About datacr4sh


Ludum Dare 30

datacr4sh's Trophies

datacr4sh's Archive

TETHER: Postmortem

Posted by
Monday, August 25th, 2014 10:54 pm

I wanted to jot down a few of the key ideas that spawned TETHER in its current state, what I learnt from making it, and what might happen to the game in future. You can play it here.

The core mechanic of the game is the concept of transferring jump momentum between the two characters, allowing the player to control jump height. This mechanic came about as a happy accident, I had intended it to simply be a single jump, but the way I coded the inputs meant that holding the buttons while a character was landing instantly made the other one jump, which visually looked like momentum was being exchanged.

Before I did anything else in the game, I made sure this act of jumping felt really good. This is extremely important to this game especially, since the player is doing nothing but jump, and they really have no choice as to when. The shake, sound and physics of the jump I took a lot of time making, and the game felt the same as it does now from about 8 hours into the comp.

An earlier version actually had up to 4 jump heights that built off of each other. I cut that idea simply because it was an unnecessary complication to the mechanic. If a player saw the arrow prompts and was suddenly bouncing around at 5 different heights, they would give up after their 2nd death, because they would not understand how they did it and what they’d have to do to replicate the “right” jump.

Now, I could just include a separate tutorial/text wall explaining the game in full. Personally I consider those approaches as bad as movie exposition: actors talking to each other about an event for the purpose of filling in the audience. It’s a visual medium, show don’t tell. In an interactive medium, showing is fine but playing is even better. This is how I went about doing that:

  • The Boost-Jump: In the “tutorial” section at the start of a fresh run, there is a small jump placed immediately before a bigger one. The bigger one requires a boost jump to clear, and since they are so close together, the player will press jump within the boost jump window upon landing, subtly forced to discover it. After that, there are a few more that need boost jumps, so that players can practice the move, and confirm what they learnt. If a player completes this part, it is skipped for future runs.
  • The Controls: The only controls are left, right, A and D. These are keys that are usually mapped to a horizontal axis of some kind in other games. I wanted to take out the arrow key prompts, but decided that it would have been too far a simplification. Players would have been thrown in with absolutely no help, and quit after their first death. What are they supposed to do if they die after trying to jump with the wrong key? Probably close the game. So I made it show up on the first playthrough, or until the player completes the “tutorial” stage.
  • Advanced Controls: The game also includes a float and a stomp move for each character. Holding the jump key after a close call will have players realize that they are falling slower, and there are plenty of situations in which the player might do that. Same goes with being about to hit an overhead spike; a player would unconsciously see if they can descend quicker by pressing the opposite direction. It’s gaming intuition to “press the button harder” or “lean into a turn”, so I bottled that and turned it into an actual useful action in game. Originally, the player could use up, down, W and S to scroll up and down the screen in order to avoid being in a tight spot, instead of the float/stomp. It felt worse, and I wanted players to commit to a jump they made, and not be able to escape the consequences in a direct way.

The game is intentionally quite hard. In the small amount of time I had, I knew I couldn’t make a huge level and expect the player to be entertained. Making a game hard not only makes it longer lasting, mastering a tricky part feels great. But this is a very thin line to tread. The lack of checkpoints was a bit of a gamble for me in this respect, but I really wanted to maximize that euphoria of mastery and really nailing something in one go, the auto-runner genre really amplifies this feeling.

Because of the difficulty, dying MUST be interesting/entertaining in some way to offset some of the frustration. The response for the chain breaking sequence took me quite aback, people really liked that part of the game which really made me happy. If someone dies in an unforgiving game and loves it, that’s really big for me.

That’s all I really want to say about it, other than my plans to bring it to mobile in an extended form. It will likely involve procedural level generation for an endless mode, as well as a short narrative. Follow my twitter for updates on that, if you like :)

Thanks for reading.

[cache: storing page]