Game Note #1: Serendipity

Returning to Grand Theft Auto: San Andreas after nearly two decades has been one of the most inspiring design experiences I’ve had this year. As a Y2K kid, I loved messing around with the obvious things: lowriders, gun mods, hidden tags and stunt jumps, tattoos, ridiculous cheats, and simply causing absolute chaos. At the time, I had no idea how deliberately Rockstar had designed the game. Replaying it now as a consumer product designer, things feel different.. This game could be one of the greatest discovery systems I have ever experienced. Instead of force-feeding players endless details and disconnected content, Rockstar creates the conditions for serendipity. The game taught me that the most valuable discovery systems not only surface unfamiliar content, but they also create the context that makes unfamiliar content meaningful to the player.

Waiting Creates Wandering

The delayed timing of certain missions is a simple example: OG Loc is one of Carl’s Grove Street friends and a wannabe rapper. After getting out of jail, he takes a job at a fast-food restaurant to support his music career. When the player goes to begin one of his missions, it is only available between 10 PM and 6 AM. Rockstar could have simply skipped forward to the correct time. Instead, nothing happens. The player has to live in the world for a few hours.

I believe this friction was intentional. While waiting, I found myself cruising around Los Santos. I stole cars I had never driven before. I sprayed graffiti. I wandered into neighborhoods I had never visited. I listened to radio stations that introduced me to American music long before Spotify existed. Eventually, I forgot that I was waiting for the mission. The cool thing is that the wandering did not feel like wasted time because discovery was constantly rewarded. Tags had counts. Cars handled differently. Exploration changed Carl’s appearance, abilities, possessions, and knowledge of the map; the serendipity felt productive.

The game was also quietly introducing me, a Taiwanese kid growing up in Taipei, to American music, cars, slang, cities, and subcultures. The 1992 Ford Thunderbird became one of my favorite cars after I discovered its in-game equivalent, the Fortune. Jane’s Addiction, Guns N’ Roses, Lynyrd Skynyrd, N.W.A., and 2Pac became part of my musical vocabulary. Yeah, the radio stations were my absolute favorite.

Modern digital products are obsessed with removing every second of waiting. Rockstar deliberately created waiting. Waiting created wandering. Wandering created discovery. Discovery created unforgettable memories.

The Thrift-Store Model

Francisco, one of my favorite product managers from Pluto TV, once described media browsing as being more like walking through a thrift store than searching a database. People do not always open a media product with a precise title in mind. Sometimes they are looking for the pleasure of stumbling across something they did not know they wanted; they want to find something strange, special, or overlooked. An old movie they can proudly claim as a favorite. Mine is Blade Runner, assuming that is still cool in 2026.

This suggests two different models of discovery. The first is directed discovery. It helps users find relevant content quickly. It optimizes task completion, reduces uncertainty, and gives people a predictable route toward something they are likely to enjoy. The second is exploratory discovery. It gives users room to wander. It creates encounters that feel accidental, even when the system has carefully designed the conditions behind them.

A person may care more about a movie, song, or object because they feel they discovered it themselves: in the corner of a video store, while driving through Los Santos, or buried on a discount shelf at Ross. These encounters create surprises. They can feel more personal than another “You May Also Like” row because the user participated in the discovery.

Algorithms optimize probability. Serendipity optimizes stories. Nobody excitedly tells a friend, “Netflix correctly predicted that I would like this.” They say, “I randomly found this strange old DVD from a flea market.” The path becomes part of the content and the story. Over time, those discoveries become part of a person’s identity.

Directed discovery	Exploratory discovery
Helps users find something relevant	Helps users encounter something unexpected
Optimizes speed and confidence	Optimizes curiosity and surprise
Predicts what the user may like	Expands what the user knows exists
Produces successful selections	Produces memorable discovery stories

Designing the Detour

The goal is not to replace directed discovery with exploratory discovery. Most people still need search, recommendations, rankings, and clear navigation. The design opportunity here is to preserve some room around those systems: spaces where users can browse without a precise goal, follow an unexpected thread, or encounter something outside the boundary of what an algorithm already knows about them.

Replaying GTA: San Andreas reminded me that media products are much more than delivery systems. They can help people build stories around how they discovered something. In an era when every recommendation algorithm is competing to predict what we will like next, the greatest competitive advantage may not be making discovery faster. It may be making discovery feel like it belonged to us all along.