Ghost That Haunts Your Puzzles


I have a new gentler difficulty curve up to level 10. Looking for new blood to try it out. Levels after 10 are not necessarily in order.

Hello friends I am working on a new thing but I tend to make my stuff too difficult so I need your input!

If you would like to help me and have a passing interest in puzzle games please visit this URL on your computer (I don’t think I support mobile yet) to access a demo of my new puzzle game.
pass: thinky_ghost

I’m trying to teach mechanics without explicitly telling you the mechanics so to see if that works I need people to try it. I’m almost certainly trying to teach too much at once, so let me know where you get stuck and I will attempt to smooth this out!

DO NOT FEEL EMBARRASSED IF YOU CANNOT BEAT MY PUZZLES my brain is uncommonly suited for them and honestly it’s more of a reflection on me because I explicitly tried to make them beatable so that’s my bad I’m bad at understanding others.

Also paging @username our resident puzzle expert to tell me if I have something interesting here.





I have to get back to work, but at 3 levels in I think these are pretty good puzzles. I like the density in levels and how well you managed the pacing so I could entertainingly fail after getting halfway through a level.

I usually think about puzzle game difficulty by itemizing, as specifically as possible, each bit of knowledge the game requires in the levels. Difficulty can be analyzed by looking at how many pieces of knowledge the user must gain in between success confirmation. Depending on the audience you’re targeting (I’d say Baba is You is a gold standard for approachable hardcore puzzle game), you may want to add some simpler iterations of your verbs in your earlier puzzles; Baba eventually dropped its guard and later chapters introduced mechanics in very complex forms, knowing that any players who’d reached that far would be able to cope. But Chapters 1 & 2 are much easier even if they have difficult offshoots as bonus levels.

#Level 1

  • Verbs
    • Can push blocks (inherited sokoban knowledge)
    • Blocks to solve blocks when pushed into them (inherited knowledge)
    • Blocks can be pushed off edges (inherited knowledge, mildly novel)
    • Don’t need all blocks (mildly novel, may trip up some people)
      • And an uneccessary block type (purple X), which increases noise
    • Teleporter (inherited knowledge, simple)
    • Blocks can be pushed through teleporters (moderately novel)
      • A simpler implementation would be: player needs to teleport to push a block from behind
    • Block pushed into teleporter to gain edge-advantage (moderately complex)
    • Hidden floors can be revealed by teleporter
      • This is a good gag – it caught me!, but not really a new puzzle mechanic. I think it’s fine and unlikely to prove challenging. I could see removing the yellow-block puzzle if you want to space it out.
  • I see two puzzles in this level - pushing the block through the teleporter, and then the yellow block back in the other way to get it away from its wall. This is pretty dense for an opening level but I prefer puzzle levels with more density in the opening, slow chapters, so it works for me. If it needs to be simplified, I would increase the linearity.

#Level 2

  • Verbs
    • Ghost spawner on arrow switch
    • Old ghosts effectively become blocks; stationary, can be pushed around
    • If spawn position is clogged, new ghost will spawn somewhere else
      • This is a bit of a subtle rule, be careful with this one
    • Player control ‘teleports’ to new location
      • This has conditions I didn’t quite understand. I think it’s: if the player steps on arrow, they will be teleported. If a ‘block’ ghost does, they are teleported instead. This isn’t visually represented well, increasing the puzzle difficulty.
    • Active ghost player is necessary to activate goal
      • Player will need to undo level to prepare by clogging near-exit spawn output
  • This is a good level and it pulls the rug out halfway through, my favorite type of puzzle-game pacing. It doesn’t reuse any of the previous levels’ mechanics, so it should head its own chapter in a more finalized pacing sequence.

#Level 3

  • Verbs
    • Multicolored lock
    • Re-order given blocks (don’t use the heart, use the diamond first)
    • Two touching blocks cancel each other out
    • Cloning through the directional arrow
      • With a trap if the player didn’t brink a block (fun trap, though)
    • Re-using the cloning to create a supply of blocks
      • This exhausted me in Baba is You, honestly – a major step up in sokoban difficulty that demands solving even if the player understood the unique-mechanic puzzle much earlier. My personal preference is that these ‘engines’ be kept spacious and easy to manipulate
  • A substantial increase in sokoban difficulty. This could be made easier by cutting down the cycle counts the block-spawner needs to be run; for example, if the locks were simply 2 hearts and 1 diamond. The multicolored door lock appears to be a red herring?
1 Like

Thanks for the feedback so far – I think I resized level 3 and may have accidentally made it a red herring. I do think level 3 is a little too tight and I busted the puzzle trying to free it up a bit. I will revisit it.

Yeah, I suspected that animations would help a lot for the mechanic on level 2, so I’ll add those soon.

