Game Evaluation - Sophia

Figure 1: Colin from our game Sophia

Me and my friend Gilbert, a student in my class, decided to create a game about mental models. We had all types of possibilities in realizing the game but since we both know 3D modeling and myself can code in C# and use Unity engine, we decided take on the challenge of making a digital Game. Knowing from the beginning that we were not graded on visuals nor on media used to transmit the purpose of the game we first focused on creating a well-structured story continuously referring to mental models.

After a lot of brain storming and idea generation, scraping and re writing levels, we came up we a pretty decent story. Also, we had to be fast yet delivering good quality since we had just three weeks to work on it.

Although the game is not yet finished and not polished, one can already play all levels and get the idea of the mental models involved. For the time being we left the narration out which will be narrated by ourselves during game play. This was done in order to keep things simple and fast to have a playable game. There are still some bugs here and there but should not effect the story yet will obviously effect the overall experience of the player. But since it is in its Alpha testing we are not worried that much.

We split the roles in making the game, where gilbert took care of creating all of the assets, textures and loading screen renderings while I programed mechanics, animations and drew sprites. The style itself of the game is simple on purpose to be easy and fast to make but yet effective. We designed everything revolving around basic shaped mostly cubes. We always kept in mind the style we were using and mental models, everything we put in levels has a purpose in effecting the player moods and emotions.

During the development of the game I was playing and testing the game several times, but there is a huge difference in playing for testing and playing for fun, even a bigger difference when you already know the outcome. Each level is designed in such a way for the player to explore and find solutions of getting the key back to the princes but while I was playing, since I knew the solution, I could not truly judge weather the situation we created was effective for the player in believing what we wanted him to believe without ever outlining it. On the other hand, I was exploring and trying every possible outcome that a player might try. Also, I was sending several builds (versions) of the game to Gilbert as he also helped me to point out bugs and fixes we could do.

Although that we knew the outcome we were still able to point out certain problems, like how the player is going to notice the trap to kill the monster in level 2 if he is not an experienced gamer? So as a fix for that in the future we will add more lighting and even a sign with a drawing on it to make it clear that it’s a trap. Still we will never say that you must kill the monster, we are just offering the possibility.

Another thing was that walking in level one and two can become tedious so we decided that we will include a sprint button to ease the pain or frustration in the player.

A list of things we intended or came up during testing but we did not manage to include are:

• Dialogue between Colin, the main character, and the princess

• Sprint button

• Gold Bar collected indicator

• Armor indicator and helmet on face showing health status – already designed just need to be programmed in the game.

• Armor itself which is already modeled - it would slow down the player

• Monster sign with key - to indicate that the monster has the key

• Trap sign – to emphasize the existence of the trap

• Final animation of princess unlocking door and kiss her boyfriend prince.

• Sound – one of the most important things we need to include to engage the player in the game. This may also help in elevating fear in level 2 when the monster smashes his arms on the ground.

One problem was that the more levels created, the harder it became to build. This is our first game, at least this big, so we encountered several problems. One of them was Lightmapping the level, this is a process where the light is precomputed and baked as a texture in order to have a better performance during gameplay. What happened was that these lightmaps where taking too much time to produce and while producing them I barely could use the computer since they consume a lot of resources. On the other hand, without lightmaps we couldn’t distribute the game to players to playtest. Finally, at some point, all lightmaps apart from level 3 where completed and immediately I uploaded the game on a cloud to spread it among our friends and game test, only to find out that everyone was either too tired or too busy to play and the game was due the day after!

Luckily, we found a test subject and she played the game with Gilbert reading the dialog for her. Gilbert did not interfere or influence the player in anyway, so that we could pick up relevant data from the experience. On level one, the player noted that it would be nice if the player could sprint and if we also would include an indicator of the gold bars, which both happened to be planned. This shows that when were testing, we were still able to pick up flaws rather accurately and not being biased. On level two the subject went strait for the monster instead of escaping away and this lead to the monster to crush the player where in the full featured game would result in death. She tried it two more times, and both of them she went back strait to the monster resulting in her character’s death. At that point it seemed obvious that it was no longer viable in continue playtesting with the same subject.

We needed an experienced gamer, preferable someone who loves puzzles. So, we searched harder and Gilbert managed two find another two. They were very helpful in pointing out difficulties in the game which we didn’t even think of and took it for granted. For example, one of them would have liked to have an indication of the furnace or at least some information before hand in order to recognize and know its function. Yes of course the princess tells Colin that he must cast the key in the furnace but by putting a signal it would be better for the player to identify a usable object from aesthetics. This would take us bake to the lesson of attractor’s. As mental models still worked as intended there were more suggestions to make them work even better. The other player suggested that we could make the monster react and come towards the player before. This way will be more effective in inflicting fear in thinking that it is chasing you rather then it turns on when you are almost under it. Something that both players mentioned and even it hurts us that we didn’t have time to include is sound. They stated that with sound the experience would have been much better and that is something that we will be working on once we have a more solid structure of the game.

Apart from flaws, bugs and incomplete content we are very happy with the outcome since we managed to put this game together in only three weeks. While still unfinished, it is playable and it also reflects the mental models we had planned, although not as effective as we intended. I believe that by further testing and working on the game, it will be able to pass the intended message to the player.