Skip to main content
Formal Specification

Hungarian Horseback Archery (Kassai Method) — Formal Specification

Hungary · equestrian-team · Confidence: low

← Back to game page

Implementation Suitability

Suitability: adaptation-requiredFidelity: probabilistic-proxyGap-fills: 1
# Lovasíjászat — AI-optimized specification ## General idea A high-tempo mounted archery simulation where an agent manages horse gallop velocity along a 99-metre straight corridor while executing rapid nocking and projectile aim decisions against three directional targets (incoming, perpendicular, outgoing) within a strict 20-second time budget. The core optimization problem balances riding speed against nocking cadence and aiming accuracy. ## Formal model - **Abstraction taken:** The continuous 99-metre run is discretized into real-time state ticks (tick duration = 0.1s; total run = 100..200 ticks). Gallop velocity `v(t)` determines track position `y(t) = y(t-1) + v(t)*dt`. Arrow flight and target hit resolution use ballistic angle checks and agent skill parameters. - **Player/agent count:** 1 to N agents, competing sequentially. - **Turn model:** Continuous-time ticked execution during active run (9 runs per match). ## Formal rules R1. State Setup: Track length `Y_max = 99.0m`. Horse starts at `y = 0.0m` with initial gallop velocity `v_0 in [5.0..10.0] m/s`. Target Tower at `y = 49.5m, x = 9.0m`. R2. Sector Boundaries: - Sector 1 (Forward Target): `y in [0.0m .. 35.0m]`. - Sector 2 (Side Target): `y in [35.0m .. 64.0m]`. - Sector 3 (Backward Target): `y in [64.0m .. 99.0m]`. R3. Agent Action Loop (per tick `dt = 0.1s`): - `adjust_pace(acceleration)`: Sets `v(t) = clamp(v(t-1) + a*dt, 5.0, 12.0)`. - `nock_arrow()`: Requires `N_ticks = floor(baseline_nock_time / nock_skill)`. - `aim_and_release(target_sector_id, aim_offset_x, aim_offset_y)`: Releases nocked arrow. R4. Arrow Resolution: - Distance to target: `D = sqrt((y(t) - 49.5)^2 + 9.0^2)`. - Hit error: `E ~ 2D_Gaussian(0, sigma_archery * (D / 10.0) * (v(t) / 6.0))`. - Scoring rings: If `E <= 0.15m` -> 4 pts; If `E <= 0.30m` -> 3 pts; If `E <= 0.45m` -> 2 pts; Else -> 0 pts. R5. Run Completion & Time Bonus: - When `y(t) >= 99.0m`, run ends at timestamp `T_final`. - If `T_final <= 20.00s`: `Run_Score = sum(arrow_points) + (20.00 - T_final)`. - If `T_final > 20.00s`: `Run_Score = 0.0`. R6. Match Structure: Each agent executes 9 independent runs. `Match_Score = sum(Run_Score_1..9)`. Highest total score wins. R7. [gap-fill] Arrow Limit: Maximum 10 arrows per run (bounded by quiver capacity). ## Edge cases - Tie at the end condition: In the event of a tie in cumulative points across 9 runs, highest single-run score wins; if still tied, fastest average run time wins. - A player with no legal move: Not applicable — agent can always hold, adjust pace, or nock/aim. - Repetition / cycles: Bounded — horse velocity `v(t) >= 5.0 m/s` guarantees crossing 99m in at most 20 seconds per run; match consists of exactly 9 runs. - Simultaneous claims on the same resource or position: Runs are purely individual and sequential; zero interaction between agents on the track. - Player elimination mid-round: If a rider exceeds 20.00s on run `k`, that run scores 0; the agent proceeds to run `k+1`. - Resource exhaustion (pieces, tokens, space, time): Arrow count resets to full at the start of each run. - Termination guarantee: Enforced by minimum forward velocity constraints and finite run limits. ## State representation - Persistent state: `match_scores[agent_id]`, `completed_runs_count`, `active_run_index`. - Per-run ephemeral state: `position_y`, `velocity_v`, `elapsed_time_t`, `nock_progress_ticks`, `is_arrow_nocked (bool)`, `arrows_fired_count`, `hits_breakdown[3]`. - Observable information: Full observability — agent tracks exact position, elapsed time, and target angles in real time. ## Action space - `set_gallop_pace(accel_m_s2)`: Adjust forward speed within valid gallop bounds. - `nock_arrow()`: Initiates nocking cycle. - `shoot_arrow(target_sector_id, aim_x, aim_y)`: Releases arrow toward sector target. ## Playability notes - Single-player-vs-AI: Highly playable as an arcade/tactical timing challenge against historical record benchmarks (e.g. Kassai's master-level scores). - Group games (n>4): Leaderboard tournament model across identical track conditions. - Physical-skill games: Dynamic horse biomechanics and blind nocking converted to continuous parameter control and probabilistic accuracy checks. - What is lost: The physical rhythm of horse canter synchronization and visceral wind resistance. ## Implementation hints Low-to-medium complexity. State updates can be stepped with a simple Euler integration loop for `y(t)`. The primary design tuning is balancing the nock delay time against the speed bonus curve to reward rapid shooting without trivializing accuracy. Suggested first milestone: implement a 1-run simulation with fixed nock delay (1.2s) and test optimal velocity policies.

Download as JSON

Download the full game corpus entry for this game as structured JSON.

Download JSON
Cite this page

Hungarian Horseback Archery (Kassai Method) (Hungary). (n.d.). Games of Countries. https://gamesofcountries.com/games/hu-hungary/lovasijaszat

@misc{gamesofcountries-hu-hungary-lovasijaszat,
  title = {Hungarian Horseback Archery (Kassai Method) (Hungary)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/hu-hungary/lovasijaszat}
}