Puzzle Pavilion

Spacechem revisited

Spacechem Revisited is a follow-up to the successful… oh, wait. My apologies. I’m revisiting Spacechem as when I wrote about it last in ye olde Pagoda I had only played through the first two or so worlds out of nine. That was good for letting people know that it existed, less good for being able to give anything beyond a general intro to the game. Let’s try and fix that.

Despite the name Spacechem is less about chemistry and more about programming. It is your duty to craft programming loops (or more specifically, two complimentary loops) that takes one or more inputs and transforms them into one or more specific outputs. Your initial tools for doing so are limited to grabbing and dropping atoms/molecules, breaking or creating bonds and rotating them (also the ever valuable sync tool which allows you to effectively pause one loop while the other completes a given process). As you proceed you gain the ability to fuse different atoms together into a single new one, add sensors that act like a railroad switch or flip-flops that can be used to alternate which path your program runs along. That final one becomes quite the doozy. In addition the game quickly proceeds to advance from single screen puzzles to ones where you are managing a series of plants each running their own interconnected processes; this complicates things in both ways one would anticipate and ways they would not (the latter would include the piping linking these various plants together).

Fortunately you need virtually no chemistry knowledge to play the game. You basically need to know how many bonds an atom can form and the atomic numbers when fusing various atoms together, both of which are given to you. What you do need is the ability to refine your designs as I am convinced that no matter how gifted one is at coming up with an initial concept in the majority of cases there will be flaws that prevent it from functioning as envisioned. Perhaps it doesn’t quite fit, perhaps it hits a wall or another molecule, perhaps you mentally missed a step, perhaps it will work for one cycle but not the ten-plus that are required. Perhaps when dealing with multiple plants the speeds of them are different enough to throw things into chaos. All I know is that there will be a “perhaps” lurking somewhere.

Now refinement can take a few different forms. The way most go is to strive for increased efficiency, and in doing this most of the problems will be solved. The game in fact encourages revisiting previously solved puzzles and refining them in such a way as each time you complete a puzzle your solution is graphed alongside everyone elses, an attempt to inspire or shame you to do better. The other form, the one that I embraced, is to say to hell with that and get the original notion working even if it results in madness such as this.

(I know, I’ve posted that screencap before. It’s still the scariest one I have!)

Don’t you pretend like that image makes any sense, I made it and I can’t even follow it. The strength of the game is that it gives you the freedom to take either route or any in between. It is an absurdly challenging game but while many of its ilk would tease you with the infinite possibilities of a blank screen before leaving only a few narrow paths to success once one reaches the latter stages, Spacechem will generally let your insane, absurd, inefficient monstrosities still pump out a passable solution up to and including the final stage as long as you can manage to make it hold together just long enough. That is what elevates it to that higher level of puzzle game and makes it an experience at least worth investigating. And hey, it comes with a very substantial demo that has more than enough content included to let one know if it is something they might be interested in.


Hanano Puzzle 2

The very first puzzle game I wrote about here was the original Hanano Puzzle. I consider it and its cousin Jelly no Puzzle to be the peak of this school of puzzle design. That lead to both a sense of excitement with the surprise release of this game and the weight of expectations, expectations that it couldn’t quite reach.

There is a tiny devil in the details on the game’s itch.io page, right there near the bottom: Programming, art, and music by Tatsunami. Level design by Lucas Watson. Tatsunami is who I’ve always referred to as qrostar, the singular person behind the original game. The other person is a fan who sent in some level packs that impressed, leading the creator to work with them a bit to polish things up and release it as an officially sanctioned thing.

This is a very nice thing to have done. The problem is that while this Lucas fellow seems to be good at puzzle design Tatsunami is a borderline genius. As such Hanano Puzzle 2 is a very nice puzzle game that has the unfortunate fate of being a twin to something truly special. This is a shame as on its own this is good, a solid 7 out of 10 with its last few puzzles in particular being very well done. I’d recommend it.

Still my main thought is how it, by existing, helps to illuminate what was remarkable about the original. There is fifteen fewer puzzles here and yet it feels like certain ideas are repeated more frequently. The sprouting flowers often created a dynamic element in the original that is almost absent here. It does probably introduce ideas a bit smoother than the original at least, that being a notable stumbling block for many.

