You only get to write after you've played
Every article has tens of hours of play behind it. We write reviews, guides, industry takes — never the kind of "encyclopedia" filler.
Why I'm still playing 2048 a decade later
No story, no characters, no multiplayer — but it still costs me two subway stops. A note on emergent difficulty.
Three counterintuitive tricks to break 200 in Snake
Spiraling isn't optimal. Corner-yielding is. My most-shared guide.
The quiet H5 renaissance: why we picked the browser back up
App Store became an ad slot. Browser games are coming back. A trend take.
Four deduction patterns to clear Expert Minesweeper
1-2-1, 1-1 edges, corner-first — the four patterns intermediate players need as muscle memory.
Opening traps in 9x9 gomoku: why AI always plays center first
9x9 isn't a smaller 15x15 — it's a different game. Opening guide from 500 matches against my own AI.
Played memory match for a week. Did my working memory actually improve?
n-back transfer is limited. Card-match transfer is even more questionable. An anti-hype review with self-test data.
Why I used raw Canvas instead of Phaser: one engineer's stubbornness
Engines save time. But the real bottleneck in H5 games is first-paint and memory. My selection notes.
We deliberately picked a midpoint between casual and hardcore
Too light = match-3 churn. Too heavy = challenge fatigue. The difficulty curve principles for our 24 games.
What Euler left us: the topology behind one-stroke puzzles
How Königsberg's 1736 seven bridges turned into modern puzzle games. Solvability needs only odd vertex counting.
The optimal solve for Flip Tiles: cracking Lights Out with linear algebra
Flip Tiles looks random but is actually GF(2) linear equations. Light-chasing solves 5x5 near-optimally.
A month of Word Drift: my vocabulary actually grew
Anagram games show real near-transfer. A self-test with data on why this transfer works.
Why "tidying" games like Drop Sort are so hard to stop
From Zeigarnik effect to the pleasure of order, why these small games hook us.
Three maze algorithms compared: why I picked recursive backtracking
Prim, Kruskal, recursive backtracking — three mainstream generators produce visibly different mazes.
Threes is better than 2048. 2048 went viral worldwide.
A 14-month-polished Threes! lost to an open-source clone. The most famous indie "beaten by a knockoff" case.
The "exponential scoring" trap in match-clear games
Why does Bubble Pop give 1 point for small groups and 90 for big? Exponential scoring creates not challenge, but a "greed trap."
Sudoku vs. Minesweeper: information entropy in two "logic games"
Why does sudoku always solve through logic while minesweeper often forces a guess? The difference is in entropy.
Sokoban is NP-hard: why AI can't solve your puzzle
Sokoban is formally proven PSPACE-complete. No known algorithm solves any instance in reasonable time. Popular science + dev note.
Sam Loyd's $1000 hoax: an unsolvable puzzle
In 1880 Sam Loyd offered $1000 for a mathematically unsolvable 15-puzzle variant. A famous marketing hoax in game history.
Connect Four was solved in 1988 — but humans still don't know
Allen and Allis proved Connect Four is a first-player win in 1988. 35 years later, most still play by intuition.
What does a reaction test actually measure? Data vs. intuition
True "reaction speed" is composed of 3 independent components; games test only the most optimizable one.
How Pong was born: Allan Alcorn's "practice assignment" that started a $70B industry
In 1972, Atari's founder told a new engineer to "practice" by making a simple game — and accidentally birthed commercial video gaming.
The day before Flappy Bird shut down: what was Dong Nguyen thinking?
Feb 2014, Flappy Bird earned $50k/day at peak when its creator voluntarily took it down. Reveals indie dev vs. "addiction economics" conflict.
60fps isn't free: 5 pitfalls I hit writing H5 game loops
setInterval, RAF, performance bottlenecks, mobile frame rate variance — battle-tested dev notes.
24 games in a year: Max and Leo's honest retrospective
BverGame series finale. Which games hit, which nobody played, what data says, what we got right and wrong.