Trope pairs
Some romance tropes are secretly the same story
A trope is supposed to be one ingredient among many. Counted across 65,510 books, several of them turn out to be inseparable, and a few pairs avoid each other so completely that the avoidance is itself a rule.
SmittenSpines · 65,510 books · published 4 August 2026
Romance is the most trope-literate genre there is. Readers don't browse by author or by cover, they browse by ingredient: enemies to lovers, forced proximity, only one bed, second chance. The vocabulary is precise enough that people build whole spreadsheets in it.
What that vocabulary hides is that the ingredients aren't independent of each other. Some pairs of tropes are wired together so tightly that naming one has already named the other, and knowing which pairs those are changes how you ought to search.
- 15×
- one night stand + secret baby, versus chance
- 7.4×
- arranged marriage + marriage of convenience
- 14%
- of the shelf carries found family, the most common trope of all
- 39
- tropes tracked across the corpus
274 books, about 18 expected
335 books
The pair that gives the game away
The strongest association in the whole corpus is one night stand and secret baby, which share 274 books against the roughly 18 you'd expect if the two were picked independently. That's a ratio of 15.1×.
Say those two out loud and the reason is obvious. They aren't two tropes. They're the first act and the second act of one plot, and the only reason the genre bothers with separate names is that readers come at the same story from different directions. Some of us want the night. Some of us want the consequence.
Once you notice that shape you start seeing it everywhere. Arranged marriage and marriage of convenience run at 7.4×. They're two names for getting married before you fall in love, and they differ mainly in who did the arranging.
Association strength
The tropes that come as a package
How much more often each pair shares a book than chance would produce. The dashed line is chance. Hover a dot for the underlying counts.
Show the numbers as a table
| Pair | Books with both | Expected | Ratio |
|---|---|---|---|
| one night stand + secret baby | 274 | 18 | 15.13× |
| arranged marriage + marriage of convenience | 335 | 45 | 7.4× |
| secret baby + single parent | 278 | 45 | 6.19× |
| forced proximity + grumpy / sunshine | 181 | 35 | 5.25× |
| brother's best friend + friends to lovers | 226 | 45 | 4.97× |
| boss / employee + one night stand | 162 | 35 | 4.63× |
| age gap + boss / employee | 177 | 42 | 4.18× |
| fated mates + soulmates | 342 | 84 | 4.08× |
| morally grey + revenge | 453 | 115 | 3.95× |
| age gap + forbidden love | 757 | 196 | 3.86× |
| arranged marriage + royalty | 724 | 216 | 3.35× |
| one night stand + single parent | 266 | 80 | 3.35× |
| second chance + secret baby | 219 | 65 | 3.35× |
| enemies to lovers + forced proximity | 952 | 288 | 3.31× |
| bodyguard + royalty | 214 | 66 | 3.25× |
| fake relationship + friends to lovers | 220 | 69 | 3.2× |
| brother's best friend + forbidden love | 418 | 131 | 3.2× |
| friends to lovers + roommates | 150 | 50 | 3.01× |
| fated mates + found family | 1,336 | 450 | 2.97× |
| forced proximity + opposites attract | 320 | 114 | 2.81× |
| second chance + single parent | 803 | 287 | 2.8× |
| friends to lovers + second chance | 556 | 203 | 2.73× |
| brother's best friend + second chance | 252 | 93 | 2.7× |
| boss / employee + forced proximity | 168 | 65 | 2.59× |
The most useful entry on that list is probably enemies to lovers and forced proximity at 3.3×, across 952 books. That one isn't a naming accident, since they really are different ideas. The genre has simply decided that hatred needs a locked room to work in. If you love enemies to lovers and have never thought of yourself as a forced-proximity reader, the shelf would like a word.
What the shelf is actually made of
Association is one question and sheer volume is another. The most talked-about tropes are not always the most written ones.
Frequency
Found family and slow burn are the genre's real staples
Share of books whose descriptive text carries each trope.
