## This weird level I made

Discuss tricky levels, and show off your diabolical creations
A Random Player
Posts: 523
Joined: Mon Jun 03, 2013 4:54 pm

### This weird level I made

Code: Select all

``[Diagonal:0:maaaaazhtaaaaazztaaaaaztammtaaazaaamtaamaaaatmaasaaaaama]``
I'm not actually sure if it's advantageous to annihilate the diagonal first, as you would need to "manufacture" more oscillated antineutrinos, which would take more moves.. But with it in the way, it takes longer to move the two manufactured antineutrinos.
I got 65 or so by not annihilating the diagonal.

Edit: I got 53 moves by annihilating about 6 of the diagonal, so maybe it's faster.
\$1 = 100¢ = (10¢)^2 = (\$0.10)^2 = \$0.01 = 1¢ [1]
Always check your units or you will have no money!

testtubegames
Posts: 1078
Joined: Mon Nov 19, 2012 7:54 pm

### Re: This weird level I made

Phew, I got it down to 52 moves, but I'm sure we're still far from perfect.

A Random Player
Posts: 523
Joined: Mon Jun 03, 2013 4:54 pm

### Re: This weird level I made

I got it down to 44 moves. I don't remember all of it, but it involved creating a "neutrino transport tube" in the end.
\$1 = 100¢ = (10¢)^2 = (\$0.10)^2 = \$0.01 = 1¢ [1]
Always check your units or you will have no money!

testtubegames
Posts: 1078
Joined: Mon Nov 19, 2012 7:54 pm

### Re: This weird level I made

Man, that's a pretty dastardly level to try and get perfect on. I'm down to 43, though. I'd guess my strategy is similar to your method, using the right-most two columns pretty exclusively.

When we're ready (and I doubt 43 is perfect, still), I can always plug it into the 'solver'. Though I get the feeling a level *this* intense may well crash the solver.

Edit: 42!

A Random Player
Posts: 523
Joined: Mon Jun 03, 2013 4:54 pm

### Re: This weird level I made

I actually never tried using only the right most columns, I tend to go to the middle for some reason. I'll try that.

Edit: I'm getting worse at this, I got 48 XP
\$1 = 100¢ = (10¢)^2 = (\$0.10)^2 = \$0.01 = 1¢ [1]
Always check your units or you will have no money!

19683
Posts: 151
Joined: Wed Jun 05, 2013 12:15 pm

### Re: This weird level I made

testtubegames wrote:Man, that's a pretty dastardly level to try and get perfect on. I'm down to 43, though. I'd guess my strategy is similar to your method, using the right-most two columns pretty exclusively.

When we're ready (and I doubt 43 is perfect, still), I can always plug it into the 'solver'. Though I get the feeling a level *this* intense may well crash the solver.

Edit: 42!
What's the link for the solver?

This is one complex level!
Binomial Theorem: ((a+b)^n)= sum k=0->k=n((n!(a^(n-k))(b^k))/(k!(n-k)!))

testtubegames
Posts: 1078
Joined: Mon Nov 19, 2012 7:54 pm

### Re: This weird level I made

No link, actually. It just exists on my computer -- a program someone made for me that can crunch through (most) levels.

I'm happy to plug in levels as requested. Down the road, if I become a bottleneck, I'd also consider releasing the solver to a wider circle.

19683
Posts: 151
Joined: Wed Jun 05, 2013 12:15 pm

### Re: This weird level I made

A Random Player wrote:

Code: Select all

``[Diagonal:0:maaaaazhtaaaaazztaaaaaztammtaaazaaamtaamaaaatmaasaaaaama]``
I'm not actually sure if it's advantageous to annihilate the diagonal first, as you would need to "manufacture" more oscillated antineutrinos, which would take more moves.. But with it in the way, it takes longer to move the two manufactured antineutrinos.
I got 65 or so by not annihilating the diagonal.

Edit: I got 53 moves by annihilating about 6 of the diagonal, so maybe it's faster.
Could you plug in this level?

Binomial Theorem: ((a+b)^n)= sum k=0->k=n((n!(a^(n-k))(b^k))/(k!(n-k)!))

testtubegames
Posts: 1078
Joined: Mon Nov 19, 2012 7:54 pm

### Re: This weird level I made

Yikes, just as I feared, this one broke the solver!

As a background, the solver works by, step by step, compiling all the possible states of the game in one (growing) list. So, at step 0 it has one state (the starting state). At step 1 it has 3 states (the starting state and the two possible moves you could make). At step 2 it has more... and so on. Normally it works fine, but memory is the upper limit. The list can only grow so long before my computer starts to hang.

This made it to Step 22... where, with 24 million states in the queue, it froze. Moral of the story: this is a tough level.

Now I suppose I *could* replace a bunch of the irrelevant antineutrinos with crates, so as to cut down on the possible moves... but even then, I can't imagine the solver could handle the level. The number of states tends to grow pretty exponentially... and we're only halfway to our expected win-state (around 40, perhaps?). Looks like we're going to have to do this one by hand...

A Random Player
Posts: 523
Joined: Mon Jun 03, 2013 4:54 pm

### Re: This weird level I made

testtubegames wrote:Yikes, just as I feared, this one broke the solver!

As a background, the solver works by, step by step, compiling all the possible states of the game in one (growing) list. So, at step 0 it has one state (the starting state). At step 1 it has 3 states (the starting state and the two possible moves you could make). At step 2 it has more... and so on. Normally it works fine, but memory is the upper limit. The list can only grow so long before my computer starts to hang.

This made it to Step 22... where, with 24 million states in the queue, it froze. Moral of the story: this is a tough level.

Now I suppose I *could* replace a bunch of the irrelevant antineutrinos with crates, so as to cut down on the possible moves... but even then, I can't imagine the solver could handle the level. The number of states tends to grow pretty exponentially... and we're only halfway to our expected win-state (around 40, perhaps?). Looks like we're going to have to do this one by hand...
Have you tried deleting unused gamestates, and checking for duplicate states? Might help a bit.
\$1 = 100¢ = (10¢)^2 = (\$0.10)^2 = \$0.01 = 1¢ [1]
Always check your units or you will have no money!