Puzzle Whiteboxing for No Evil

Puzzle Whiteboxing for No Evil

This blog will journey the puzzle whiteboxing process for one of my recent projects, No Evil.

 

Introduction to Puzzle Whiteboxing

Players begin in the same room together and travel down a corridor. The length of this is kept short enough so that players can interact with each other for long enough without feeling like the introduction drags on. They are then forced to split up with one player going into the room below and the other player continuing forward.

A full overview of the maze. Cylinders and cones represent landmarks. Each cylinder is a crane or tower of coloured crates whilst the cone is an observation tower (a unique landmark to signal the end of the maze). Each crate blocks a part of the maze that the player explores. Once the colour of the crate that the crane is carrying is identified, the crane will move the crate to another section of the maze. This has been colour coordinated in the picture above. For example, the cylinder with the red cross can have the positions of the crates move to either of the red circles within the maze. Green crosses indicate other potential places for cranes based on player feedback and how many other positions are needed. Walls for the maze are used as a placeholder in this instance, as they do not necessarily need to be walls. Rather, it can be a case of debris to give the map a more open feel, or it can be closed off, with breaches in the ceiling to allow the crates to move. This was later developed into the aesthetic of a cargo bay, with crates being used as the walls of the maze. Emphasis on sensory deprivation was key when designing the levels and revising any changes suggested by playtesters or team members.

The red cross in the picture above denotes the first section of the maze that players encounter. This represents an open hallway where the player will have the option on which path to take. This has the opportunity to be expanded or kept the same size upon discussion. An option for Box positioning to give randomised gameplay elements was considered but was not implemented as it was tasked as a stretch goal. Corridors have been kept narrow to have the contrast of claustrophobic gameplay but the comfort of having someone guide the player through the maze.

The overall size of the maze should be sufficient enough given the player will be slowed down due to the lack of vision. Considerations need to be made for failure states as gameplay progresses. As this is an introduction, failure states aren’t important to implement but will be given future consideration.

 

Puzzle Prototyping and Whiteboxing

The puzzle itself requires one player to be inside the maze and will be tasked with reaching the end. Crates will be blocking the players path, requiring the other player – the navigator – to help the player by pressing the corresponding button relating to the crate that is blocking the player in the mazes path. The crates themselves will move to a predetermined point in the level. Pressing the same button again will move the crates back to the original position. Some of the crates in the level can be used to plug up pitfalls within the game world for players to traverse to different areas of the maze. The twist involves the navigator having colour removed whilst the player in the maze has very limited vision. Below are the different buttons that are being used in the first version of the whitebox.

The room for the navigator also houses a map containing the different locations within the maze. This is useful for communicating landmarks that help the navigator triangulate the position of the player within the maze.

Revised Implementation on Cranes

New cranes have been added to the whitebox to present more variety in the colours that can be communicated. These new cranes are marked with a Red ‘X’.

Therefore, new crate placements will need to be justified. These will be represented in the diagram below, but can always be opened up after further playtesting. Each crane has been represented with its own colour through an ‘X’. The scribbles located on the maze itself represent potential locations that the boxes can spawn during the initialisation of the game, and where they can be moved too once players move them. Each scribble and ‘X’ is colour coded to symbolise which crane does what crate.

One problem with the current implementation is how does the player guiding the player in the maze regroup after they have left the maze. Whilst one way is having a corridor open and that player walking to meet up, that would take too much time and be a boring walk for both parties. One idea that could show through playtesting would be the other player moving through the maze after the other player but having the maze change up. This can be done through cranes falling over after the player opens the door, leaving only one path for the other player to traverse through. This can create a sense of surprise for both players and gives a unique way for the players to regroup. Below is a diagram illustrating how the cranes would look once they have collapsed, and a critical path that the player can follow to regroup with their partner.

The critical path in this example is illustrated with a red line. The animation will not be necessary as both players won’t have eyesight on any of the cranes. The player navigating will have clear feedback telling them to move through the maze (Lighting, Voiceovers).

Whilst the landmark of the crane does help to guide players towards the exit of the maze, it should be noted that a path within the maze would also help with navigation of the maze. Further additions to landmarks could include numbering the cranes to show a steady sense of progression. However, Including different shaped lines in the maze may help the player inside the maze. This could be put forward towards playtesters to see their opinions. But for now, the diagram below shows the different types of paths. Colour has been used on the diagram to give a clearer understanding.