Again I feel bad spending all this time being critical of a swell enough game, one that is basically a free gift. Most of the puzzles are good, it is stronger than a number of puzzlers I’ve written about previously here. It just has the burden of that name.


Oh man did someone mention Slitherlink a year ago? I’ve played so much of that on the DS that I can play even the hardest ones using either lines-only or X’s-only. I’ve tried both!

1 Like

As a heads up, all three of the Squeenix GO puzzle games (Hitman GO, Lara Croft GO and Deus Ex GO) are currently available for free from Amazon Underground if you have an android phone or a Kindle Fire that can run them. I’ve played through much of the Hitman entry and it is a pretty solid little puzzle game with an amazingly charming board game aesthetic. Might do a full write-up on it when I finish it up.

I think they have certain permissions required so that you get a pop-up ad when starting up, if stuff like this concerns you be forewarned.

1 Like

They’re also all 99 cents on the Google Play Store for Android.

I just finished wrapping up Lara Croft Go and it’s great. Excited for the Hitman and DX versions.

I just started (well, restarted) Hitman Go but I can’t play too many missions in a row or it starts to feel reeeeaaal repetitive.

I think one of my main criticisms against Hitman GO is that it waits too long to unveil some of its more interesting mechanical twists. It seems to like to introduce a new playing piece once or twice per level at about the 1/4 and 3/4 mark, and while it makes sense I think it would have been aided by front loading at least a few of these towards the first third of the game. As an example, what might have been the last mechanical twist added in may be the most interesting one in the game and I’d have liked to have seen it used more than it was.

hello, some friends are trying to get a cute puzzle game on steam! please help them out if you feel so inclined




All the Zachtronics games are on sale in a bundle.

Hitman GO

I finally got around to finishing up all the objectives in this game, so I should probably write something about it.

The game is basically an odd puzzle/stealth hybrid in the skin of a board game. Each stage has its own layout with an eventual goal you must make your way to and various enemy/guard pieces in your way. You move first, then the guards move, repeating as necessary. This helps enforce the stealth element as you can’t directly go after most enemies; if you step into their line of sight they have the next move and will strike you down. You can still take them down from the side or behind (or if you can arrange things so that they are the ones to step next to you) or you can try to just avoid them altogether (some stages require you to off at least a few enemies).

The game is laid out as a series of board games, each one containing 8 or 15 stages built around a given set of target assassinations. Once past the first few intro stages each one has three objectives, the mandatory “reach the goal” one and two optional ones. Most of these optional ones are based around completing the stage in a given number of moves or killing/not killing every enemy with a few other more exotic ones sprinkled in here or there. They often make completing the puzzles more interesting and can produce some unique challenges.

The game generally throws in one or two new mechanical twists in each group of stages. These are usually a new enemy type with its own unique behavior. You get guard dogs that can see you from an extra space away and will chase you around the board, two enemies that stand back to back to prevent you from sneaking up on them from behind and snipers who can shoot you if you step along any row of spaces among others. There are also some items that pop up from time to time that help you reach your goal or otherwise mix things up, be them rifles that let you fire at a certain set of spaces or moving walkways which, I mean I don’t think I’ve ever objected to the inclusion of a moving walkway.

Overall the game is good but it takes a good while to ramp up, especially if you are only trying to clear a stage and are unconcerned with any optional objectives. This makes sense in a way as this was likely aimed at an audience not familiar with this kind of puzzle game. If you are more familiar with them you will likely speed through the first few boards before starting to meet much resistance, and even then the game rarely becomes that difficult. There is also the issue that some of them most interesting items are the very last ones given to the player.

The very last item introduced in the second half of the last true board is a stopwatch square. When you step on it you skip a turn while every enemy takes an extra turn, which sounds odd but is actually a big, game-changing thing. Enemies generally go through a cycle of moves, and due to the way the alternation of your and their movements work (plus having to avoid stepping into their lines of sight) you are basically set up to only interact with them on every “even” step. The stopwatch item breaks your synchronization with them and lines you up with their “odd” steps instead. Given that each stopwatch can only be used once and that you must interact with some things on an even step and some on an odd one in order to successfully complete the puzzle… it is basically the kind of mechanic that someone could build an entire puzzle game around. Hitman GO goes 75 or so levels before introducing it and ends shortly after. They left a lot of potential on the table.

