Korean pop culture, in numbers

Titles

Remove Ma Dong-seok and 99 more: Korean Netflix still holds

Of 666 Korean titles with a recorded cast, 649 form a single connected body through shared actors. We expected a few very busy people to be holding it together. Take out the 100 busiest and 94.5% is still one piece.

The table this is written from: /actors

#ma-dong-seok#hwang-jung-min#korean drama#korean film#netflix#wikidata#casting#korea

Join every Korean film and series that has reached a Netflix Top 10 to the actors credited in it, and a shape appears that is hard to see from any one title.

666 titles. 1,391 actors. 3,579 casting slots. Draw a line between two titles whenever they share a credited actor, and the median title is connected to 16 others. 12.12: The Day is connected to 114. Only 15 titles — 2.3% — share nobody with anything else.

Follow those lines and 649 of the 666 titles, 97.4%, form a single connected body. There are 17 separate groups in total: one containing almost everything, and 15 titles sitting alone.

We assumed a few people were holding it together. They are not.

The obvious explanation is a small repertory of very busy actors. Ma Dong-seok appears in 21 of these titles, Hwang Jung-min in 20, Lee Byung-hun in 18, Sul Kyung-gu and Ha Jung-woo in 17 each. Thirty-eight actors appear in ten or more. Pull those people out and the whole thing should fall apart.

We tested it. It does not.

Actors removed Titles left Largest connected body Median links per title
none 666 649 (97.4%) 16
10 busiest 654 636 (97.2%) 14
20 busiest 654 634 (96.9%) 12
38 busiest (everyone with 10+ titles) 645 624 (96.7%) 11
50 busiest 639 617 (96.6%) 10
100 busiest 613 579 (94.5%) 7

Remove the hundred busiest actors in Korean Netflix-charting film and television — one in fourteen of everyone in the set — and the remaining 613 titles are still 94.5% one connected body.

The median number of links does fall, from 16 to 7. Busy actors do a great deal of the connecting. They are simply not load-bearing: take them out and the structure stands on everyone else.

For comparison, removing the hundred least busy actors changes almost nothing, as you would expect: 649 of 665 titles, 97.6%.

Where the connection actually comes from

Half the actors in this set appear in exactly one title. 673 of 1,391 — 48.4%. Only 204 appear in five or more.

So the connectivity is not the work of a famous few. It is the accumulated effect of a very large number of people who did two or three of these, overlapping in different combinations. A production does not need to hire a star to be joined to the rest; it needs to hire almost anyone who has worked before.

Title Linked to Cast recorded
12.12: The Day 114 titles 28
Mr. Sunshine 109 32
Squid Game 74 16
The Thieves 69 11
Pinocchio 70 43

Note that The Thieves reaches 69 titles on eleven recorded cast members while Pinocchio reaches 70 — one on eleven recorded cast members, the other on forty-three. How connected a title is has little to do with how large its cast is.

What would change this number

A link is one shared actor. That is deliberately weak. A single character actor with a scene in two unrelated films joins them here. The median-links figure is the honest companion to the 97.5%: it says the typical title is not hanging on by one thread but sitting in a mesh of sixteen.

Wikidata’s cast lists are incomplete, and they skew towards names someone thought worth writing down. That cuts both ways: the true number of links is higher than we measure, and the share of linking done by famous people is lower than it looks here.

277 titles have no cast recorded at all — 943 Korean titles are matched to Wikidata items and only 666 carry a P161 statement. Those 276 are not counted as isolated; they are absent.

Only Korean citizens are counted. International co-productions lose their non-Korean cast, and with them the links to everything else those people have been in.

And this is cast alone. Directors, writers, producers, studios and agencies are not in it. If the question is whether Korean screen output is made by an overlapping pool of people, cast is one answer, and almost certainly the most conservative one available.


What changed on 2 September 2026

We re-collected the cast data on 25 August 2026 and this piece now reports that count. The set moved from 668 titles to 666 and from 1,395 actors to 1,391, because Wikidata’s cast statements changed underneath us in both directions. Every figure above was recounted from the new file.