The colours are shown as such:

• Red: Critical pathway that the other player should take once the cranes have fallen down.
• Blue: Critical pathway that the first player should take to the exit
• Pink: Extended pathway to the potential key (or dead-end)
• Cyan: Extended pathway to the potential key (or dead-end)
• Lime: Scenic route 

Changes to Design – Phase 1 (navigator helps the player in the maze)

After playtesting sessions, it was discovered that there was a lack of interior space for the player to explore through. It was also found that the North East section of the map would rarely be explored if at all. This was because there were no points of interest to consider. Therefore, changes needed to be made. There was a possibility of 3 changes to choose from. Each of which is detailed below with a diagram and explanation.

Option 1 would involve completely removing the area. This would be accomplished by adding a pathway that cuts across the maze and removing the top areas. Whilst this does solve the problem, it felt that more could still be done with the area given proper time. As a worst-case, this approach could be used to keep the idea simple to work with. 

Option 2 would include joining the north section and removing the southern connector. This forces the player to at least move through a part of the maze. However, this not only removes choice for the player but also creates longer corridors; something that is currently trying to be avoided. Therefore, a combination approach will be adapted to allow for player exploration whilst still keeping original parts of the maze intact.

The conception for this plan involves including extra interiors for the player to explore. As a stretch goal, it’s possible to populate these buildings with extra puzzles that the player can solve on their own, or through communication with information that the player navigating has. Each of these interiors occupies a section of empty space near the maze to ensure the player doesn’t have to wander far away from the maze. The main inclusion is the interior at the top of the map that bridges the connection between the start and the North East section of the map, giving players freedom to explore that section of the maze should they choose too. Below is the final incorporation of the extra buildings.

Playtesting also showed us that whilst players recognised the landmarks, they weren’t able to distinguish between each of them apart from their size. To try and rectify this, each of the landmarks has a unique icon on the sides and from above. This will allow for the player in the maze to see what building is closest to them when looking from any angle and the player navigating will be able to see the buildings icon from a birds-eye view. Buildings and landmarks will have a distinct landmark location on the side and on the top of the building. This helps both the player in the maze and the player navigating to have a grounded understanding of where the player is. These will take the form of chess pieces whereby black pieces will denote landmarks (cranes and the observation post) and white pieces will denote places of interest (buildings across the map). These colours have been chosen so they can contrast well against the backdrop of the buildings. Below are screenshots from the updated in-game map, and where these symbols have been placed for the player to navigate through:

Puzzle Whiteboxing Changes

During playtesting, it was found that the game world lacked any direction that informed the player as to why they’re doing what they’re doing. The puzzle lacked any context to the players situation which meant that players wouldn’t try to remember which button did what, but instead would move through each button in a way that didn’t promote lateral thinking. Therefore, designing a puzzle with a clear way to identify how each button works will prevent this type of spamming on each button. Should spamming continue, deterrents will be put in place to help prevent this type of puzzle-solving behaviour. 

Solution 1: Each crate has a number that can easily be identified by the player in the maze. This number is relayed to the player within the navigation room. Turning OFF the colour module will allow for numbers to appear on the buttons. Each of these numbers correspond to the numbers on the crates. Whilst this is the simplest solution and does empower players through giving the solution by making the player take away something they find comfort in, it is very simplistic and doesn’t have much depth to the puzzle element.

Solution 2: Each crate has a roman numeral associated with each crate. The crates in the maze will not be coloured. In the navigation room there will be a coloured chart that gives a value to each colour. For example, Red would equal I and Blue could equal V. Each of the buttons will have a combination of colours such as one half being red and the other half being blue. Should a crate have the number VI on the crate then the answer would be the red and blue button as they both equal the same number (6). This solution adds more of a puzzle element to the game. However, there is no reason for the player to initially turn off their colour vision. Therefore, another aspect in the navigation room would need to be developed in order for the player to need to use their colour module in an empowering way. One example of this would be using the map to highlight key areas in the maze for the player in the maze to move around.

Solution 3: Having the colour module turned on or off will reveal different types of information around the navigation room. These types of hints would give the player the location of items in the maze to relay to the player in the maze. Clues and building locations would be seen with the colour vision off and the colour vision on will allow for the corresponding colour of the crate in the maze to correspond with the colour of the button in the navigation room. Whilst this solution allows for a mixture of different reasons to flick between having colour vision on and off and contextualising this to the player, it does lack in puzzle solving thinking as the relationship to each button is very much one to one, unlike the complexity of the multiple colours mentioned in solution 2.

