Lee Jung-jae Drew 10% of Squid Game's Wikipedia Peak. Hyun Bin Drew 34% of a Smaller One

Twelve Korean titles, a mix of films and shows whose English Wikipedia pages are known to have spiked at some point, give a clean pattern when you ask one question of each: on the single day each title’s own page peaked, how much of that peak’s traffic went to its best-known cast member, measured the same way?

The bigger the peak, the smaller the actor’s share of it

Sorted by the title’s own peak, the pattern runs almost straight down: Parasite’s 2.7 million-view peak day sent only 3.9% of that traffic to its most-looked-up actor, Song Kang-ho. Crash Landing on You’s much smaller 131,083-view peak sent more than a third of it — 34.2% — to Hyun Bin. Across all 12, the correlation between a title’s own peak size and its lead actor’s share is -0.68 (t = -2.92 on 10 degrees of freedom).

The lead actor was not our pick

We did not choose who counts as each title’s “lead” before measuring. We pulled every credited cast member’s page and read all of them on the title’s own peak day, then used whichever one actually drew the most views. Guessing in advance would have been wrong twice here — Hellbound’s biggest draw is Kim Hyun-joo, not the ensemble’s top-billed name, and the same is true for The Silent Sea, where Gong Yoo leads.

What could be doing this instead of what it looks like

This sits beside our running fictional-vs-real count, which asks a related question — whether a title’s own page draws more English-language attention than the people in it. Two things push against reading this new pattern as “megahits are about the story, not the star.” First, the biggest peaks in this set — Parasite, Squid Game — are also the titles with the most credited cast members we could measure (9 and 14), which mechanically spreads the same pool of attention across more pages and shrinks any one person’s slice. Second, cast coverage is uneven: Crash Landing on You is missing 10 of a possible 25 cast pages from this count, and a title missing rows could show a share that is higher or lower than its true one for that reason alone, not because of anything about audience attention. We’re publishing the correlation with both of those caveats attached, not as a settled account of why it holds.

Every value here is same-day: the actor’s page is read on the exact date the title itself peaked. Actor pages commonly peak a day or two after a title’s own peak, once a cast is being searched by name rather than the title’s headline. That makes every share in this table a floor on how much of a title’s attention eventually reaches its lead — not the final word on it.