The Tropes These Apps Make Most Are Not the Ones People Watch Most
Contract marriage is 20% of everything these platforms post and sits near the bottom on views. Mafia is 7% and sits at the top. We classified 2,221 videos to find out.
What we did
We took the official YouTube channels of ten short-drama platforms, kept the 2,221 full-length uploads (fragments excluded by measured duration), and classified each one by title keyword against twelve recurring themes. Then we put production volume and median views side by side.
They do not line up.
| Theme | Videos | Share | Median views |
|---|---|---|---|
| Mafia / power | 161 | 7.2% | 14,780 |
| Campus / youth | 69 | 3.1% | 12,725 |
| Military / agent | 37 | 1.7% | 10,280 |
| Forbidden relationship | 164 | 7.4% | 10,224 |
| Billionaire / dynasty | 363 | 16.3% | 7,400 |
| Secret baby | 225 | 10.1% | 6,906 |
| Rags to riches | 93 | 4.2% | 6,860 |
| Contract marriage | 444 | 20.0% | 6,671 |
| Secret identity | 222 | 10.0% | 6,547 |
| Genius / prodigy | 39 | 1.8% | 5,935 |
| Werewolf / supernatural | 159 | 7.2% | 5,466 |
| Revenge / rebirth | 254 | 11.4% | 5,273 |
The mismatch
The two biggest production lines sit in the bottom half on views. Contract marriage is one video in five and lands ninth of twelve. Billionaire stories are one in six and land fifth.
The three smallest sit at or near the top. Mafia, campus and military are 7.2%, 3.1% and 1.7% of output respectively, and they take first, second and third on median views.
Mafia gets 2.2 times the median views of contract marriage while being made a third as often.
Before you read too much into that
Four limits, all of which matter.
These percentages do not add to 100. They add to 100.4%, because a video can carry more than one theme — nine of the 2,221 do. Never sum them into a total. The denominator is 2,221 content videos, not the 2,230 in the video column and not the 2,974-video full corpus.
The classification is a title-keyword regex, not human judgment. It has known contamination: don matches inside don't, and captain catches a hockey captain. Nothing has been hand-checked. Read every count as approximate.
Two samples are too small to carry a conclusion. Military has 37 videos and genius 39. A dozen unusual uploads would move either one substantially. We would not build an argument on those rows.
Channel windows differ. Each channel’s sample spans anywhere from 41 to 917 days, so views are not strictly comparable across channels, and the theme medians inherit that. This is the caveat that most limits the mismatch finding, and we would rather state it than bury it.
We also revised this table once already. An earlier version had campus at 20,823 and military at 17,311 median views; correcting how we separated full-length content from fragments brought them to 12,725 and 10,280. The mismatch survived the correction, but it got smaller — so do not let anyone, including us, describe it as dramatic.
What might explain it
We can offer readings. We cannot test them with this data, and it would be dishonest to present any of them as a finding.
A promotional artefact. Platforms may push the themes they consider safest to their broadest audience, and YouTube’s algorithm rewards other things. Views here measure marketing reach, not in-app viewing.
A saturation effect. With 444 contract-marriage videos in the sample, any single one competes hardest against the others in its own category.
A scarcity effect. With 37 military uploads, whoever wants that story has almost nowhere else to go.
Which of these is doing the work — or whether it is something else entirely — is not something 2,221 titles can tell us.
Where the app catalogues disagree with the channels
Our other dataset is the platforms’ own catalogues: 1,197 series from NetShort, ReelShort and ShortMax, with the labels each platform assigns.
Werewolf makes a good comparison. It is 7.2% of YouTube uploads (159 of 2,221) but only 17 of the 1,197 series in our catalogue carry a platform Werewolf tag — 1.4%. A fivefold gap.
There are two honest explanations and we cannot pick between them:
- Platforms push werewolf content harder on YouTube than its catalogue share justifies.
- Our catalogue covers three platforms, and the ones filing under Werewolf may simply not be among them.
The second is more likely than it sounds, because of something worth knowing: all 17 sit on NetShort. ShortMax’s 463 series include zero — not because it has no werewolf stories, but because its 26 categories are genres and none of them is a trope label. Its filing system has no shelf for the concept.
That is the general lesson. When you compare “how much of X exists” across platforms, you are usually measuring who labels things, not what got made.
What this is useful for
If you are picking something to watch: the small categories at the top of the table are where the format’s best-performing work sits, at least by the one measure available to us. Mafia, campus and military are underproduced relative to how they do.
If you are trying to understand the format: the two dominant premises — a marriage of convenience, and a very rich family — account for over a third of everything posted. That is the centre of gravity, whether or not it is what performs.
If you write about this category: the numbers above are reproducible. The corpus, the classifier and the theme definitions are all fixed in scripts, so anyone can recount them and get the same answer. That is more than we can say for most trope rankings on this subject, including the ones that look more confident than this one.
Themes classified from the titles of 2,221 full-length uploads on the official YouTube channels of ten short-drama platforms, corpus collected 18 August 2026 and re-checked 19 August. View figures are YouTube promotional performance and do not describe in-app viewing. Catalogue comparisons use our own 1,197-series dataset read from platform pages on 19 August 2026.
Use the tools