It was decided that Solution 2 would be the fix to this problem, primarily for its ability to make the player switch between the two colour modules to work out an answer to the problem. (See Map Room Puzzle Functionality document for further information)

 

Changes to design of the Puzzle Whiteboxing – Phase 2 (Player in maze helps navigator)

Once the player in the maze reaches the end, the level subverts and the navigator must now be the one to enter the maze whilst the other player. Not only does this extend the time the level is played for, but also changes how the players have been playing the game. The player in the maze will be able to see the other player from a vantage point that is gained at the end of the level. Currently, no buttons are within this vantage point that can interact with the game world as the navigator moves through the maze. However, this can be a possibility later in development. As cranes fall in the level as the player in the maze reaches the end, routes within the maze have been cut off. The diagram below illustrates the blue lines whereby the cranes will cut off irrelevant points on the map. The red boxes surrounding some of the blue lines indicate where the player will be able to move over these obstacles. This indicates new areas where the player in the maze couldn’t pass through originally. Another screenshot indicates the positions of these cranes within the whitebox itself.

The player will have the top right section of the map cut off with cranes tilted to block certain maze paths. This also makes it easier for the player in the observation post to better guide their partner through the maze. New routes for the navigator along with blocked paths provide a fresh twist for subverting the original maze whilst not needing extra development time for assets or programming. The possibility for adding extra functionality could be a possible function but is currently a stretch goal for our team.

 

End Of Maze Changes

To allow for a clear end game and to ensure that both players can regroup in a natural way rather than be teleported. To incorporate this change, the watchtower at the end of the maze has been converted into a proper watchtower in the whitebox. Once the player reaches the end of the maze, stairs will allow the player to move up and support the second player as they make their way through the maze. The cranes falling over once player one should be considered for this action as it allows for a new challenge for navigating the maze. Below is a screenshot that highlights this change.

Pressure Plates

During playtesting, it was apparent that players did not know when to separate from each other without being told externally; the game does not tell the player that they could go one way. Therefore, players should be provided with a means to progress at the cost of leaving one person within the other room to be in charge of navigation. This was decided through the use of pressure plates. By blocking the entrance of one room and presenting two pressure plates (as shown in the image below), players should be incentivised to stand on these. Once stood on, the path ahead will open whilst a door will close the other player inside. In order to release this player, they must reach the end of their maze for their door to open. 

It is possible for these pressure plates to be placed in different locations. For example, if one was placed in the navigation room, players could figure out who would want to stand in what area, and could guess what role they will play in the upcoming puzzle. This gives players a chance to decide who wants to do what should they have the lateral thinking to arrive at this conclusion.

Furthermore, a pressure plate will be located at the end of the maze. This pressure plate will signify the end of the maze. Once this is stood on, the door in front will open. Once the player moves through, the pressure plate on the other side will close this door and open the door in the navigation room, allowing the other player to regroup by going through the maze. This setup is also pictured below such as the red pressure plate opens the door to the exit and the blue pressure plate closes the door behind and opens the door where the other player has been navigating.

It is important to change how the maze is laid out when the second player is moving through, ultimately swapping the roles of each player. Changing the layout of the maze will involve moving the cranes as shown previously to block off certain routes. In order to communicate this, the player that has moved through the maze will be able to access the observation post to provide real time feedback to each player. An early screenshot of the view from the observation post is shown below.

Ceiling/Roofing

Different sections of the maze will have different types of ceiling. Not only for a variety in the environments that the player explores through, but also to elicit different emotions based on the type of room that the player enters. Current ceiling explorations have been detailed below.

 

Starting Room/Crew Quarters

This type of design is to enable both a feeling of freedom, but also a feeling of claustrophobia as the ceiling reduces in width as the length increases. This can be changed to allow for more or less layers.

Walkway Towards Separation

To keep inline with giving an open environment, a concave ceiling was employed. This will allow players to feel at ease before hitting the end of the hallway. At the end of this section,  a flat ceiling is used (pictured below). This could be changed to reflect the same design as the one previously mentioned to continue down the hallway, or to have a sharp change in the ceiling.

Main Lobby Area

As before, the lobby area deploys a large open area for the player. However, as the player will be blind at this stage, it may be possible that they won’t be able to see the top of the room. This is a good thing as the player will feel a sense of uncertainty, only to be filled with some comfort when their team mate guides them through the maze.

Within the Maze

The maze itself has been opened up fully to reflect the openness of a cargo bay. It also has the added feature of justifying how cranes can move from one part of the maze to the other. However, as the player in the maze won’t be able to see much in front of them, it’s plausible that covering up the ceiling with a flat surface or with another ceiling design.

Connector (White Knight)

Through playtesting, it was apparent that the top right section of the maze was hardly moved through. Therefore, to give it a reason for players to explore, a connector building has been implemented. This connects the top sections of the maze together should the player decide to go East rather than South. Below is an interior picture to illustrate the insides, a diagram in the section ‘Changes to Design’ gives an overview of this.