Show the numbers as a table
| Trope | Books | Share |
|---|---|---|
| found family | 9,323 | 14.2% |
| slow burn | 9,201 | 14.0% |
| forbidden love | 7,328 | 11.2% |
| enemies to lovers | 7,032 | 10.7% |
| royalty | 5,741 | 8.8% |
| second chance | 5,227 | 8.0% |
| coming of age | 5,134 | 7.8% |
| revenge | 3,689 | 5.6% |
| single parent | 3,594 | 5.5% |
| fated mates | 3,159 | 4.8% |
| opposites attract | 2,786 | 4.3% |
| forced proximity | 2,680 | 4.1% |
| friends to lovers | 2,548 | 3.9% |
| arranged marriage | 2,469 | 3.8% |
| morally grey | 2,037 | 3.1% |
| fake relationship | 1,770 | 2.7% |
| age gap | 1,754 | 2.7% |
| soulmates | 1,737 | 2.7% |
| love triangle | 1,611 | 2.5% |
| boss / employee | 1,583 | 2.4% |
| one night stand | 1,449 | 2.2% |
| roommates | 1,282 | 2.0% |
| insta-love | 1,235 | 1.9% |
| marriage of convenience | 1,201 | 1.8% |
| brother's best friend | 1,168 | 1.8% |
| grumpy / sunshine | 843 | 1.3% |
| hidden identity | 840 | 1.3% |
| secret baby | 819 | 1.3% |
| bodyguard | 751 | 1.1% |
| redemption | 735 | 1.1% |
Found family tops the list at 14.2%, slow burn is a hair behind, and enemies to lovers, which has by far the loudest voice online, sits at 10.7%. Grumpy/sunshine, which dominates the discourse to an almost silly degree, appears in only 1.3% of books.
That gap between how much a trope gets talked about and how often it gets written is worth keeping in mind. A recommendation engine trained on the chatter would badly over-serve the loud tropes. One trained on the shelf alone would under-serve them. Neither is what a reader wants.
The pairs that refuse each other
The inverse is just as informative and gets discussed far less. Some combinations aren't rare by accident. They're rare because they can't both be true of the same book.
Anti-associations
Combinations the genre almost never writes
Pairs that share far fewer books than chance would predict. Each of these would be expected in at least 20 books; most manage single figures.
Show the numbers as a table
| Pair | Books with both | Expected | Ratio |
|---|---|---|---|
| boss / employee + coming of age | 6 | 124 | 0.05× |
| friends to lovers + morally grey | 6 | 79 | 0.08× |
| boss / employee + fated mates | 8 | 76 | 0.1× |
| arranged marriage + roommates | 5 | 48 | 0.1× |
| grumpy / sunshine + love triangle | 2 | 21 | 0.1× |
| coming of age + marriage of convenience | 10 | 94 | 0.11× |
| morally grey + opposites attract | 11 | 87 | 0.13× |
| arranged marriage + boss / employee | 9 | 60 | 0.15× |
| coming of age + touch her and die | 4 | 26 | 0.16× |
| opposites attract + revenge | 26 | 157 | 0.17× |
Some are just the arithmetic of setting. Coming-of-age stories don't happen in offices, so boss/employee and coming of age barely meet. Others are more revealing. Grumpy/sunshine and love triangle share 2 books in the entire corpus. A grumpy/sunshine book is a two-person promise and a third person breaks it, so the genre declines to write that.
Friends to lovers with a morally grey love interest is another near-void. The appeal of friends to lovers is safety. The appeal of morally grey is that you can never quite be sure. You can want either one. You apparently cannot want both at the same time.
Why this changes how you should search
Trope search as it usually works treats your tags as independent filters: tick three boxes, get books with all three. The counts here say that model is wrong in both directions.
- Some of your tags are redundant. Asking for both halves of a fused pair narrows nothing and makes the results look thinner than the shelf really is.
- Some of your tags are in conflict. When a search comes back with three books it should say so, because “the genre almost never puts these together” is a genuinely useful answer and much better than three bad matches dressed up as good ones.
- And the tag you didn't tick might be implied. Wanting enemies to lovers usually means wanting the locked room too.
A catalogue that knows the shape of its own shelf can tell you which of those three you're in, which is a more honest thing to hand someone than a list.
Checking our work
Every figure here comes from one corpus with known limits. It is a snapshot, it leans heavily towards books published between 2020 and 2024, and it works from what gets said about books rather than from their full text. The methodology page sets out what that does and doesn't let us claim, including the measurement we ran to find out how often our own counting is wrong.
The waiting list
Get one email when the search box opens
One email, then nothing. It isn't a newsletter and there's no drip sequence waiting behind it with “5 tips” in the subject line. If that ever changes, you get to say no first.
This form isn't connected to anything yet, and pressing it stores nothing.