The final project I designed is a mystery game which is now built up in Unity.

The original story was about a middle-aged man woke up in a strange island only found out that his car was crushed and the sea level was rising. And the player’s task was to help him escape from the island. However, since I’m so interested in the program, plus people love those games which would apply their detective skills, I decided to make it more complicated and tricky. With some feedbacks, the new story is adapted as follows. The player wakes up in a cave in an isolated island, and the only clue left for him/her is an instructional note given by an unknown person telling him the rules being on this island and his 10mins life duration, and the key words include cars, buckets, gas(showing in the game), etc.  After the player gets away from the cave, he/she would not be allowed to come back to the cave again. Instead, he/she would see a crashed car and a middle-aged man and some debris scattered around the island. Following the notes, player could easily find anoth oil drum and he/she could pull it toward another crashed car, which would then lead him/her to another clue -- a secret hidden in the valley. He/she would risk getting lost in the valley without any progress and wasting the limited time, since it's my goal to construct the scene as a huge maze -- and even myself have constantly been lost in the scene when testing. Alternatively, he/she could take the plunge explore the center of the island. If he/she is prudential and lucky enough, he/she would directly find the whale, which holds the fourth hint. Still he/she would have to find out food from the rest of buckets and feed the whale to get the last instruction, which is to find the bench. Here in this scene, the bench is equivalent to success, the player could just wait for the boat and finally escape from the island.

I love my idea, but there are also several things that don’t work which force me to make changes at some places. First, the man is able to jump on the mountain even on a precipitous cliff. In the map-setting, I want to divide the whole island into several regions which would make the map as a maze, but this buggy makes the plan hard to accomplish since the players could easily cross different regions by climbing mountains. Second, the speed of the rising of the sea level is hard to control, so that the original storyline is difficult to achieve. Third, my original idea is to make it as a detective story. To be more specific, I want to set up at least five clues, and put them in a certain sequence which means every previous clue is a trigger of the next one. As a result, player can follow the hints one by one and eventually escape from the island. But unfortunately, I’m not able to accomplish the one by one trigger sequence and have the scene still remains realistic, so I have to put everything in a same status and using notes to guide the player, which makes the game less interesting than what I originally thought about. Besides the three major problems, there are also others. Such as, I cannot figure out a way to set up hands on a First Person Version like what Counter Strike (actually most of the FPS game) did.

The online Unity tutorials from Lynda.com and YouTube were fairly helpful. Aside from the basic of the Unity we learnt in class, in those tutorials, I learned how to use the navigate mesh to set the scope of walking in the game, as well as inputing sounds, making visual effect, and designing the Playmaker status to logically trigger events and stories.

The way I pivot for the project is constantly changing to the play mode and test whether the triggers are successfully placed and whether the First Person's viewpoint looks realistic, such as to make sure the ratio of the size of the car is the same as the ratio in the reality. I learned a lesson that before putting everything in a program, I need to ensure every single trigger of the whole story is well-organized, otherwise it will waste a lot of time checking for errors.

Unity is tricky to some extent. It really hard to handle at the beginning, but we have to be patient. And one thing needs special attention is that we should consider both entering the event and exiting one since FSM setting is a logic loop. One tiny mistake could lead to a serious failure.