The clash between Karmine Corp and Cloud9 peaked at just over one million viewers on the opening day of Worlds 2026. That’s the headline from the Play-In stage, and on its own it reads as a decline: down 59.2% from the equivalent stage a year ago. The number is accurate. It is also measuring the wrong thing.
Last year’s Play-In was a single series. T1, six-time world champions, needed one win over Invictus Gaming to reach the main event, and that match set the all-time peak-viewership record for a Worlds opener: 2.5 million. This year’s Play-In expanded to four teams playing a double-elimination bracket. Karmine Corp, Cloud9, FURIA, and MVK Esports all needed a result; none of them carried T1’s gravitational pull. Comparing this year’s ceiling to last year’s isn’t a measurement of League’s health. It’s a measurement of what happens when a single blockbuster match gets replaced by a tournament.
The Number That Actually Grew
Total Hours Watched tells a different story: roughly 14.8 million this year, about twice last year’s total. That increase is real, and it’s also arithmetic. Four teams playing a multi-game bracket generate more broadcast hours than two teams playing one series, independent of how many people are watching at any given moment. Riot gets to report growth. Advertisers who bought against peak concurrent viewers, the number that decides how many eyes actually hit a single ad read, got a number that fell by more than half. Both claims describe the same four days. Neither is lying. They’re just not measuring the same thing.
What Karmine Corp Is Actually Selling
The one audience segment that grew in absolute terms was French-language viewership, which doubled, driven entirely by Karmine Corp. Kamet0 Kebir, the club’s co-owner and a streamer in his own right, finished second by peak viewers across the entire stage, behind only Caedrel. That’s not incidental. Karmine Corp bought its LEC slot in a 2023 raise that valued the club at roughly 20 million euros; a franchise seat like that is a bet that a fanbase can be mobilized on demand, repeatedly, regardless of where the team actually finishes in the bracket. The 3:0 sweep over MVK Esports is a result. The audience that shows up because Kameto is live is the asset the club was built to monetize.
207 Channels, Not One Audience
207 unique channels broadcast the Play-In stage, up from 180 last year. That’s growth too, and it points the same direction as most of League’s recent viewership data: the audience isn’t concentrating on one official feed, it’s spreading across co-streams, and the value that would once have accrued to a single broadcast now gets split 207 ways. English still led, at just over a third of total hours watched. Everything else is a partial claim on a smaller, more specific audience.
The headline number doesn’t tell you whether the Play-In worked. It tells you how many games were played. The real question is which metric Riot and its broadcast partners are actually selling against: peak spectacle, or aggregate hours? This stage just proved the two can move in opposite directions at the same time.
