Rat Race

Starting page
Leaderboard

Inspiration

We looked at current trends that is popular among university students in Singapore, such that they are able to relate to. One of the current trends is LinkedIn, as we are in the phase where we are finding for internships and jobs, some of us will make the extra effort to decorate and polish our LinkedIn posts. So we drew inspiration from that and decided to go with this particular idea for this hackathon.

Our game is also inspired from The Password Game. The main interface involves a text box, and interesting rules to create a "strong" password. Although rules are not prominent in a LinkedIn post, they naturally serve as guidelines shaped by the hustle culture of the platform. Seeing how the game fits our idea, we decide to implement it.

What it does

This is a rule-based game. Players will have to type out a LinkedIn-like post, but there are quirky rules to follow, and it also runs on a life-system (there will be 8 lives). As the levels progress, there will be harder rules to follow.

Press the submit button to move on to the next level. If all rules have been fulfilled, they can progress to the next rule, and in turn adding points to their total score. If they didn't fulfill all rules, a life will be lost. The game will end if they lose all their lives. At the end of the game, they will be submitted to the leaderboard with their username and score.

How we built it

The project is built using React.js (front-end), Express.js (backend) and SQLite (database).

Challenges we ran into

The first challenge we ran into was to ensure people are typing actual LinkedIn posts. Since it is a text box, users can actually type anything and form incomplete sentences. As long as they fulfil the rules (e.g. include keywords), they can carry on.

We have two methods in place to tackle this challenge. The first method is to use the Language Tools API to check for grammar. The text is sent to the server, and it returns in the form of a list of grammatical errors. The presence of errors will be considered as failing the rule. The second method is to use Gemini API. Both the text and the prompt is sent to the API, and it returns with Valid / Invalid. With these two methods, we can ensure the legitimacy of the LinkedIn posts.