Inspiration

Throughout sixth form, we played poker together every lunchtime. Over time, we realised that success went beyond the cards we were dealt; by reading our opponents’ behaviour and anticipating their decisions, we could gain a strategic edge through skill and psychological insight.

What it does

T- Tracking over time, The bot records a players decisions across multiple rounds and identifies patterns I- Intelligence, The bot learns how the player makes decisions, such as when they take risks or play cautiously M- Memory, The system remembers previous rounds and builds a player profiles that evolves as they play. E- Evolution, The player's profile changes over time as their strategy and behaviour develops

How we built it

We used OpenAI to develop the foundation of the game, then independently coded the bots’ playing styles and data collection functionality.

Challenges we ran into

One of the key challenges we encountered was developing a reliable way to quantify opponents’ playing styles, particularly their aggression and tendency to play weaker hands, while statistically estimating the probability of winning with a given hand. This was complicated by the limited availability of data at the start of a game, forcing us to determine how the bot could make effective decisions with little prior information and how much data it needed to collect before identifying reliable patterns and adapting its strategy accordingly.

Accomplishments that we're proud of

It is a functional game. A lot of time went into research of how data management and collection actually works and how we can program a bot into using the relevant data effectively

What we learned

We learned how much data needs to be collected and managed before we can identify reliable trends and patterns.

What's next for PokerBot

Our next goal is to develop multiple distinct player profiles that simulate a variety of playing styles and expand the game to support multiple opponents simultaneously. This will enable the bot to adapt its strategy to different personalities and behaviours, while making decisions in a more complex, multiplayer environment.