dogs 2

                                      (fan level)











This ongoing personal project is intended to be used as an exercise to expand my level design knowledge. The idea is to take the systems of Watch Dogs 2 and design my own level. I want to carry on working on it but this is what I've done so far. I'll show what I've done after roughly 2 weeks in development.       



  • Sketched and blocked out the layout of the level.

  • Scripted the mission.

  • Scripted the mechanics used for the level.

Level Design Pillars:

  • Put player in charge of their actions.

  • 360 approach.

  • Make player feel like ultimate hacker.

Level Overview

  • Play Time: 5-10 min.

  • Setting: Wave, Modern Fictional Streaming Music Provider.

Ideation Process


During the exiting first steps of the ideation process, I like to let myself go and draw a variety of ideas for possible top-downs. At this stage, I usually focus on brainstorming beats and layouts that would suit the intended gameplay as well as gathering as many references as possible.   

Sometimes, I would jump into the engine or level editor to quickly blockout those ideas that seem interesting. By doing so I can "answer questions" like for example verticality, that the limited top-down cannot do for me.   



This time, I wanted to add a new tool to my workflow called Sharp 3D. I used this very iterative tool to get an overview of the possible building where the level happens. Even though the building is fictional, it was important that it would feel like an actual structure to suit Watch Dogs' world design realism.

Once I've sketched a top-down that I'm pleased with, I start doing what I call "blockouts passes". I tent to start as general as possible and then slowly focusing more on the minutia of the level. With no exception this time, I looked at things like cover or enemy placement later when there was a space to play and move around.

Mechanics & the gym

As part of my workflow I like to make use of a "Gym". A Gym is a space set up to test different mechanics, scales and metrics. This became more helpful than ever because I wanted my level's metrics to be as similar as possible to those of WD2.  

Since I want to keep the same workflow that I would follow if this was a team project I created a "Library". A Library is a space to store the modular assets that I would use for the final blockout, before art comes in. Except for the trees, the pipes and the human models, all the assets were created by making static meshes of Unreal's BSP. 

To achieve a better feeling of the mechanics and systems of WD2 in my level, I figured that it was important to prototype similar mechanics and systems. I made a selection of those WD2's features that the player would use and so I gave myself some days to step out of the Level Design and start scripting.   

Camera Hack, camera interaction with 

Junction Boxes and AI reaction to Junction Boxes.

Deploy Jumper and Jumper interaction with JBoxes. 

Interaction with Scissors Lift.

This was an amazing way to grow familiar with WD2's systems as well as to improve my Blueprint skills and my logical thinking. My main focus is to always find optimal approaches which implementations would take as little time as possible.

For example, for the Scissors Lift I used Unreal's vehicle pack.

1. I placed the base box, the part of the lift that doesn't rise up, around the actual mesh of the vehicle. 

2. Disabled collisions between these two meshes. 

3. Used the already exiting vehicle logic to drive around.


Then I would only need to focus on the behavior of the part that the player rises up plus some other minor things.   


By analyzing WD2 I realized that many of the missions follow this core loop

1. Scan mission area with surveillance  cameras.

2. Sneak past guards and security systems.

3. Hack specific devices to proceed mission.

4. Escape the mission area.


I wanted to keep this loop intact and challenge 

myself to design a level around it to get a result that will, hopefully, resemble WD2. In the future I will expand the level design to make sure that it suits playstyles like: aggressor or pacifist.   

This mission is intended to take place start/mid-game. As many missions in the original game, I wanted to keep simple and straight forward objectives and let the player figure out things on their own. 

Agency, AI & Covers

I wanted the player to predict outcomes and be as active as possible when taking decisions, such as WD2 does. Thus, I placed the surveillance camera network in such way that the player could scan almost the entire mission area if they want to.  

From the very start, I wanted the player to know what they could expect from this level. I thought that by adding multiple entry points to the building, some more hidden than others, the level would set its tone of openness and freedom. Besides this may encourage the player's creativity and make them feel smart when succeeding  with their plans.

I used the AI as my way of suggesting, not forcing, the player to do certain things in the level. For example, by placing them in the way, the player would either choose to confront them or find another route to pass by them. 

I also planed AI and cover placement to create challenges which the player could approach in multiple ways. Creating a foundation for playstyles such as: pacifist and stealth.

I kept the number of AI down to a maximum of two in each corridor or room to not

overwhelm and outnumber the player. This may promote the player's sense of control. 

Breaking agency

To make full use of the Jumper, I took the liberty of designing an objective that can only be cleared with the robot friend. I wanted to try to incorporate a more linear beat within the freedom of the level.

I believe that the possible player's reactions could be:


1. Total frustration since they don't have another choice than to clear the objective as intended.

2. A sense of cleverness and surprise when realizing what the vent leads to. 


Once I've done external playtesting, I would know whether to keep this beat or iterate on it.