Korean pop culture, in numbers

titles

We wrote that a market has one Korean slot. Measured, two-thirds of that was arithmetic

When a Korean title reaches a market's top three, the other Korean titles there fall by 0.395. Non-Korean arrivals cost 0.014. That looks decisive until you notice the two groups start from different heights.

On 9 August we published a sentence on our hard-markets page explaining why some countries only ever show the biggest Korean titles:

there is one Korean slot and the biggest title takes it

It was an explanation, not a measurement. We had not tested it. This is the test.

What a Korean arrival looks like

Take every week where a Korean title entered a market’s top three without being on that chart the week before — 2,258 of them. Count the other Korean titles on that chart, four weeks before against four weeks from the arrival. Then do exactly the same for arrivals that are not Korean.

Arrivals at the top 3 Cases Other Korean titles before Change after
Korean title arrives 2,258 1.916 -0.395
Any other title arrives 36,654 0.563 -0.014

A twenty-eight-fold difference. Published on its own it would have looked like proof.

Why it is not

Read the third column. The Korean cases start at 1.916 other Korean titles; the control starts at 0.563. A Korean title reaches a market’s top three most often in markets that already carry several Korean titles — and a chart that starts high falls whatever arrives, because there is more to lose.

So we grouped both sides by how many other Korean titles were there to begin with, and compared like with like:

Other Korean titles before Korean arrivals Change Other arrivals Change Difference
0 643 +0.130 21,459 +0.157 -0.027
1 527 -0.237 10,634 -0.188 -0.049
2 347 -0.519 2,808 -0.296 -0.222
3 357 -0.617 1,294 -0.503 -0.114
4 or more 384 -1.173 459 -0.875 -0.298

The control falls almost as far in every band. Weighted to the same starting levels, a non-Korean arrival would have cost -0.273; the Korean arrivals cost -0.395. What is left is -0.122, and 68% of the raw gap was the starting level rather than the arrival.

What survives

Something does. A Korean title reaching the top three of a market costs that market about a tenth of another Korean title, over and above what any arrival costs. In the markets that already carry four or more, it costs about three-tenths.

And the top row of the table is the honest limit on the original sentence. Where there was no other Korean title, a Korean arrival brings slightly fewer than a non-Korean one does — +0.130 against +0.157. In the markets we described as having “one Korean slot”, the slot does not behave like a slot at all. Whatever is happening there, it is not one title displacing another.

So the sentence was not wrong. It was much smaller than it sounded, and it does not hold in the markets we wrote it about.

What this cannot tell you

It shows places, not reasons. A Korean title that disappears the week a bigger one arrives may be ending its own run on schedule; nothing in a rank table separates that from displacement, and a four-week window cannot either.

It is also not viewing. A title pushed to eleventh place is invisible here at any level of watching, so “lost its place” is a statement about a chart, not about an audience.

Where these numbers come from

Sources

  • Netflix — Tudum weekly Top 10 country lists, 492,900 rows read, 2021-07-04 to 2026-07-26, 93 markets, Films and TV counted as separate charts, Russia excluded · https://www.netflix.com/tudum/top10

What we checked

  • Non-Korean arrivals at the same top-three positions are used as a control, so the comparison is against what any arrival does rather than against nothing
  • The control is weighted to the same distribution of starting levels as the Korean cases, because a chart that starts high falls whatever arrives
  • The arriving title is removed from the count on both sides, so the fall cannot be the arriving title itself
  • The band with no other Korean title present is reported separately, because it is where the claim fails

What we left out, and why

  • Russia, excluded across this publication because Netflix withdrew and the remaining weeks are not comparable
  • Any reading of why a title left. A run ending on its own and a run displaced look identical in a rank table

The data behind this

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 /crowding.

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

← All 74 articlesFront page