Build your second characterHelp shape fit-based game discovery.
Join waitlist
The Codex/Recommendation

Your Favorite Games Are Better Filters Than Your Whole Library

When you own or wishlist too many games, the fastest way to choose is not browsing everything again. It is reading the pattern inside the few games you always reopen, forgive, or recommend.


A giant library creates a strange kind of amnesia.

You can point to hundreds of games you own, wishlist, or still mean to try, but when it is time to pick one for tonight, all that abundance stops helping.

A lot of players respond by searching wider.

That is usually the wrong move.

01Your favorites are not nostalgia. They are evidence.

Most players already have a pattern, even if they do not describe it cleanly.

There are games you reopen when you want relief. There are games you forgive for their flaws because the payoff feels worth it. There are games you recommend instantly because they scratch a very specific itch you trust.

Those are not random attachments.

They are evidence about what kind of friction, rhythm, reward, and mood you reliably respond to.

The strongest recommendation signal may be the shape shared by the games you keep coming back to, not the size of the pile you have not touched yet.

02The full library is too noisy to think with

A giant library mixes too many reasons for ownership:

  • discount impulses
  • old ambitions
  • curious one-day picks
  • social obligation buys
  • comfort staples
  • games you respect more than you actually want to play

That mix is normal.

It also means the whole shelf is a bad filter when you are tired and trying to decide fast.

The fact that you own a game does not mean it belongs in tonight's shortlist.

03Favorites reveal the kind of fit you actually trust

Look at three or four games you repeatedly finish, replay, or happily recommend.

What do they have in common?

Maybe they start cleanly. Maybe they create momentum fast. Maybe they let you experiment without punishing every mistake. Maybe they feel intense without becoming exhausting. Maybe they give you closure in one sitting.

Those shared traits are often more useful than broad genre labels.

Two strategy games can demand completely different energy. Two cozy games can offer completely different kinds of comfort. Two RPGs can ask for wildly different tolerance for setup and drift.

04A better recommendation starts from repeats, not prestige

A lot of bad picks happen because players choose from admiration instead of pattern.

They open the prestigious game, the acclaimed game, the important game, or the game they feel they should finally understand.

Then the session drags.

Not because the game is weak.

Because it did not match the kind of experience the player reliably says yes to when nobody is watching.

That is where a better recommendation layer should help.

It should be able to say:

  • you usually like systems that become readable quickly
  • you forgive repetition when the build expression is strong
  • you come back to games that reward one more run energy
  • you bounce when the first hour feels like admin

Those are reasons a player can inspect and argue with.

05What Snowbll should preserve

Snowbll should not pretend to judge whether a game is objectively great.

Its useful job is narrower.

Read the pattern in your favorites. Compare that pattern against the larger shelf. Show a smaller set of matches. Explain the reason.

That boundary is what keeps the recommendation honest.

The persona can notice your repeats and point at the likely fit. The gamer still decides whether tonight is the night for that game.

06One practical question before you browse

Before you scroll the whole library again, ask this:

Which game(s) do I almost always trust when I want a good session, and what do they reliably give me?

That answer does not solve every choice.

But it usually gives you a better filter than another hour of browsing games you only liked in theory.

Snowbll is building a game discovery layer focused on taste, persona, and fit. You describe what you want; we return a few close matches, not a long list.

Phase 0 - the search side only. The catalogue is unverified and the AI parses your intent; it does not judge whether a game is good. AI recommends. Humans decide.