What the Numbers Actually Belong To
The most-watched de_cache match in recorded history happened in January 2017, at the ELEAGUE Major in Atlanta. Astralis and Fnatic met in the semifinals. At overtime, 785,394 concurrent viewers were watching. That figure has not been surpassed. It also has nothing to do with CS2.
The next three entries on the all-time list come from the same two-year window: PGL Major Kraków 2017 (708,376), ELEAGUE Major 2018 (673,856), the same Atlanta event (550,012). Every top-five peak belongs to the 2017-2018 Major cycle, when CS:GO’s viewership was still accelerating. De_cache was removed from Active Duty in March 2019. The competitive era that made those numbers closed with it.
Nostalgia carries viewership value. Nostalgia for a specific map is not the same thing as nostalgia for the conditions that built those numbers in the first place.
What Valve Is Actually Acquiring
Valve purchased de_cache from FMPONE on the day of its CS2 Workshop release in May 2025. The official CS2 version arrived April 28, 2026: eleven months of internal development for a map the community had already finished.
Valve did not need eleven months to make cache playable. A working version existed. What they spent eleven months on was control: new Source 2 assets, a scheduled release window, and the ability to drop the map as an official Valve event rather than a Workshop port with a footnote.
The first thing they purchased was the intellectual property. The second was the authority to decide when and how the map returned. FMPONE’s Workshop remake was available. Valve’s version arrives with a teaser image that spread for weeks before the update landed and a tweet from the official CS account. That is not a content update. It is a product launch wearing a patch note.
Whether This Moves the Audience
CS2’s competitive scene runs on different structural logic than 2017. Roster volatility is high. The rivalries that gave de_cache its viewership weight — Astralis against the field across multiple Majors, Virtus.Pro as the map’s recurring protagonist — have no clear equivalents now. Team identities shift faster than audiences can anchor to them, and audiences follow teams more reliably than they follow maps.
The figure that deserves more attention is not the 785,000 peak from Atlanta. It is the question of what configuration of teams, rivalries, and stakes produced that number. A familiar map in an unfamiliar competitive environment may generate short-term traffic. The 2017 numbers are not a forecast. They are a record of what was possible under conditions that no longer exist.
Whether the current game can give de_cache a reason to approach them is the open question. Map pool additions do not answer it.
Lex Northbyte 
Elias Rowe