City, FIFA & Moneyball: Interpreting Places

The world is vast, and there are more places to go than we could ever know.

Travelling through the South of France this summer brought that fact home to me once again. Villages unfamiliar even to the French people I spoke with, and astonishing landscapes, were hidden throughout the region. Yet search results were dominated by the same major cities and tourist attractions. AI was good at organising places that were already well documented, but it did not venture far enough into those with only a faint digital trace.

In the end, I had to go looking for them myself. I moved between Google Maps and a tangle of search tools, searching for places that matched what I was after and stitching them into a route. Some places only began to reveal themselves after I had gone several layers deep. A few of them became the most memorable parts of the journey.

Does finding a remarkable place always have to depend on individual persistence and blind luck? Could the process of discovering somewhere that truly suits a person become more systematic and precise?

Looking back, discovery has been a recurring theme throughout my career. At Netflix, the heart of my work was helping audiences discover titles. Our team specialised in identifying what was distinctive about each title and translating those qualities into visual codes and expressions tailored to the different tastes of individual members. The same title could offer different entry points depending on which subject, character or mood was brought forward.

When I started Kosmosis, the subject of discovery shifted to Korea. Rather than explaining Korea once again as a global K-phenomenon, I wanted to explore Korea as it is actually lived, through the scenes and sensibilities that often pass unnoticed in everyday life. The focus gradually moved from titles to Korea, and then towards the moods, scenes and places that make up Korea, each time becoming smaller and more specific.

Around the time I began wondering whether that same sensibility could be applied to places themselves, another thought became difficult to ignore. AI is rapidly changing the way we live, yet the way we search for places has barely changed since the arrival of the internet.

The way we book travel has barely changed. (Source: Flying Angels)

Travel still begins by choosing a destination and a date. We decide on Jeju, Tokyo or Paris, then search for flights and hotels. Everyday life follows much the same pattern. We choose a neighbourhood in which to meet, then look for restaurants and cafés nearby.

Everything begins with a destination.

But our desire for a place does not always begin with its name. Sometimes we want to feel as far removed from our familiar surroundings as possible. At other times, we want to find something familiar in an unfamiliar city. There are days when we want to disappear into a crowd, and others when we want to walk alone for hours.

Yet we still begin our search by saying, “I want to go here.”
The sentence in our minds is often closer to, “I want to go somewhere like this.”
What if discovery could begin with that sentence instead?

This brought me back to the idea I explored in my previous essay, The Language of Places. However precisely we understand a person’s tastes and needs, it is difficult to connect them with a place if places themselves are described only as “historic,” “hip” or “local.”

Places, too, need to be legible at a higher resolution.

Different places need to be understood and compared through a shared set of criteria, allowing the qualities of each place to form a distinct profile. The “language of places” is ultimately the common framework that makes this possible.

I began calling this experiment DiscoveryOS. It is less an app for recommending destinations than a framework for interpreting cities and neighbourhoods through the same criteria, then connecting their profiles with the kinds of experiences different people are seeking.

Two unlikely references helped me imagine how such a system might work.

The first came from a football game. In EA SPORTS FC, still better known to many as FIFA, every player is represented through dozens of attributes. Some describe fundamental abilities such as pace, passing, shooting, defending and physicality. Others capture less immediately visible qualities, including vision, positioning, reactions and composure.

Players with similar overall ratings can feel entirely different in a match. One becomes dangerous whenever space opens for a long-range shot. Another constantly moves without the ball, creating space where none appeared to exist. Another remains composed at a decisive moment and attempts the boldest option.

Whenever the ratings are updated for a new season, real footballers pay close attention to both their overall score and their individual attributes. The numbers represent more than the performance of a game character. They show how the football world has interpreted them as players.

A single overall score cannot explain a player. Their distinctive profile emerges from the way multiple attributes come together.

Perhaps it is not the rating he dislikes, but the photo. (Source: FIFPlay)

The second reference was Moneyball, a book and film that has stayed with me for years. If FIFA showed how multiple attributes could form a profile, Moneyball showed why choosing the right attributes matters.

Baseball was already a sport rich in statistics. What Billy Beane and the Oakland Athletics changed was not the amount of data available, but the criteria used to interpret it. By evaluating players through measures such as on-base percentage, they reconsidered those overlooked by conventional scouting and discovered value the market had failed to recognise.

The important thing was not more data. It was a different set of criteria capable of revealing hidden value.

A scene from *Moneyball*, in which players overlooked by conventional criteria are re-evaluated through a different set of stats. (*Moneyball*, 2011)

Cities already have an abundance of statistics. Global competitiveness and liveability studies compare transport, safety, economic scale, housing conditions and cultural infrastructure. These are all valuable measures, but they mostly describe how effectively a city functions.

They tell us little about the kind of day we might spend there, how it might feel to move through it, or who might find it most compelling.

The problem was not that places lacked statistics. It was that we lacked the statistics needed for discovery.

The contrast I experienced between the new and old centres of Mokpo came immediately to mind. Pyeonghwa Square sits at the heart of the city’s newer district. Along the waterfront, it resembles a smaller version of Haeundae in Busan. A few blocks inland, it could be a new town almost anywhere in Gyeonggi Province. Wide, straight roads are lined with Japanese-sign taverns advertising inexpensive beer and the familiar franchises found across the country.

Measured through convenience and activity, the new district would probably score higher. Yet the old centre around Mokpo Station, an area often referred to as Mokwon-dong, still contains old buildings, narrow lanes and the accumulated time of a port city, even as people and commerce continue to leave it.

Where people gathered, Mokpo was difficult to feel. Where Mokpo felt most itself, people were disappearing.

This is not an argument that one side is better than the other. It is an example of how completely a city can change depending on the criteria through which we read it.

I began with Seoul, the city I know best, placing its neighbourhoods side by side to understand what distinguished them.

Myeongdong and Dongdaemun both carry long histories and strong identities. But in Myeongdong, little everyday life remains beyond tourism and consumption. In Dongdaemun, commerce and production, labour and migrant life continue to overlap in dense layers.

Seongsu and Dosan Park both attract new brands and spaces. But in Seongsu, production, experimentation and consumption coexist. Dosan Park is closer to the consumption of tastes that have already been carefully refined.

Neighbourhoods that appear similar through one measure can reveal entirely different profiles through another.

At first, I assumed that a sufficiently detailed profile of a place would be enough to match it with the right person. But the same alley can feel entirely different when walked alone or with someone else, on a first visit or after many returns. The place we want today may not be the place we need on our next journey.

We are not always the same visitor.

A place becomes an experience only when its existing character meets the condition of the person moving through it.

That was when the basic structure of DiscoveryOS began to emerge. One side is the character a place already possesses, which I call Place Character. The other is the experience created as a person moves through it, which I call Visitor Experience.

The starting point of DiscoveryOS is the intersection between a place’s distinctive profile and the experience a person wants at a particular moment.

The next two essays will explore the twelve elements that make up these two dimensions. The first will look at six elements that shape the character of a place. The second will examine another six that determine how the same place can become a different experience for each visitor.

DiscoveryOS may eventually evolve into something quite different, perhaps as more people who share this premise become involved. For now, its ambition is simple: to draw a concrete blueprint for what becomes possible when the discovery of a place begins with an individual’s tastes and needs.