Your outline of my levels looks a lot like my google doc of what I’m teaching w/ each :slight_smile:

I think I undervalue the dopamine hit from solving individual puzzles (even if simple) and overvalue the sort of “puzzle zen moment” of getting tripped up or having a eureka moment, so I could totally spread out these lessons.

ah, that means you’re clearly communicating, then!

For level 2: an animation that shows the copying clearly happening and a red X on blocked spawns would likely clarify things a lot.

Agreed, and made more succinct with your edit. The multicolored door lock was a light introduction to the concept, but you I think you do need to open it to beat the level, unless I misunderstand what you mean by red herring.

I never look in this forum so I’d have never seen this if not alerted, so good call.

I am up to level 102 and I think the puzzles have been rather well done so far (I can’t figure out 102 so I can’t judge it other than saying it is on first blush rather tricky). I don’t have the level of breaking down that Busted provided in me, but I will say that I personally prefer an opening… I won’t say non-puzzle but for me the joy in a puzzle game is more figuring out how to use the rules to solve a seemingly difficult problem rather than trying to figure out what the rules themselves are. This tosses you very much in the deep end and while I think the first puzzle works well enough as a “figure it out yourself” intro puzzle the second one is pretty complicated on its own as an intro to the mechanic.

That said once I learned the mechanics via these first two levels you produced some well thought out puzzles that definitely aren’t pushovers without being like… absolute madness beyond reasonable comprehension (judgement on 102 pending). I think them being relatively small in scope helps them as there are at most times relatively few moves to make, so while hard you only have so many options available to you at a given time.

The one with the teleporters against the walls (4 I think) I wish gave a more initially clear visual indication regarding which way they work, the red line is very thin and while relatively easy to figure out the red line on the… destination one that indicated that side being blocked because its partner was currently up against the wall is something I did not catch immediately when it changed and hence took me a sec to deduce. It is a minor thing.

Okay I went back and finished up 102 after another several minutes, pretty good puzzle.

So yeah, I think the base concept is pretty strong and you’ve designed some solid puzzles with it. I think you may need to teach the cloning thing a bit better and there is some visual stuff I’d make a bit clearer, but overall I’d say this is a strong start.

1 Like

Yeah, I’m playing around with more visual cues for it.

This is definitely less subtle but still not perfect maybe.


EDIT: I think it’s a bit confusing visually for walls vs. switches vs. things you can push. I’ll work on it.

Yeah, I agree with this too. I’ll think about how to better visualize it.

How’d you fair with number 5? Was it obvious what needed to happen once you saw the goal and what you were given, or did it take a bit?

I completed all the puzzles. They are all very good imo.

Puzzles 1-3 are good introductions and I don’t really have anything to say about them that hasn’t been said.

Puzzle 4 was the hardest for me imo. First, it took me a while to understand the conditions that made the a particular side of a teleporter pushable vs. teleportable — though part of that can certainly be attributed to my difficulty distinguishing between red and green (something something accessibility options). Second, the epiphany that teleporters can be teleported through teleporters took me a bit longs to get, but once I figured that out the puzzle wasn’t too hard. This puzzle is good, though perhaps it would be better to introduce those two concepts in separate levels and save this puzzle for just a bit later (though perhaps this would result in this puzzle being too easy (idk)).

Puzzle 5’s epiphany that cloning a teleporter that leads to a pocket dimension clones the pocket dimension as well was probably significantly more of a leap of logic than the stuff in puzzle 4, but the possibility space of the level was small enough that it didn’t take me too long to come up with that idea and test it. I was grinning when I found the solution.

I liked puzzles 101 and 102. They were both novel extensions of the mechanics you had already established, and yet still decently difficult. If you can make [x] more non-introductory levels of this caliber then I’d say that you’d have yourself a solid puzzle game on your hands (where [x] is equal to the desired scope of the game).

1 Like

With puzzle 5 (spoilers I guess) I initially went with the “well if I just push a yellow diamond up there followed by the heart that should duplicate both of them” approach which I was fairly certain was too easy but I found useful to do as I like to physically do the “too obvious” solutions for difficult puzzles just to see exactly how and why they fail. When I saw that was clearly a dead end I spent a minute or so seeing if there was any way the warp blocks could be used to remove the diamond from that last arrow space, which clearly was not possible. However, when trying it just for kicks I did get the warp block duplicated below which was a big “oh hey, that’s useful” moment. I realized almost immediately after that I could just push the diamond and heart in there, duplicate the box twice and have enough of them to reach the goal.

I was crossed up with some of the logistics afterwards for a bit, first I cleared the facing down warp box out of the way to have more room to push those two blocks out of the up facing warp block, but that resulted in the upper block being stuck on the top row and being unreachable (I believe I duplicated it once and thought that might be enough, but when it happened again with the other duplicated warp block I realized it wasn’t quite right). I then messed around with pushing one warp block into the other, which quickly proved to be unviable. After that and pushing the blocks back and forth between the two warp blocks on their original rows for a bit I realized I should just push them into the upper one, then move the lower one out of the way at which point the puzzle was solvable.

