Five Nights at Wall Street

Watch the trends! Can you increase your stock portfolio in our trading game?

Comment

Inspiration

Our fascination with the complexities of the stock market sparked the creation of "Five Nights at Wall Street." This title echoes the five days that encapsulate the highs and lows of market engagement, resonating with the intensity that traders endure. Seamlessly fitting into the theme of this year's HooHacks Hackathon, the game serves as a playground for aspiring traders to sharpen their day-trading skills within a risk-free environment. It's designed to simulate the thrill of real-time trading and the strategic decision-making required to profit from stock market trends. Through immersive gameplay, players can learn to analyze the volatility of stock values and hone their ability to forecast and react to market movements.

What it does

Become a day-trader! We created a stock market game where players act as day-traders on a time compressed day in order to improve their decision making in real-time. Our game simulates 5 consecutive historical days randomly in the past 10 years for a random stock on the S&P 500. "Five Nights at Wall Street" plunges players into the fast-paced world of stock trading. Our game's heartbeat is a dynamic graph displaying real-time stock values, moving from the opening bell at 9:30 AM to the closing bell at 4:00 PM, all within a compressed timeframe of 30 seconds. After being briefed by the instructional modal, players dive into the trading arena with a starting capital of $10,000. They'll execute buy and sell orders while the clock ticks down each day. The "Next Day" button propels them into the following day's trading session. The game's swift pace and substantial capital allow players to explore various trading strategies over the course of five days. Finally, they'll see the results of their decisions reflected in their profit margins.

How we built it

We used 1 minute interval intraday data from Alpha Vantage's API, which we extracted from the month CSV for each ticker on the S&P 500. Since API calls add up and can be expensive, we stored the month data for future use. Thanks to Intel's Developer Cloud, we are able to host large amounts of S&P 500 CSV data ranging from the last 10 years.

A majority of our frontend is vanilla HTML/CSS/JS + tailwind CSS, but our most notable elements is our dynamically updating chart, which we were able to create using Highcharts. This unique chart adds 1 minute intervals from the stock data to the graph every 50ms. This feature gives players an immersive and interactive environment, simulating the pressure and decision-making tempo of day trading.

Challenges we ran into

Constructing a real-time graph that could handle the swift changes in stock prices without lag was a technological hurdle. Finding cost-effective hosting solutions that could accommodate our large datasets without compromising on performance was another significant challenge. Navigating through intricate stock data, forming the correct queries, and ensuring accuracy demanded meticulous attention and posed a steep learning curve.

Accomplishments that we're proud of

We're particularly proud of our graph's ability to accurately depict the ebb and flow of stock prices in real-time, synchronized with user interactions. Our innovative approach to data storage—archiving queried CSVs—safeguards against potential API issues. Also, the hackathon's theme is captured in our game's title, "Five Nights at Wall Street," and the cohesive design of the interface are testaments to the team's attention to detail and collaborative spirit.

What we learned

A lesson learned was the importance of comprehensive project logistics planning. Adequate foresight into the required steps, resources, and tools is crucial for smooth execution. Delving into the project, we also gained deeper insights into the significance of efficient data handling and the power of user-centric design in crafting engaging applications.

What's next for Five Nights at Wall Street

The team is extremely excited to push this project forward – we aim to introduce layered complexity with varying difficulty settings. Advanced levels may introduce tighter time constraints and multi-stock management to simulate a more demanding trading floor. Personalization is on the horizon, with plans to empower users to select specific companies to trade. We're also exploring the potential of integrating an AI assistant, trained to interpret market trends, that could provide users with insights to refine their trading strategies. As we continue to develop the game, we anticipate creating a more robust and nuanced educational tool for aspiring traders.

Additionally, we were thinking of ways we can create a fine-tuned instance of LlAmA-2 with improved NLU and reasoning capabilities with RAG using the stock trends as datastores in order to better understand market trends in different sector, and with such knowledge, help the user understand ways to improve their day-trading decision making.

Share this project:

Updates

Submission history