🍋 Lemonade Stand 🍋

Game Overview

The Lemonade Stand Game is a simulation where players compete to run the most successful lemonade stand. Each day, players make decisions about:

How many lemons to buy
How much lemonade to prepare
What price to charge per cup
How much to spend on advertising

Players can compete against each other or against AI opponents. The game runs for a specified number of days, and the player with the most cash at the end wins.

AI Function Requirements

Function Signature

def your_ai_name(player_state, game_state, prices, previous_info):
    # Your logic here...
    return {
        'lemons_bought': 0,      # Integer: Number of lemons to buy
        'lemonade_prepared': 0,  # Integer: Number of cups to prepare
        'price_per_cup': 1.00,   # Float: Price to charge per cup
        'advertising_spend': 0   # Integer: Amount to spend on advertising
    }

Input Parameters

Parameter	Type	Description
player_state	dict	cash (Decimal): Current cash balance lemons (int): Current lemon inventory
game_state	dict	current_round (int): Current round number total_rounds (int): Total number of rounds in the game weather (str): Current weather type (e.g., 'Sunny', 'Cloudy', 'Heatwave')
prices	dict	lemon (Decimal): Current price per lemon
previous_info	dict or None	None on round 1 Contains previous round's information including: round_number: Previous round number total_potential_customers: Total customers available last round your_sales: Cups you sold last round your_decision: Your previous decisions all_players_decisions: All players' previous decisions

Game Parameters

Basic Settings

Game Length: 5-365 days
Starting Cash: $10-$500
Base Lemon Price: $0.01-$10.00 (fluctuates daily with ±20% random variation)
Base Customers Per Player: 5-50 per day (randomly selected each day with ±25% variation)

Customer Behaviour

Price Sensitivity (0.0-1.0):
- Higher values make customers more likely to choose lower prices.
- Lower values make customers less price-conscious.
- Applied as a multiplier to price comparisons between stands.
Sign Sensitivity (0.0-1.0):
- Higher values make advertising more effective.
- Lower values reduce the impact of advertising.
- Applied as a multiplier to advertising effects.

How Customers Choose a Stand

Each stand is assigned an attractiveness score based on:
- Price: Lower prices increase attractiveness (weighted by price sensitivity).
- Advertising: More advertising increases attractiveness (weighted by sign sensitivity, relative to total advertising spend).
For each customer, a stand is chosen at random, with the probability proportional to each stand's attractiveness score.
If a stand runs out of stock, it is removed from the pool for the rest of the day.
There are no individual customer types or budgets; all customers use the same market-wide sensitivities.
Note: This is a weighted random choice, not a Boltzmann/softmax function.

Weather Effects

Each weather type has:
- Customer Modifier: Multiplies base customer count (0.1-3.0).
- Likelihood: Relative chance of occurring (1-10).
Weather is randomly selected each day based on likelihoods.
Customer count is multiplied by the weather's modifier.

Events

Events can occur randomly each day based on their chance (0-50%).
Each event has:
- Chance: Daily probability of occurring.
- Duration: How long the event lasts (1-10 days).
- Effects: Can modify prices, customer behaviour, etc.
Only one event can be active at a time.
Events affecting ingredients modify prices by their multiplier (0.1-10.0).

Daily Simulation Process

Weather is randomly selected based on likelihoods.
Base customer count is calculated and modified by weather.
Event chance is checked and applied if triggered.
Each customer:
- Evaluates all stands based on price and advertising (using market-wide sensitivities).
- Chooses a stand at random, weighted by each stand's attractiveness score.
- If their chosen stand is out of stock, they do not buy.
Results are calculated and stored for the next day.