How many different Korean titles has your country actually seen?

Two groups, 93 markets

Every figure is a median of the markets in that group, not a total. In the median Asian market it takes 54 different Korean titles to account for half of all Korean chart appearances; everywhere else, 17. The single biggest Korean title holds 2.8% of the Korean places in the median Asian market against 14.6% elsewhere.

The two ends

The difference is not which title leads. The same title leads 2 of the five widest markets as well as all five of the narrowest — it leads Indonesia with 2.4% and Latvia with 46.1%. Those two numbers describe completely different viewing experiences of the same show.

The objection, tested

A market with fewer Korean places will look more concentrated for arithmetic reasons alone. A country with only 84 Korean chart appearances cannot spread them over 247 titles. On that reading this page says nothing except "some markets chart less Korean content", which we already published. So we tested it, by comparing only markets that hold a similar number of Korean places.

The Asian markets in that band are Hong Kong, Japan, Philippines, Singapore. The others are Bangladesh, Bolivia, Maldives, Nigeria, Peru, Qatar, Saudi Arabia. The raw gap in titles-to-fill-half was 37 (54 against 17); inside the matched band it is 25 (54 against 29). 32.4% of the gap was volume, and the rest is not. That is the honest size of it: smaller than the first table suggests, and not zero.

The width inside the Asian group

A group of ten is not ten identical markets. The number of titles needed to fill half runs from 33 to 117 inside it.

As many different Korean titles have charted there as in its neighbours, but far fewer of them carry half the places. We are not going to explain that here — a large domestic catalogue and a separate history with Korean television are both plausible and this data distinguishes none of it. We name it because a group with one member behaving differently is worth saying out loud rather than leaving inside a median. The full story →

Every market, in alphabetical order

Sorted by name, not by size. Look up your own; the order carries no meaning. Every market name below is a link — 93 of them open a page with that market's own titles, up to 20 of them, week by week. A market needs at least 8 charting titles to get a page; we do not publish a page thin enough to say nothing. Medians across the 93 markets, for comparison on those pages: 4.5% of chart places · 97 different titles · 19 titles filling half.

Shaded rows are the 10 Asian markets grouped above. Every figure is measured over the same 265 weeks with the same identification, so a market low on this table charted fewer Korean titles — it does not mean fewer were available or fewer were watched.

How sure are we that these titles are Korean?

Every title above is matched against works Wikidata records as made in South Korea — by Wikidata item number where we hold one, and by English name otherwise. We asked, for all 938 titles in this table, which countries made a film or series of that name. Four answers came back.

By viewing, this catalogue is far more certain than it is by title count. 82.7% of Korean viewing hours sit on titles only Korean works carry — but only 51.4% of the titles do. The two are not in conflict: the largest titles are the unmistakable ones, and the uncertainty lives in the tail, where a single common English word can name a Korean series and a foreign one at once. A figure weighted by places or hours inherits the first number; a count of distinct titles — the “Different titles” column above — inherits the second.

A name is not an identifier, and this check learned that the hard way. Asked by name alone, Wikidata reported no Korean work behind Land, Deliver Us from Evil or Detour — because it files them as LAND, Deliver Us From Evil and DETOUR, and a label match is case-sensitive. Only the foreign works came back, which reads exactly like “this is not Korean.” We now ask by identifier instead, using the Wikidata item numbers already stored alongside these titles. 40 titles carrying 1,413 of the 37,750 places on this page are found only that way.

The last row of the first table is what is left over: foreign works carry the name and no Korean item is tied to it. That is not a finding that the title is foreign — it is the part we have not settled. It covers 87 of the 37,750 Korean places here, or 0.23%. We name them rather than quietly drop or keep them:

What this check still cannot do: a shared name stays shared. We can say how large that group is, which is what the first table is for, but not which of the two works charted in Manila on a given week. And a title Wikidata has never heard of comes back with no country, which is not evidence in either direction.