Happy Halloween! In the spirit of the season, I will take this opportunity to talk about the dead. Or, to be less melodramatic (and less seasonally relevant), another one of the world designs we decided to abandon along the way.

After we had cut out the Time and No-Paint worlds, we felt that our remaining worlds did not provide enough variety and as such we scrambled to come up with new mechanics – we did a lot of experimentation and came up with a lot of ideas, most of which were not interesting enough to warrant an entire world or straying too far from the core concepts of the game. One idea we spent a good amount of time on though, is the world that we tentatively called “Scroll”.


The concept of this world is a bit difficult to explain (one of the reasons we decided to cut it), but in short the front- and backworld were played on different background layers. This meant that backworld was much smaller than frontworld and this would affect scrolling – when the player moved through frontworld, backworld would scroll really slowly and consequently moving through backworld meant that frontworld would scroll really quickly.

In the example level above, the distance from the middle platform to the hovering platform is way too large for the player to make the jump, but by making sure both platforms are still in the frontworld while the air between them is in backworld the player could cause the jump to take them the entire distance since the level was caused to scroll much faster.

We initially liked this idea a lot since it had some major benefits – one of our core design goals is to make sure that using the worlds in tandem is encouraged – if a puzzle could instead be solved as easily by switching the entire world with a keypress, we would not be playing to the unique strengths of the painting mechanics. Scroll encouraged puzzles that took advantage of transitions and it had good uses of one-world physics. In addition, neither world had any inherent advantage and the mechanic worked by enabling movement rather than restricting it so we avoided the problems we had with the Time levels.

Some of the inspiration for Scroll came from Ludo Land’s recently released Four Sided Fantasy – a really nice game that you should check out.


The displayed level is another example – by pushing the statue to the backworld platform, the player could take advantage of the different scroll speeds and move the platform around by walking in frontworld to get the statue to the trigger.

We added a few more objects and built some 5-10 puzzles with this mechanic before we decided to give it up, there were some problems with the individual puzzles but the main reasons we decided to give up on the entire world was as follows;

  • Building this world was technically complex, it had lots of edge cases that placed heavy demands on the game engine to work around. While we try not to shy away from ideas just because the implementation is challenging, building this world would have taken a lot more time than other worlds, and the systems we had to change would impose restrictions on other mechanics as well so it became very costly.
  • To get back to the previous point of explaining this world being difficult, while playing it we noticed that it was sometimes very difficult to anticipate what would happen when you moved between worlds. While Backworlds is fundamentally about understanding how to use the worlds to solve puzzles, we do not want the threshold to basic understanding to be high.
  • Since the worlds were different sizes, the frontworld always had to be very large which meant that you frequently had to walk long distances. This made traversal really tedious.
  • The mechanic made the camera have a very active role and made it more or less impossible to change it later without having to redesign levels. We are still thinking about camera movement (more on that in a later post) and did not want to commit to something we maybe could not change later.

… So ultimately the world was cut. Sadly I do not have any video of how this worked in practice as the functionality made the physics code difficult to work with and I removed it after we decided to cut the feature. But from a development standpoint, our reasoning behind this decision was very different from any other world.


In development news, we are iterating on design and art and I am still rooting out remnants of the gamejam mindset when I find them. Lately, this took the form of a physics change – for dynamic objects, we resolved their solid parts at the beginning of every frame but this made the actual simulation order-dependent since some objects would simulate against old versions of other objects. Changing that has been a careful process, but hopefully it will lead to more stable levels.

In other news, Indiecade occurred earlier this month and as always it was full of fun and inspiring games – the aforementioned Four Sided Fantasy included. A personal favorite this year was the time-looping adventure Elsinore,  keep an eye on that one early next year.