TECHSTERS
Road Map For RoboCup 2D Soccer (RCSS)
Project Overview
• Objective: Pass and Defense
2. Passing Strategy
Delaunay Triangulation:
• Establish player connections by forming triangles to visualize passing lanes.
• Create a dynamic grid to update player positions and connections in real-time.
• Pathfinding with A*:
• G-cost: Calculate based on the distance to the goal from the passer.
• H-cost: Estimate based on the probability of interception, incorporating opponent
positions and movements.
• Finding the most optimal pass/series of passes
3. Defensive Strategy
• Zone Defense:
• Assign one defender to act as a "buddy" to the goalkeeper when an opponent enters
the defense zone.
• Implement algorithms to identify and predict opponent shooting zones.
• Goal Coverage:
• The goalkeeper prioritizes coverage of high-probability shooting areas.
• The buddy defender covers secondary areas, maintaining a flexible response
to the opponent's position.
• Opponent Marking:
• Implement a marking system where defenders block passing lanes and line of sight to
opponent players.
• Use heuristics to prioritize marking high-threat players based on proximity and ball
possession.
• Using MCTS (Monte Carlo Tree Seacrh) for defensive action
• Traverse the tree from the root node (current game state) to a leaf node using the
Upper Confidence Bound for Trees (UCB1) to balance exploration and exploitation.
• When reaching a leaf node, expand it by generating possible moves for the defenders
and opponents.
• Perform simulations by randomly selecting actions for both teams until a terminal
state (goal scored, out of bounds) is reached.
• Update node statistics based on the outcome of simulations (win/loss for the
defensive team).
4. Offensive Strategy
• Entering the Opponent's Defense:
• Upon entering the opponent's defense zone, analyze the goal's open sides using a
combination of brute force and strategic algorithms.
• Calculate optimal shooting angles and probabilities based on player positioning and
opponent behavior.
• Shooting Decision-Making:
• Integrate a "Behind Score Factor" to increase shooting probability when trailing in
score and time is limited.
• Develop a shooting selection algorithm that balances risk and reward based on the
game context.