That said, it is a pretty satisfying puzzle game to dip into for a few stages at a time. The basic stealth gimmick is a neat one and once you get a ways into the game the puzzles offer just enough resistance to make things interesting. I would also be remiss if I didn’t point out how utterly charming the aesthetic is. Everything is crafted to look like pieces on a(n unrealistically detailed) gameboard and I just adore it. The game itself isn’t something one will likely love, but it offers enough enjoyment to be worth checking out (it is still free or 99 cents on some android storefronts) and hopefully later Eidos GO games will build on the strong base this one laid out.

If you liked Hitman, I highly recommend both Lara Croft and Deus Ex Go, both of which look better AND play better. Lara Croft is the best, I think - it abandons the flat board motif and goes vertical, better to fit the character’s game-history of climbing stuff and plumbing tombs.

All three games are essentially similar (e.g. snakes in Lara = blue dudes in Hitman, lizards in Lara = dogs in Hitman), but it’s cute the way they interpret the original games’ mechanics as little powers or enemies unique to each game. Lara has more pressure plates and traps and whatnot. Deus Ex has a stealth power that makes you invisible for one step. Etc.

I’m glad to hear that. Unfortunately I seemingly cannot run those later entries on my Kindle Fire HD (which is what I play most of these mobile games on) so I’m sorta stuck until I upgrade something. Either that or until I get bored and try the pc/console versions, but they seem like a better fit on mobile.

I’ve also been playing through Pushmo. I won’t do a big write-up on it as Warpzone did a really great one back in the previous topic here. I will say that while I like it I don’t love it. I don’t think I can name a puzzle game that needed an editor more than this one, it seems to have a bit less than 250 puzzles and probably should have just picked the best 60-70 and made that the game instead.

More importantly, while I respect the concept of dealing with 3d space it never truly got its hooks in me. As a rule I generally don’t play puzzle games after a certain point at night or else it plays out in my head while I’m trying to sleep and I then get less of it. I could (and have) played Pushmo right before sleep and had no problems. I think it leans more in the trial and error direction rather than a “figuring out a linchpin” one and that is why. Also not a fan of a large number of the puzzles being built around discovering if you had to preserve a pathway earlier on and forcing a restart if you didn’t figure it out fast enough.

Other than that it is legitimately pretty good! It just has things holding it back from being something much more remarkable.

I have the same rule except for programming!

So hey, CrossCells, the new game from the guy behind the Hexcells trilogy and SquareCells, came out earlier today and I’ll likely write something about it when I finish it up.


I have to note one thing, and that is that the mechanic introduced in stages 21-25 is just poorly explained in that a large portion of people who are playing the game are having to struggle to figure out how the heck it actually works. I had to go to the Steam forum just to see what the rules actually were.

However! There is a much simpler answer, and apparently after a patch it will end up this way regardless. After you play stage 21, skip ahead and do puzzle 26 next, then return to puzzle 22. 26 teaches you an aspect of a mechanic that is required to solve puzzles 23-25.

Other than that headache the game is good so far.

EDIT: So the dev got enough feedback to patch in a new puzzle that does a much better job of introducing this concept, so problem solved and one can just play as normal.

The comparison is broad but not quite so much as to creep into reductionist territory

1 Like


I hope more people are playing this after the $10 Humble Bundle sale.

This is an unbelievable game that everyone on Select Button needs to play. I held off on getting it at $30 since that seemed too steep a price point for what looked like a very elaborate Sokoban (in general, I hate Sokobans) despite the praise it was getting. I was incredibly wrong:

  1. SSR isn’t really a Sokoban since it’s much more than moving “boxes” into predetermined areas. It’s design also fixes everything frustrating and workmanlike of Sokobans, with small workspaces, limited movement of the sausages, and most importantly the forgiving nature of the game (there are only 2 or 3 ways to truly “trap” a sausage, so it’s easy to adjust things after realizing you don’t have things exactly set up the way you like; none of the puzzles require hundreds of moves done in a specific order with multiple moving parts).
  2. In retrospect, the game is truly worth $30. It is 30 hours long and is constantly re-inventing itself.

There are so many things I love about this game - the lack of pretensions, the density of ideas, the constant surprises, the fact that it actually has a plot. I can’t recommend it enough.