So it took me a bit, but while in writing that reads like a lot there wasn’t really a lot of dead time going “what the heck do I do now?” If I didn’t stumble upon duplicating the warp box via trying something obviously wrong perhaps that would have taken me a bit longer to figure out (this is why it is always worth trying obviously wrong stuff). I think it took more time than 3 and less time than 4 if that helps?

FWIW I think having the warp blocks (both the 4 and 5 versions) be both pushable and warpable, but only in certain directions, is a rather clever little twist that can open things up in some interesting directions.

I don’t really like that duplicator animation BTW. I hate to criticize without coming up with an alternative… but I’m struggling to come up with one. I hesitate to disagree with the actual game designer as that likely means I’m wrong, but while it could use some minor… non-giant brown line visual refinement I think having a puzzle before this one in the theoretical full game that more clearly teaches the mechanic would help with the initial murkiness more.

…I ponder if having the new duplicated ghost that player control switches to having like its eyes flash red or something would perhaps be better. I worry about adding any really detailed animation as it would necessitate play slowing down a little bit and delay taking that next move (I just played a puzzle game that did this and it can be irksome), but a quick flash should be almost instantaneous while still being enough to hopefully grab player attention? The actual duplication mechanics wouldn’t be spelled out, but again I don’t know that needs an animation makeover as much as a more direct introduction. Also I don’t make games so I might be an idiot.

1 Like

Hey, I’ve only designed 3 complete “games”, this being the 4th, so I don’t pretend I know much more than you do! I appreciate any and all feedback regardless, and I’ll be able to suss out what I think is right or not. A quick flash or palette swap of the object might be all that’s needed. I’ll try it out and see if I like it. I think you’re right that it’s confusing until you understand it, and then it’s simple, so maybe a better introduction is all that’s needed.

Yeah, animation blocking is a huge pet peeve of mine so currently moving animations get cancelled when you enter a new move before they’re done, and some animations (like falling in pits) always play out even if you’re 3 moves ahead.

Glad you both liked the mechanics in 4 and 5. Coming up with teleporters + cloning is what compelled me to make the game, and I worked backwards from there to get mechanics I thought would work well with those. Obviously, it can get a lot more complicated, but I do think there’s actually a ton of design space to make both less complex and more complex levels.

Every time I make a level I get like 3 other ideas for levels, so my level ideas list is growing much faster than I can implement them.

Again, appreciate all the feedback so far. Between y’all and a friend I gave it to, I think the right decision is to smooth the early curve more so each level isn’t such a jump. I think part of the problem is that I was trying to make a demo that got to the concept in level 5, so I had to establish a lot to get there in a short number of levels. And it’s hard to demo a curve without building out a whole curve.

Unfortunately I can’t get y’all to forget what you learned so it’ll be harder to validate a second time, but I think I have a better feel now for what one looks like.

1 Like

How about:

  • Anim shows the old character ‘deactivating’ (light goes out, soul leaves in a puff of smoke (can a soul leave a ghost? Implying a ghost has a body?)
  • Simultaneously, a similar animation plays in reverse spawning the new object. This doesn’t have to be exactly the same, but it should be visually similar. I’d target 500-700ms, or however long your standard ‘turn’ length is.

I’d do…purple smoke clouds. Can’t go wrong with purple smoke clouds

1 Like

FWIW I thought about this during development and intentionally used big ol’ shapes for the doors/keys but didn’t (and still don’t) have a good idea how to visualize the teleporter stuff. The number of pixels are definitely limiting – I’ll workshop it. I don’t think @username is color blind (please correct me if I’m wrong) but he had similar issues with it too.

I am not colorblind and TBF could see the red line once I knew it was there, it just doesn’t really stick out much which makes it initially not all that noticeable and even when it is known it doesn’t really… pop I guess? The limited pixels definitely makes it trickier. I don’t think it is a huge issue but I would put it on a list of things to try and workshop a bit.

Game kept freezing on me in Safari, as I type this out I realize I might know where the problem is because I forgot I was using Safari

Good to know! Am trying a different platform than usual. I’ll see if that’s a known issue. I can probably produce a mac binary too eventually.

EDIT: What timescale were you getting freezes on? Minutes? I did a spot test and I wasn’t able to reproduce the freezes, but I did notice it’s significantly harder to exit fullscreen in Safari (and likewise hard to access to esc menu while fullscreen in other browsers, not that I’m currently using the esc menu for much).

New animation for copy:

New “blocked TP” graphic:


That copying animation looks a lot better than the prior one, and the warping blocks are much easier to visually parse this way.

Forgot to mention this before, odd request but is there a chance that the undo button can also undo hitting the restart button? I have no idea if that is easy to implement (if it isn’t it is far from a big deal) but I sometimes have a habit of getting those two mixed up in games and appreciate it when I can undo that particular mistake.

1 Like