The Octagon (White Rook)

As the number of interiors in the maze was lacking, extra rooms were put into the game to encourage players to engage with the maze rather than just running to the end of the level. This in itself should hopefully put a bit of strain on communications as the player in the maze will want to explore other things compared to the goal that their partner wants to do. This level of tension should produce an interesting experience for both players. The extra interiors, both with Connector and The Octagon may be fleshed out with puzzles later in the game. However this will primarily be a stretch goal. Below is a screenshot of the interior of The Octagon.

The Barn (White Bishop)

As with The Octagon, the primary use of this building is to create a sense of exploration around the maze that may cause a conflict of interest between players. As mentioned before, this may include some sort of puzzle or other type of reward. However this should only be mentioned as a stretch goal. Below is a screenshot of the interior. Once further playtesting has occurred. These buildings may be removed based on player feedback.

It should be noted that these rooms would be populated with additional puzzles to solve. However, due to time constraints, these have been cut. 

Fail States

Fail states have been represented in the game through box meshes. The idea is for the player to fall down these holes should they not be careful. Any instance of failure will mean that the player is teleported back to the start. All cranes will have their positions unchanged if a fail state is triggered. Feedback will be presented to the player as a visual indicator on their UI.

 

Penalties

After gathering results from playtesting sessions, it was clear that the solution to the main puzzle in terms of the navigator was to spam the buttons on the walls that moved the crates in the maze. Therefore, a solution was needed that made sure the player refrained from just spamming the buttons. Two problems were discovered with the design of the puzzle meant that the puzzle itself didn’t make sense to the player. 

Firstly, there wasn’t a relationship between the buttons in the navigation room and the crates in the maze; there needed to be a link between these buttons to insight thought that will allow players to think before resorting to spamming buttons. 

Secondly, there was nothing preventing players from spamming in the first place. Even if players have a reason to work out the correct answer, there is no penalty for spamming the buttons and solving the puzzle using brute force. Therefore, having restrictions in place as a deterrent to prevent players from abusing an attempt at brute force is the ideal solution. Possible solutions could involve:

• Having audio feedback so players know when they have pressed a button.
• Delaying the next time a player can press another button.
• Disabling the players’ module for a certain amount of time, having the modules inverse will visually tell the player not to spam the buttons as its important to change between the modules when solving the puzzles.
• Restarting the player in the maze to the starting position, although this would be seen as a last resort due to the jarring nature the player in the maze would feel.

 

Playtesting will need to incur to see if these types of preventions are suitable for the game that we are making or alternatively to see if better solutions could be made. What is important is making sure that the players are grounded in a puzzle that makes sense within the game world. Having no context means that players don’t understand that they are in a puzzle environment and that communication is needed to solve the puzzle. 

 

Whiteboxing Assets

As a guide to asset development, a separate document is provided detailing all available assets used to build the level. As for asset placement, Assets throughout the map should include some exposed wiring around the walls. No obstacles should block the way through apart from the crates attached to the crane. However, parts of the environment may block player access as long as it’s not part of the main pathways. For example, boxes or furniture in the corner of a room. Parts of the level can be opened up to allow for a closer look at the cranes to allow for identification of where the player in the maze is located. Lighting will play a big part in both the maze and in the navigation room. The navigation room will use light to highlight the buttons for the player to press, the map to help navigate the player and any clues within the room that will help with communication between the players. The player in the maze will have light to easily identify the maze locations, crane identification and crate identification.

 

Puzzle Whiteboxing the Crane Falling Section

Due to limited resources and time constraints, implementing phase 2 (pg 14 to 16) of the maze whereby the cranes fall down to provide the second player the opportunity to move through the maze has been cut. Whilst this was intended to give more features for the second player to do, this has been balanced out by making the puzzle more of a puzzle rather than an opportunity for the player to spam through each of the provided buttons. This extension to the pre existing puzzle should allow for the player in the navigation room to have more engagement with the game through communication and helping the other player in the maze. This also includes the tower at the end of the maze for which one of the players can walk up to gain a vantage point across the map.

 

Implementation of Assets into Puzzle Whiteboxing Level

Once Assets have been built, changes can be made from the Whitebox to include more detailed environments. Note: A duplicate was kept of the original Whitebox incase problems occurred when swapping out assets. Asset changing started with simple external assets before moving onto building the main portions of the maze.

 

Cranes

Cranes were the first asset to change out due to their ease and simplicity. This also allowed a familiarity of the assets to be established before longer tasks needed to be undertaken. The screenshot below showcases the implementation of these cranes. The chess pieces have been left in for identification purposes but will be taken out and replaced with something more appropriate in the final version.