The finding did not change. Removing the hundred busiest actors still leaves 94.5% of the set in one connected body, where the earlier count said 94.6%. We are printing the old and new figures side by side rather than quietly swapping them, because a number that moves without saying so is a number no one should trust.

The body holds without its busiest people. It is built from pairs — and those barely repeat. Read how rarely two actors work together twice →

What changed on 3 September 2026

We re-collected the keyed title roster from Wikidata. One more charted title matched a Korean Wikidata item, and it has no cast recorded.

Was Now
Charted titles matched to a Korean Wikidata item 942 943
Of those, titles with at least one cast member 666 666 (unchanged)
Titles with no cast recorded 276 277

The join did not change. What grew is the roster it is measured against, so the gap we name here is one title wider than it was.

Where these numbers come from

Sources

  • Netflix — Top 10 weekly lists (Tudum), global and per-country, used to decide which titles enter the set · https://www.netflix.com/tudum/top10
  • Wikidata — Country of origin (P495 = Q884) to identify Korean titles; cast member (P161) with citizenship (P27 = Q884) for the cast of each · https://query.wikidata.org

What we checked

  • Titles and people are joined on Wikidata Q-numbers, not on title text. Matching by text put cast on only 317 of 1,005 titles because article names differ from chart names; matching on Q-numbers reaches 666 of 943
  • This piece uses no pageview data at all. Nothing here depends on whether a person has an English Wikipedia article, so the selection bias that shapes our attention panels is absent
  • A link means two titles share at least one credited actor. That is a weak relation on purpose, and it is why we report the median number of links per title (16) beside the size of the connected body, rather than the body alone
  • The robustness test removes actors in order of how many titles they appear in, and is reported against the mirror-image test that removes the same number of the least busy actors. Both are shown
  • 943 of the titles Netflix has charted are matched to a Korean Wikidata item. 666 of those have at least one cast member recorded. The 277 with none are absent from this analysis and counted here

What we left out, and why

  • Any claim about the whole Korean industry. This is what reached a Netflix Top 10, which is a small and unrepresentative slice of what Korea makes
  • Directors, writers, producers, studios and agencies. The link here is cast only. The industry may be far more connected through people we did not measure
  • Foreign cast. We filter to Korean citizenship (P27 = Q884), so international co-productions lose their non-Korean cast and the links those would create
  • Any claim that Wikidata's cast lists are complete. They are not, and they skew towards well-known names — which understates how many links exist and overstates how much of the linking famous people do

The data behind this

Written from the same data

Take this away — 10 cards

The figures in this article, as images. Free to repost with the address on them.

Remove Ma Dong-seok and 99 more: Korean Netflix still holds — card 1 of 5Remove Ma Dong-seok and 99 more: Korean Netflix still holds — card 2 of 5Remove Ma Dong-seok and 99 more: Korean Netflix still holds — card 3 of 5Remove Ma Dong-seok and 99 more: Korean Netflix still holds — card 4 of 5Remove Ma Dong-seok and 99 more: Korean Netflix still holds — card 5 of 5Remove Ma Dong-seok and 99 more: Korean Netflix still holds — tall card 1 of 5Remove Ma Dong-seok and 99 more: Korean Netflix still holds — tall card 2 of 5Remove Ma Dong-seok and 99 more: Korean Netflix still holds — tall card 3 of 5Remove Ma Dong-seok and 99 more: Korean Netflix still holds — tall card 4 of 5Remove Ma Dong-seok and 99 more: Korean Netflix still holds — tall card 5 of 5

If you work in this business

Everything on this page is measured from the same weekly Netflix lists, per market and per title. We publish the pages free; what we sell is the same measurement cut to one company's catalogue — including the figures behind /actors and /titles.

What a company sheet contains, and what it cannot tell you →

← All 237 articlesFront page