Posts tagged Game Dev
The first RTP alert was replaced in six days: from ±1.5 points to "can't judge yet"
Just past 10,000 equal-stake rounds, the demo slot gives 57.9% false alerts with a ±1.5-point band. Its replacement withholds judgement for unqualified models.
Setting tolerances for Monte Carlo tests: standard error, false rejections and bit-exact replay
Monte Carlo tolerances can use z times the standard error plus an approximation allowance. Fixed limits may reject valid results or miss bias as N changes.
FreeCell deal numbers: reproducing Microsoft's deals, the Internet FreeCell Project and #11982
Reproduce Microsoft FreeCell deals in Python, check deals 1 and 11982 against published layouts, and see why a deal number alone does not guarantee solvability.
Tetris 7-Bag: The I-Piece Drought Bound and How to Count a Drought
The 7-bag caps I-piece gaps at 12; pure random and idealised NES rerolls have no cap. In the NES model, a 12-piece drought has odds of 14.0% after an I and 12.5% in any window.
The Third SET Card: Mod 3 Pair Completion and the 1,080 SETs
Any two distinct SET cards determine a unique third card via mod 3. This gives 1,080 SETs in 81 cards and at most 66 lookups to search a 12-card layout.
Uniform Solvable 15-Puzzle Boards: Parity Test and Two Generators
Uniform shuffling with a fixed parity-correcting swap gives every solvable 15-puzzle board the same chance. Random walks guarantee a solution but give uneven samples.
Sampling Bias in Maze Generation: Randomised Kruskal vs Wilson's Algorithm
Kruskal samples spanning trees unevenly; Wilson samples uniformly. Across all 7! edge orders on a 2x3 grid, Kruskal produces each tree either 360 or 300 times.
No-Guess Minesweeper Boards: First-Click Rules and Solver Verification
A solver can filter no-guess Minesweeper boards from the chosen start. Opening odds and completion rates help choose between retries and perturbation.
Solvable Lights Out Boards and Level Generation with GF(2) Linear Algebra
GF(2) elimination shows why a quarter of 5x5 Lights Out boards are solvable. Pressing each button with probability 1/2 from a dark board samples them uniformly.
Unpredictable Game RNG in Production, Replayable in Tests: Separating the Two in Go
Use a Source interface to keep production RNG unpredictable and tests reproducible. Covers HMAC-seeded ChaCha8, recorded draws for disputes and an AES-CTR deadlock.
The Faster You Mine, the Faster It Tips Over: A Game Economy as a Dynamic System
Sensible prices can still produce runaway wealth once output, upgrades and costs feed back into each other. Difference equations, feedback, delay and Monte Carlo on a mining economy.
Pity Is Not a Higher Drop Rate: Gacha Pity and Gear Upgrades as State Machines
Pity counters and upgrade downgrades make each roll depend on history, so model them as states. Three rules with the same average differ more than threefold for the unluckiest 1%.
A 1% Drop Rate and 100 Tries Without Luck: The Geometric Distribution and Weighted Sampling
With a 1% drop rate, about 37% of players still have nothing after 100 tries and the unluckiest 5% need 299 or more. The geometric distribution, then two ways to sample a loot table.
Same Ship, Different Flight at 30 and 144 FPS: Integration and Fixed Timesteps
Multiplying by deltaTime is not enough to make physics frame-rate independent. Update order, damping and how time is accumulated all leak through; measured in Python.
Facing Is Not Position: Vectors, Dot and Cross Products and Coordinate Spaces in Games
How a game decides whether a target is ahead, which way to turn and how to slide along a wall: dot products, the 2D cross product and coordinate transforms, run in Python.