Crates

Puzzle crates simply needed to have their static mesh changed from the original UE4 cube to the puzzle box mesh that has been made. This allows the retention of any scripts and keeps everything working how it was originally. Crates also changed the appearance of buildings within the game by keeping the functionality of the room but replacing the way it is built with. Each change is detailed below: 

 

Spawn Section and Corridor

Replacing these assets allowed for a more futuristic look to the whitebox without sacrificing too much in space. The spawn room size was reduced to accommodate for these new assets. The corridor leading down towards the pressure plates has also had its look changed, with the original ceiling being discarded for a cleaner look. Doorways themselves are clearly labelled through detail in the environment rather than blank hallways.

 

Navigation Room and Maze beginning

The pressure plates outside of each of these respective rooms have been given a more realistic look alongside sci-fi doors that make the level further represent the theme. The navigation room itself has been reduced in size. This is primarily due to playtesters walking around the previous space and not spending the time helping their partner navigate the maze. Reducing the amount of space in the room should allow for players to be more focused on helping the player. In addition, increasing the amount of game information that the player can look at and interact with should also keep player engagement high and lead to communication with their partner. 

The maze beginning has also been updated, now with clear doorways and Celestial Comforts logos on the walls of the interior. This was done so that if players were to fall into any pitfalls or activate any fail states that would teleport them back to the starting room, then the logos on the walls should familiarise players where they are, and should allow them to continue the game without being confused on their whereabouts.

 

Whiteboxing the Maze Interior

The maze itself has been constructed through the use of the crate asset. Cylinders have been replaced by crane models and White walls have been replaced with cranes. Aesthetically, crates have been used to give the maze a different look and feel. Rather than just have straight walls, crates will allow for a more natural feel to the maze and give more opportunities for less tight turns within the maze. The crates themselves can also be positioned to give the effect of them falling over. This allows for a different look at the environment that can make the player feel uncertain, thus facilitating more communication between them and their partner. Stacking the crates on top of each other can also allow for more ways of identifying landmarks within the game as stacked crates in different positions.

 

Sound’s use in Puzzle Whiteboxing

Sound is important for the feedback of player actions. The sound itself has been implemented into key pickups, button presses, doors closing and pressure plates. All of these are essential to gameplay as it lets the player know that an action has been performed. This feedback makes the player aware of the game world rather than question if the action they’ve performed has been conducted or not. Following on from this, the decision to not include ambient sound is an important one for a few reasons. Primarily so that players can talk to each other clearly and not have any distractions from ambient noise as well as focus on the main feedback sounds that players will hear when an action is performed. As a secondary effect, players will feel isolated by hearing the silence in this section of the game, and should therefore heighten the need for communication between each other.

 

Lighting

Lighting within the interior has been used for 2 instances. Firstly, for lighting the inside of the maze so the player can see the different paths they can take. It was also used to light assets that players can use to navigate. This can include cranes, buildings or stacked crates. The use of light in this instance should draw players’ attention to areas of the maze and question their partner about what objects are over by the light.

 

Main Menu Implementation

A simple main menu was needed within our game to give a transition rather than just starting the game abruptly. At this stage in development, only a simple menu is needed. However, should the development of the game continue, extra features would be included both for accessibility and gameplay. UI had been considered during development, but with the scale of the project and the payoff from including it into the game, it was scrapped in light of extra time to work on polishing and bug fixes. 

The menu introduces the player to the logo of ‘Celestial Comforts’ for the first time in the game. A logo that will be prevalent throughout the game. Having this logo early in the game gives a recognisable tool that is used in later sections of the game to re-orientate players should they trigger a fail state and need to respawn within the game.

Once players click the play game button, they are presented with this second screen. As this is only a technical demo, this message will be displayed to players to advise them to take a different route that doesn’t complicate the gameplay. However, this is only advice and players are still welcome to try the other section of gameplay. The full game would include full-colour blind accessibility options to aid with players’ sight. 

 

Final Polish of Puzzle Whiteboxing

After adding new lighting effects to give the maze more of a space feel, the maze itself has been completed. Extra additions could include adding in the cranes falling down section to subvert the maze and reverse the roles of both players. Further considerations could also have been made for giving players inside the maze more control in opening sections for the player in the maze navigation section to have more tools to access that help the person in the maze. However, the maze itself functions and was well received in the hands of playtesters.

 

For more information on how to create a brilliant whiteboxed environment, I recommend reading this blog by Yuzhi Zhu Steven. In it he whiteboxes a level based on “Prometheus” and “Alien”.