My only complaints are the game has two huge, discouraging hurdles. The first is the entire introductory island. I understand the game wants to throw you into the deep end to get used to experimenting, thinking outside the box, discovering things on your own (since that is the point of the entire game) but it basically requires a new player to hit their head against a wall for an hour before everything clicks. The first island is also the least interesting and I feel most people won’t be sold on the game until they get the “revelation” at the start of island 2 that makes you realize where the game is really going.

The second hurdle is the “Great Tower” level in Island 2. This is such an intimidating puzzle just on a surface level that’s compounded by how it introduces multiple new rules. It’s the meanest puzzle in the game and I know it forces people to stop playing. “The Great Tower” seems to say “If you’re not comfortable dealing with elaborate, scary puzzles like this, give up now” when the rest of the game is anything but that. SSR is actually a very welcoming, friendly game from then onward. “The Great Tower” feels like a filter that was never truly necessary.


Obduction was on sale on GOG and was excited to try it out. But I also couldn’t resist re-purchasing the grandaddy of adventure puzzle games, which I had been waiting to replay after cheating my way through it when I was like 12 years old.


Riven has been discussed to death and I don’t have much to contribute. A few random thoughts on replaying the game in 2017:

  1. The game 100% holds up and is actually better than it was 20 years ago. You don’t have to swap CDs anymore so it takes less than a 2 minutes to get anywhere (assuming you click fast and accurately) in order to re-examine something specific.

  2. Sadly, I still remembered the solutions to all the big puzzles of the game and the biggest surprises. Credit where credit is due: retaining these 20 year old memories so distinctly probably isn’t due to some great memory on my part but rather due to how memorable Riven is.

  3. It’s unfortunate Riven was basically a dead-end of game design (are there any other “anthropology theme park” games?). But if Riven was a tough sell in 1997 (when obtuse adventure game was still a genre), it’d be even more niche today. Riven was the ultimate pet project – an insane, AAA budget (self-funded), humorless, introspective game that expects you to spend about 10 hours wandering around until you hopefully get enough insight to solve it’s 3 or 4 puzzles. I consider Riven a one-time gift.


The opening 30 minutes of Obduction are great. Environmental graphics are nearing Riven’s photorealistic pre-rendered scenes and Obduction’s opening sells this well. The game drops you into an unfamiliar and bizarre world, let’s you explore a bit, and expects you to try to piece some information together. I’m not inclined to expand on this part of the game since I don’t want to ruin it for anyone who may play the game.

Unfortunately, the game steadily gets worse as you go on.

Riven was a game that expected lots of the player throughout it’s runtime, very little is telegraphed. Riven asked you to embrace the mystery, never gave you any hints, and ultimately rewarded you with a satisfying explanation for everything. This is the ideal for a puzzle game – making the player observe,contemplate, and experiment to what seems to be the limit and then rewarding them appropriately.

Obduction strangely expects less and less of the player as it goes on. The enticing mystery of the opening 30 minutes quickly becomes…a straightforward, cliche sci fi story. The exciting but intimidating goal of figuring out what is going on becomes…the more boring but concrete goal activating 3 identical “beacons.” The alien number system can be solved early on…but if you wait a few hours the game just spells it out for you (and worse still, it’s never really used for any puzzles). Most of the puzzles are just finding codes written down and knowing where to input them, no contemplation necessary.

The game climaxes with a large transporation puzzle (the central mechanic of the game) that is just a huge misfire. It’s not interesting, requires no in-game knowledge, and adds nothing to the game’s world. Worst of all, although it is not particularly difficult it literally requires 30 to 45 minutes of busy work. Before this point I was pretty neutral on the game but afterward I was angry at how little it respected my time.

The whole game feels undercooked. Maybe it’s because (as I later found out) the game makers underballed their Kickstarter goal, expecting to make 2-3x their goal but really only matching it. Maybe it’s the changing expectations of what gamers want. I do feel this game would have had more of an impact if it had come out before The Witness, which is a far superior game.

The game isn’t a disaster, it’s worth checking out for some of the environments, but it may be more worthwhile just replaying Riven rather than playing this.


Just grabbed this, and coincidentally played till level 23.

I was a bit scared-off by the maths, but happy to throw a couple of bob the guy’s way since i like his other, underpriced games so much. Enjoying it so far! glad he patched in the explanation of the directional stuff though. It still took me a few minutes to realise it, becuase the arrows are quite faint and small on my screen.

