Insights on solo development, AI tools, and building digital products as an independent developer.
Why a solo developer should ship free browser games at all. A portfolio, a funnel, a reason for a stranger to click your name, and the leverage of link not download.
When you play a star out of his natural role, World 11 docks him. Here is how I turned the fuzzy sense that a winger is wasted at fullback into a rating system.
How I designed the campaign auto-battler in Orchard Deck. Eight combat keywords, six tribal synergies, a readable line of five cards, and replays that beat live control.
How I built the orchard in Orchard Deck around wall-clock growth and vitality that decays over nine harvests, turning worn-out cards into fuel instead of trash.
Scoping a solo game so you ship it. The feature complete vertical slice, cutting scope without guilt, the endless systems trap, and knowing when a game is done enough.
How team strength becomes chances and chances become goals in World 11. Building a probability model that produces believable scorelines and the occasional upset.
How a slot-machine tech tree became the progression backbone of Orchard Deck. Three reels grow to five, jackpots arrive, and the odds visibly bend your way.
How the museum in Orchard Deck turns cards you donate into permanent production multipliers, and why a one-time sacrifice feels heavier than a repeatable cost.
An honest case for Godot 4 as a solo developer, from shipping two free browser games. Free with no revenue share, GDScript, native web export, and the real trade offs.
Same-nation and same-decade bonuses make World 11 a real draft. Here is how chemistry sets up the tug of war between the obvious star and the player who fits.
Ten arcade minigames inside Orchard Deck without turning it into a grind. Entry fees, payouts, once-per-day rewards, and level gating that keeps them optional.
How I built breeding in Orchard Deck, where fusing two cards consumes both parents for a rarer child, and why the destructive cost is the point.
A balance bot runs hundreds of optimal teams through the full gauntlet and reports the champion rate. That dial replaced the playtest team a solo developer does not have.
I have no QA team, so I built a headless smoke suite in Godot that instances every scene, round trips saves, sweeps 300 battle seeds, and enumerates the slot odds.
All of World 11's balance lives in data files. Tuning the difficulty means editing JSON and rerunning, never touching the simulation. Here is why that split matters.