Boris Chen [new] →

The result was not a list, but a A series of horizontal blocks, arranged from top to bottom. If two players are in the same tier, flip a coin. If they are in different tiers, start the better one without hesitation.

But the raw data was ugly. Chen, who moonlights as a design enthusiast (he has cited Piet Mondrian’s grid-based abstract art as an influence), decided to publish the results on a simple GitHub page. He used a clean, color-coded CSS grid. Red for Tier 1. Orange for Tier 2. Yellow for Tier 3. boris chen

That insight was revolutionary. Chen realized that traditional numerical rankings create a false sense of precision. The gap between the 5th-ranked wide receiver and the 8th-ranked one might be tiny (a bad route, a dropped pass), while the gap between the 8th and the 15th might be enormous. The result was not a list, but a

So, he borrowed a technique from machine learning: The Algorithm Behind the Art Chen began scraping consensus rankings from the industry’s most accurate experts (sources like FantasyPros, ESPN, and Rotoworld). Instead of averaging their numbers, he applied a clustering algorithm—similar to how Netflix groups similar movies or how biologists classify species—to group players who were statistically indistinguishable from one another. But the raw data was ugly

In the sprawling, chaotic ecosystem of fantasy football, information is currency. Every Sunday, millions of managers drown in a tsunami of stats: targets, air yards, rushing attempts, defensive matchups, and weather forecasts. The difference between a championship trophy and a last-place punishment often comes down to one question: Who do I start?