Similar to the player design process, the level design process started off with basic sketches, with the aim of teaching the player how to play through level design — rather than having a hand holding tutorial. So the sketches began, with designs and gameplay features that would end up being dropped until a later time, but they gave me designs to work off and begin prototyping the game with.
There are three goals for each level:
- 1- Complete the level
- 2- Complete the level within Par
- 3- Collect all the zombie heads
The level Par is the lowest amount of waypoints possible used to reach the exit, the player can use as many as they wish, but shouldn’t be able to use any less.
I wanted the first level to be easy, and to help the player learn what the goal is — place waypoints to create a path to the exit. The first level is easy, it’s impossible for the player to die and only 1 waypoint is needed to be placed. There are zombies, but they’re trapped behind a set of barricades, at such a distance that they will see the player and try to attack him, but be unable to because of the obstruction — thus showing the play that zombies will chase Jim if they see him.
In the screenshots below, you can see in the sketch that the start position isn’t in line with the exit, this was changed during development to ensure that the player would go past the zombies. We can also see how the UI started out, and how it finished — though it did go through a couple of other looks before reaching the finished state.
When the zombie head collectables were added, during play tests, players were avoiding the heads as well as the zombies — it was my hope, initially, that the UI in the top left (along with the level Par for level 1) was obvious enough to show that these are to be collected, and not avoided. Due to this bit of feedback the instructions will be getting an update to add this info in.
On to level 2, another tutorial level, which is exactly the same as the first level, except for 3 differences:
- 1- The barricade blocking the zombies was removed
- 2- The collectable heads have been moved down 2 rows
- 3- The player has moved up 1 row, and is no longer in line with the exit
The barricades removed to allow the zombies to chase the player, and turn him into a piece of steak if caught. The collectable heads moved down to show the player (gamer) the correct path, by putting 3 squares in between the zombies and the heads, I’m attempting to show that this is the distance that the zombies can see — 3 squares infront of themselves. Lastly, the player was moved 1 row up to show that it’s possible to walk on the pavements, as well as the roads, many testers were unaware of this at the start.
In the paper sketch, for level 2, you’ll notice that there are two cars present, this level was originally planned to demonstrate how to climb objects, however this feature was left out to be introduced into later levels — eventually being left out of the game entirely, until a later time.
From early on in development, I knew I wanted each level to be set during the day and at night — night time having more zombies, less light, and a harder solution.
The solution for level 1 at night is the same as it is during the day, the main purpose of this level is to display that night time has more zombies. It is level 2 where the player is taught that night time has a different solution to the day time version. The Par for night time levels is usually higher than the day time versions.
To make the level creation easier, so that any changes for the day time version wouldn’t then need to be copied to a separate scene for the night version, each level has 1 scene for both day and night. When a level is loaded it grabs a public variable, set at the main menu, to load the correct setup, it finds all the spawn points for the enemies and all the lights in the level. The enemies have two types of spawners, a normal one that spawns enemies for both day and night, and then a night spawner, which is only turned on if the level flag is set to night. All the lights are turned on by default, the script goes through turning off the lights which aren’t needed for the level flag.
In the screenshot below, you can see 4 types of spawn points:
- Green: Player start position. Though the player is not spawned, as it’s already in the level, it is just moved into the correct position
- Red: Exit point.
- Yellow: Enemy spawn points, for day and night
- Dark Yellow: Enemy spawn points, for night only
- Blue: Collectable head spawn points, both day and night spawn points are the same colour here, but easily tracked which is which in the hierarchy as they are childs of separate empty game objects
Something I found when creating a couple of Portal 2 maps, was that it’s incredibly difficult to judge your own puzzles — you know the answer, is it too easy? too hard? This was even more prevalent developing Shh, Zombies — each level needed two lots of puzzles/ solutions, one of which affected the other. Game statistics were added early on in development, to track player deaths and successful completions, and a lot of iterations were done before reaching the launch designs for each level.
Since release, gathering more stats it’s become apparent that the bigger levels aren’t keeping the interest of the players, level 4 had a 50% drop rate of people completing it, compared to levels 1–3. I believe this to be because of the bigger size in level from 3 to 4, and the difficulty spike isn’t quite right — an update was released which changed the position of the player start to make the solution a bit simpler. Moving forward with the new set of levels currently in development, I’m using this statistic to create better levels, I’m already happier with the ones completed thus far and think them to be better than a few from the original bunch.
I’m going to finish this blog post off with a screen of level 11, the first level in the next update, which shows off the look of a new enemy type — if you’ve played Shh, Zombies already, reached level 10 and have a keen eye, you’ll already have spotted a preview of the new enemy type.