Protip: If you have his other games and you’re buying through Steam it’s slightly cheaper to buy the bundle. the other games get excluded, but you get a little discount.

Was just poking round Github and found a Hexcells level editor: https://github.com/oprypin/sixcells

Also there are a bunch of level solvers. Which seems strange, but I guess they’re more programming exercises then things people need/want to use?

Edit: Hmm I guess handy for making levels too. But Sixcells has one built in.

Edit: And here’s am open-source version: https://github.com/roSievers/elm-sweeper

Re: Riven, I think it is a game I appreciate more than I enjoyed, but I also played it under less than ideal conditions. A few dozen hard crashes work contrary to the mood the game is going for, I’d wager. I think being as “pure” as it is works to its disadvantage the few times things don’t work ideally. There are five “balls” you must find across the game in order to solve a mandatory puzzle. One of them is off the beaten path in a dense area that despite searching through multiple times I never found, and would have never found. Most games would cheat and find a way to work against such a thing happening; Riven will happily leave you stuck forever, or having to go through literally every screen clicking randomly. Again, appreciate it more than I enjoyed it.

I will have to look into this Hexcells fan scene.

I have been lazy and had something written up for CrossCells sitting half-finished on my comp for a bit now, a mention of it is a good excuse to finally finish it up.


The most recent game by Matthew Brown of Hexcells and SquareCells fame (the fact that the Cells isn’t capitalized in the former irks me more than it should), CrossCells is another fine numerical/logic puzzle game. It is still very much in the vein of those other games, albeit one that has moved a step beyond numbers and into mathematics.

The Cells games have always been played by determining which cells are to be marked and which are to be removed; that has not changed here. What has changed is a bit of how this is determined. Hex- and Square- have a bit of a similarity to Minesweeper in that the number in a cell can give info regarding the cells in its immediate vicinity, and this combined with other sources of info would be used to determine the correct solution. In CrossCells the numbers in each cell are just numbers (well technically numbers and either a + or x sign), the goal being to leave only the numbers/signs left that produce the value outside that given row or column.

This base concept is modified by several complicating factors (I swear I’ve written a version of this line so many times about various puzzle games…) that are gradually introduced to make it more than just basic arithmetic. The initial challenge is comparing the required final values for a given row and column and to use this to determine which cells to leave marked to solve for both of them. The number outside a row or column will then at times be bracketed, referring to how many cells are involved instead of an actual sum; a 4 in this case would not require the numbers to add up to four, just that four total cells are marked. Later on a section of the playing field will be surrounded by a box with a number outside referring to this entire quadrant rather than a single row or column. Once multiplication becomes involved you have to then deal with the direction of the order of operations. As an example if you have 1 + 4 x 2, by the game’s logic going in one direction that equals 10 and in the other direction it equals 9. Once this is established the game hits its stride.

This is a bit of a problem though, as by this point you are at around puzzle 30 in a game that has 50 total. It would be unfair to say that everything up until that point was a gradually escalating tutorial, but the sentiment isn’t without merit. The game takes a long time to introduce everything in five puzzle chunks and it takes roughly half the game to do so. I think it may have the most included puzzles of any Cells game and yet it almost feels like the least. I said earlier that I hoped SquareCells would get a follow-up like Hexcells did to further develop things. CrossCells almost needs a follow-up as there is just too much potential meat left on this particular bone.

The initial version of the game had a few missteps (the introduction to the direction you add/multiply making a difference was very rough in particular) but those have been mostly sorted out by now. One touch I did appreciate was that after SquareCells moved to uniformly square playing fields this game moved back to having more unique layouts, not as distinct as in Hexcells but enough to add a bit more variety.

So yeah, good game, not a revelation but it does lay a very solid base that hopefully will be revisited. Hexcells didn’t get great until Plus after all.

Also hey, Steam sale at the moment. CrossCells isn’t discounted on its own (still only $2.99) but the five game bundle that has all of these games is.

Bonus tidbit

I’m currently playing Sokobond, a chemistry themed sokoban-like. I am through maybe half of the included puzzles which isn’t enough to offer a full opinion, but there is a Steam sale that is currently marking the game at 70% off so let me just say that up to this point it has been a pleasant enough puzzle experience.

1 Like