Hi people, recently I have found this page of Center for Computer games Research and it looks like, they are accepting PhDs! So I have write down few proposal, please give me feedback so I can form solid Statement of Purpose from it. It is 3 of them, 1 main and 2 spare. Fell free to be bad and have a lot of criticism.
Here it is:
Real-World data based game content - basic idea is to generate game content based on actual real-world data. There could be multiple uses for this and it is wide area. Examples:
1. Games content delivery - you can deliver game content based on geographical location of player, in different places you can play different levels of game. It can be used for example for propagation of cultural attractions, where if you visit museum or gallery you can play game on your phone, and levels of this game will be personalized to that museum or gallery. Goal of this would be to build a game, which delivers game content depending on geographical position and other easily obtained data (time) from client (player).
2. Solving real world problems - lets imagine a computer game, which reflects real traffic situation and quest of the player is to get from point A to point B. There could be multiple players and one who can get there in best way will be the winner and actually his solution can be delivered to someone, who will be using GPS and would like to get from point A to point B. Next example is to have game with climate situation and the quest of the player will be to manipulate with the weather. Based on player actions then we can observe the resulting situation and compare it to actual situation evolved after time. Those are simple examples, we can think about more complex models, which provides player with ability to solve real-world problems with computer games, based on real-world content. One bright example can be found here.
3. Gameplay in real-world environment - Here you can imagine computer games, mostly on mobile phones, which are using augmented reality to provide game content. Provided game content is also dependant on current surrounding of the player. This would be nice combination with point 1. Under this point one can also discuss game-content based on real world models. Just imagine you have a model of a car, you would move your mobile phone around that car, the car get 3D into the mobile phone and you can play the game with that car. Actually this, what I am describing in last sentence is based on video-based rendering and I have already some results in that area.
Those are 3 main possible areas, where to use real-world data based game content. If none of them would be suitable, I have few ideas here for other themes for thesis. For this other themes I have not deep background, but it is not problem for me to work hard to get it. So here are few other ideas:
Indirect control gameplay - The name is little bit mysterious, but have you ever play Black and White game? This is one of a kind game and it is using this indirect feedback model. You have a creature, which is acting in the game according his own will and you as player can punish or reward it. Also this creature is observing players behaviour and actually it reflect players style of play. I think it is pretty interesting area to observe. For example in strategy games, when one is playing for a race and he will build a new base. So he will have 2 bases. Then he can promote and general, which will take care for the old base and he can concentrate on the new base. The player can give some indirect orders to the general and also promote/demote him. This can result into forming generals character. This is example for strategy game, same can be done for racing game, where teams are racing against each other and in one race there is multiple racers, who need to cooperate. Here it can be also interesting to get the good balance between player feedback and agent autonomous behaviour.
Adaptive computer-player cooperation in gameplay - I have read your papers on adaptive content and how to maximize player fun. This was one of other factors, which can be taken in consideration. Besides player there could be another character in the game controlled by the computer and this character will help the player based on his current results. So if he is doing well, the character will interact less, otherwise he will help more. This could be really nice addition to generated game content, because it has no delay in response to player actions (if we are thinking about Mario game, then computer controlled character can help also on the same screen so player does not need to wait to move to another screen).
