top of page

Tower In The Forest
Level Design Project

Tower in the Forest was a personal project I made for my Level Design II course while at Champlain College.

The goal was to create a bite-sized version of the main game's experience, and to gain familiarity with the Level design process.

Status - Complete (9 Weeks)

Skills - Level Design, Puzzle Design, Lua

Design Goal - Non-Linearity

Going into this project, I sought to try and emulate the basic structure of a single Grimrock II overworld area. Without going into too much detail, Grimrock II levels are typically set up as a series of small, independent challenges, which each reward the player with either items or a power gem, the game's primary collectable.


To that end, I divided my level up into a similar manner, separating it into similar smaller chunks that were self-contained puzzles. Each rewarded the player with a Golden Key and an item required to get into the tower at the center of the map.

Screenshot 2023-04-24 at 10.37.57 PM.png

Fig. 1a - Initial Level Layout (Uses In-Engine Tiles)

The Golden Keys were not necessary for progression, but could instead be turned in at the forge to gain upgraded equipment for one of the player's four characters.

As the scope of the level was limited enough that actual XP gain wouldn't allow the players to gain additional experience levels, this was the player's main way of gaining strength.

Design Issues - Onboarding

One of the issues that I ran into when creating this level was the problem of onboarding. Since this was made for a pre-existing game, it didn't have time to teach the base game's mechanics as extensively as the game itself could.

On one hand, this isn't necessarily a problem for the final product; in theory, anyone who played this game would already own and have at least started the base game before downloading any custom dungeons.

However, I still took steps to address it, particularly for testers who were unfamiliar with the base game. Additionally, I think that onboarding is important regardless, to establish the basic mechanics that are going to be used to a potential player.

Screenshot 2023-04-24 at 11.28.05 PM.png

Fig. 2a - Starting Area w/Blocker

To that end, I started the characters out in a completely safe area, which contained equipment designed for each party member. This forced the player to re-learn item interaction, and gave them a good set of starting equipment for their characters.

This area was sealed off by a breakable set of twigs, which both reminds the player that this mechanic exists, AND forces them to learn the basic combat mechanics in order to progress further.

Design Process - Rapid Iteration + Rescoping

To facilitate creating the level, I followed a simple process of iteration; the level was first proposed in the abstract, using a document which went over each intended feature of the level, all of the essential items in it, and included gathered reference images that I could use to get a feel for the level's place in the world.

Fig. 1a - Level Proposal Document

From there, I moved onto a blockout stage. At this point in time, I primarily focused on building out the level with only the essential elements, such as doors, key items, and any environmental features required for puzzle progression.

Screenshot 2023-04-25 at 1.17.12 PM.png

Fig. 2a - Maze Blockout

The benefit of doing things this way was that I could focus on functionality before decoration; getting a sense of how the puzzles worked in isolation and making sure they were fun on their own. However, as I moved into actually needing to decorate the levels, I ran into a problem.

Actually boxing off the area so that the player could not see the outside skybox required more space than I initially guessed, requiring me to "shrink" the level inward to claim more space. This meant that I had to rescale all of the existing areas to fit within these constraints, causing a few puzzles to be reworked.

Ultimately, the problem was solved, and I think it actually improved some of the puzzles by forcing me to compartmentalize them, removing downtime and padding from the actual challenges.

bottom of page