7: Level Framework, Data Collection
With our build objects created, we’ve moved on to the UFO and alien assets. We have replaced our previous goal (star block colliding with the star symbol) with the newly modeled and textured UFO and alien. This meant adjusting the collision point so that the goal is now achieved when the player gets the UFO aligned with the alien’s feet. As we continue developing assets, we will be adding an idle animation for the alien while the player creates their structure, a celebratory animation for when the UFO is placed properly, and a lift-off sequence once the alien materializes into the UFO.
To match the new art style, we will be making background element changes. We plan on creating alien landscapes to tie in with the build objects. The environments are going to be low in detail with vibrant colors to engage our audience.
Level Framework & Creation
From here, all the level designer needs to do is integrate level-specific assets and features, such as pitfalls or cliffs, as well as change the location of the stranded alien. With this system now in place, we have now begun to implement the tier 2, or wider base, level designs. We aim for a playtest with these levels late next week.
Last week we decided to go with the drag mechanic instead of the click-and-drop method. One implication of this is that items dragged out of the inventory already have physics applied and so can’t pass through cliffs or blockades designed within the level. To better accommodate these levels, we moved the inventory to the top of the screen (above cliffs and blockades) and resized the objects.
The ability to collect analyzable data from the students who play the game is key to the long-term success of this project. We are using HCII’s DataShop format to organize this data. DataShop was designed around action-response pairs, but decisions in our structure building game are subtler and more multi-faceted than that, so one of the challenges for HCII will be to figure out the most effective way to capture data through which a child’s decision-making processes can be inferred. Statistical correlation across large pools of children is likely to be an important way of doing this. We’ll be testing our first transfer of data from the game to DataShop next week, and then HCII will recommend adjustments to the data we collect throughout the semester.