheap" solutions for most of the puzzles.

The second point is more subtle. With certain solutions, it was possible to eliminate a cycle or so by behaving differently on the sixth output, since you knew you didn't have to set things up again for the next go-around. I think I only did this once, but now I don't remember on which puzzle.

Anyway, based on previous Zachtronics games, when the leaderboards reset, you'll probably need to go through and re-execute your solutions in order to post them on the leaderboards again.

I'll have to revisit a puzzle or two thanks to Point #2 because I got a little sloppy on some of the cycle optimizations.

Important note: They reverted that second point in another patch almost immediately:

Tonight's update removes the restrictions on non-looping solutions that were added in the past two days. It's possible that we'll try something else in this area, but for now it's back to what it used to be!

Ah, good to know. I think I'll still make sure it runs smoothly; I'm up for something light tonight.

Really enjoying this game. The Mahjong variant mini game is much better than the Freecell variant in Shenzhen.

The only thing I think its missing is an inspired meta game. Though the game is fundamentally a puzzle game, its also about crafting. I want to run an alchemy workshop with little black boxes on my shelves ticking away at my solutions to various problems to fulfill orders that come in. Maybe an elemental shortage forces me to rethink the way I solve manufacturing rocket fuel, or I just want to crank out rope for a few hours to help build a new airship while I noodle away at some rounds of Sigmar's Garden.

Got over my stain removed hump and did 3 more solutions last night (well, 2 and the teeniest bit of fixing to do on the last).
So good.

I'm quite proud of the first pass of my Sword Alloy solution!

Spoiler:

IMAGE(https://imgur.com/rMCswNQ.gif)

Which with some slight modification, becomes my very compact Purified Gold machine.

Spoiler:

IMAGE(https://imgur.com/z1SRzMU.gif)

The latest update adds a fast-forward / breakpoint function by way of ctrl+clicking an instruction during playback. Very nice!

Holy cow clockworkhouse, that 17 cycles on the first puzzle is nuts! Nice job.

Can people shoot me a friend request on Steam? I'm robc04

I have to say - Ted and MightyMooquack are pretty awesome at this game. ClockworkHouse has also had some really good scores. I can tie some of your scores, but others just leave me totally perplexed - such as 14 for area on Health Tonic. Absolutely amazing!

I've got quite a few friends that play this, so I don't get to see all of my friends' scores since the leaderboard is so small - so the rest of you may be right up there on a lot of the puzzles too. Is there any way to view all friends' scores that I'm missing?

robc wrote:

I have to say - Ted and MightyMooquack are pretty awesome at this game. ClockworkHouse has also had some really good scores. I can tie some of your scores, but others just leave me totally perplexed - such as 14 for area on Health Tonic. Absolutely amazing!

I've got quite a few friends that play this, so I don't get to see all of my friends' scores since the leaderboard is so small - so the rest of you may be right up there on a lot of the puzzles too. Is there any way to view all friends' scores that I'm missing?

Ah you flatter me!

To be honest most of my labors took a lot longer than I care to admit. I actually started burning myself out and not having much fun. I wised up and switched gears halfway in Chapter 2 by only optimizing for area. By the end of Chapter 3 I went with anything that worked and then optimized retroactively when the fancy struck. I started having way more fun, but that's just me.

In regards to optimizing for area in I-Health Tonic there's a trick to it, as you've probably suspected. This trick plays a key role in optimizing for area in most subsequent challenges. Here's a vague hint:

Spoiler:

This challenge has something new about it as compared to all the other puzzles before. Can you spot it?

I don't think there's a way to see all friends' scores, unfortunately.

I'm not going to read that spoiler.

Robear wrote:

I'm not going to read that spoiler. :-)

For now I'm not because I like figuring out stuff myself, but if I remember maybe I'll look at it once I get through the game on my own. Ted did say it was vague, so maybe it won't give away too much.

I will ask Ted if the following idea is it though...

Spoiler:

Does it involve rotating the initial input instead of moving it off of the dispenser.

robc wrote:
Robear wrote:

I'm not going to read that spoiler. :-)

For now I'm not because I like figuring out stuff myself, but if I remember maybe I'll look at it once I get through the game on my own. Ted did say it was vague, so maybe it won't give away too much.

I will ask Ted if the following idea is it though...

Spoiler:

Does it involve rotating the initial input instead of moving it off of the dispenser.

Got it in one!

Just picked this up... going to add some of you to my steam friends.. I have a feeling I'm too much of a brute force "just make it work" kind of guy but I love how this game makes me think about problem solving.

Ted wrote:
robc wrote:
Robear wrote:

I'm not going to read that spoiler. :-)

For now I'm not because I like figuring out stuff myself, but if I remember maybe I'll look at it once I get through the game on my own. Ted did say it was vague, so maybe it won't give away too much.

I will ask Ted if the following idea is it though...

Spoiler:

Does it involve rotating the initial input instead of moving it off of the dispenser.

Got it in one!

Yea!

Carlbear95 wrote:

Just picked this up... going to add some of you to my steam friends.. I have a feeling I'm too much of a brute force "just make it work" kind of guy but I love how this game makes me think about problem solving.

Many of my solutions are probably not elegant - especially when going for cycles.

As I said above, Carl, the game has 3 major axes you can design on. You can change the challenge just by optimizing. What's *really* interesting is figuring out why you design things the way you do; what your particular blend of priorities naturally is.

It might as well be called "Rohrschach".

I beat one of clocky's cycle scores! Pretty sure that means I beat the game.

Chaz wrote:

I beat one of clocky's cycle scores! Pretty sure that means I beat the game.

You beat life itself!

I picked this up and I sent out a few friend requests. So far the presentation has done a great job of getting me into the game. So far I seem to be prioritizing cost.

"You arrange the party in the streets, I'll arrange the party in the sheets" is both hilarious, and not a line I expected from this game.

Sorry for skimming. I'm digging the game, but does anyone know why only one of my friends scores is showing? I can only compare my own scores to those of Robc04, but plenty of my friends have played the game (not least of all Robear, who gifted it to me). I notice there is a hermit mode but I don't have it checked. Any ideas? Do you all have hermit mode checked?

Not that I'm aware of... Pretty sure that the leaderboards have been reset a few times by patches. Maybe none of us has played since the last one?

I updated one of my solutions - Rocket Fuel - so maybe it will include me now?

kergguz wrote:

Sorry for skimming. I'm digging the game, but does anyone know why only one of my friends scores is showing? I can only compare my own scores to those of Robc04, but plenty of my friends have played the game (not least of all Robear, who gifted it to me). I notice there is a hermit mode but I don't have it checked. Any ideas? Do you all have hermit mode checked?

I've noticed this as well. I'm guessing it's the leaderboard reset that Robear mentioned. I sent you a friend request so you can feel good about beating my scores :).