Game design and the core loop

More is not more fun

When Kai and Remy first vibe coded Hyper Shards, they kept asking the AI for more: more ships, more weapons, more menus. The game got bigger, but it did not get more fun.

Fun does not come from the number of features. It comes from the thing players do over and over, and how good that feels. Game designers start there, so this lesson does too.

Mechanics and the core loop

• Hyper ShardsDodge bullets, blast aliens, collect shards, upgrade the ship. Then the next wave, a little harder.
• Garden DefensePlant, defend Sprig from pests, earn coins, grow new plants. Then the next wave.

A mechanic is one action the rules let a player do: move, shoot, plant, collect. The core loop is the small set of mechanics a player repeats again and again, often every half a minute or so. Players might spend hours inside it, so it has to feel good every single time.

Each step feeds the next

In a strong loop, every step gives you something for the next one. In Hyper Shards, shards pay for upgrades, upgrades help you blast more aliens, and more aliens drop more shards. That chain is what makes players say "just one more wave".

Good loops also hold small choices. A shard that drops right next to an alien is worth grabbing, but you have to fly into danger to get it. Risk and reward choices like this keep the same loop interesting for hours.

Loops inside loops

• Short goalsSeconds: dodge this bullet, survive this wave.
• Medium goalsMinutes: beat the boss, unlock the next plant.
• Long goalsDays: finish every level, climb the leaderboard.
• RewardsA sound, shards, a new plant, a medal. Each one says "you did it" and points to the next goal.

Games have loops of different sizes, one inside another, each with its own goals.

Scope and the MVP

Scope is how big a project is: how many features, levels and characters it has. Bigger scope means more time to build, test and fix.

An MVP (minimum viable product) is the smallest version of a game that is still fun. For Hyper Shards: one ship, one alien type and waves that get faster, drawn as plain coloured shapes. Designers call this a grey box prototype. If plain shapes are fun, the loop works. If they are not, art and extra levels will not save it.

Cutting features

Every feature costs time to build, time to test, and time to fix for as long as the game exists. So studios ask one question about each idea: does it make the core loop better?

If yes, it goes on the plan. If not, it goes on a "maybe later" list. Professional studios cut features all the time, sometimes after months of work. Cutting is not giving up. It is how a small team actually finishes a game.

A one-page design

Studios write their decisions down so the whole team, and the AI, builds the same game. A short design document covers four things:

• The pitchOne sentence: what the game is and why it is fun.
• The core loopThe repeated actions, in order, and what each one feeds.
• The playerWho it is for, and on which device they play.
• MVP and maybe-laterWhat we build first, and what waits.

Bolt says: Paste the one-page design into your prompt when you ask Claude Opus 5.5 for a feature, so the AI knows what the game is for. Next lesson: level design, turning the core loop into waves and maps worth playing.