Last on Goodreads, first on AO3 for its size: what fan fiction saw in Heated Rivalry

The 30-second read

Two hockey romances were announced for television eight months apart: Off Campus at Prime Video in October 2024 and Heated Rivalry at Crave in June 2025. By November 2025, The Deal, the first Off Campus book, had about a million Goodreads ratings. Heated Rivalry had about 42,000. Then it aired, and won a record 16 Canadian Screen Awards and a Peabody.

I looked at Archive of Our Own (AO3), where fans post their own stories, and counted fan works on each couple's tag for every 1,000 Goodreads ratings. I call it the maker rate. Before their shows, The Deal sat at 0.02 and Heated Rivalry at 13.6. Three books can't show that fan writing predicts anything. They show that the book with the fewest ratings had by far the most fan writing for its size.

A question I want to ask out loud first

Fan writing is made by fans, for fans. Should anyone be counting it? And underneath that: is there a way to understand what love stories people are hungry for without being extractive about it? I've been in fandom spaces for years, and I don't think those questions have easy answers. They're part of what this piece is about.

How I counted. Every AO3 number here is a tag total, the single count printed at the top of a public tag page, read logged out and by hand. Most of the past counts come from Internet Archive captures rather than AO3 itself. AO3's Terms of Service forbid scraping content to commercialize it. I didn't scrape AO3, and I followed its terms closely: no bots, no stories opened or saved, no look at who wrote anything. I read about 30 public pages by hand, mostly over the evening of September 28, with the Wattpad cards read on September 29 and a few reference checks on October 4 and 10, and kept only the totals printed at the top of each page.

Before the shows: last in ratings, first for its size

I looked at the two places I actually spend my reading life: Goodreads, where I log what I read, and AO3, where fans post their own stories about characters they love. My hunch was that AO3 would punch above Heated Rivalry's ratings count long before the show. It did. Once I counted fan works against ratings, the two books swapped places.

The maker rate is public fan works on a couple's AO3 tag for every 1,000 Goodreads ratings. Before each show came out, The Deal sat at 0.02. Heated Rivalry sat at 13.6. In raw numbers, though, Heated Rivalry had only 566 works a month before its show, and Red, White & Royal Blue, which comes up next, had 3,018. It's first for its size, not first by volume.

My first thought was that I'd just found the difference between M/M and M/F romance. So I added Red, White & Royal Blue: two men, like Heated Rivalry, but with a ratings count much closer to The Deal's. Its maker rate was 3.8. That's far above The Deal, but still less than a third of Heated Rivalry's. M/M probably helps, but it doesn't explain the whole gap.

Three books can't prove fan writing predicts anything, and I'm not going to pretend they can. What they do show is smaller, and I find it more interesting. The book with the fewest ratings had by far the most fan writing for its size. That sent me looking for a reason, and I came back with a theory.

First, when everything happened. The dates matter, because every "before" number in this piece comes from a different distance before its show.

On Goodreads it came last. Counted my way, it came first.

A note on The Deal: its rate rests on 21 works, so a handful either way moves it a lot. The comparison that does the real work is with Red, White & Royal Blue.

After the shows: every rate went up

Heated Rivalry's fan writing didn't start big. Its couple tag held 71 works in June 2022, 223 in December 2023, and 566 a month before the show. On September 28, 2026 it showed 45,712. The net increase since that last capture is about 99% of the September 28 count.

All three books gained fan writing faster than they gained ratings. Heated Rivalry just did it on a different scale. Its lead over Red, White & Royal Blue grew. Its lead over The Deal shrank, because The Deal started from almost nothing and grew its rate the most.

On Goodreads, the order didn't change.

Heated Rivalry is still the smallest of the three on Goodreads, by about three to one against The Deal. Both things are true at once because the two sites count different things.

AO3 vs Wattpad: two archives, two kinds of writing

I wanted to know if this was an AO3 quirk, so I checked a second fan fiction site, Wattpad. The order was the same, but the gap was much smaller. On AO3, Heated Rivalry's rate is 316 times The Deal's. On Wattpad, it's about 7 times.

Part of the reason may be what each site's tag actually holds. On both sites, writers choose their own tags. The difference is what happens next. On AO3, volunteers link writers' tags to one shared couple tag, and that tag is for stories about the couple. On Wattpad, a tag stays exactly as the writer typed it, and some stories use the tag for something other than the couple, like crossovers or "inspired by" stories.

One more thing I noticed: of the Wattpad heatedrivalry stories I could date, most were started after the show. The tag page said 344 stories on September 28, and 338 loaded when I scrolled it on September 29. I dated 22 of those 338: the 9 oldest, and every 26th story after that. Eight were started before the premiere, and one of those is a 2020 science fiction story, a good reminder that a tag isn't a topic. The other 14 were all started after it. Wattpad's story numbers rise with time, so the sample suggests only those eight, out of 338, came before the show. I haven't dated every story, so it's a sample, not a full count.

My theory: open and closed stories

This next part is a theory, not a finding. It's me starting to put something into words, and it needs far more data than three books. I'll keep working on it in the Notebook.

I think stories tend to be closed or open.

A closed story delivers a complete, satisfying arc. A to B to C, and then it wraps up. In romance that's often exactly the design: you read it, love it, put it down and reach for the next one.

An open story leaves room between the pages. Sometimes on purpose: time passing off the page, a relationship readers only see in fragments, a world so big it can't all be shown. The example I keep thinking of is Tolkien's The Lord of the Rings. AO3's umbrella tag for Tolkien's works, which also covers the films and his other books, holds over 115,000 works. Sometimes the room is accidental: a loose thread, an ending that arrives before readers are ready to leave.

Fans have names for how they fill that room. A canon divergence asks what if one thing had gone differently. An alternate universe drops the characters into a new setting. A fix-it rewrites an ending readers couldn't accept.

If I'm right, a high maker rate doesn't measure how good a book is, or even how popular. It may partly measure how much room it leaves.

Heated Rivalry had a small, devoted group writing into its margins for years before a show was announced. That's what I'd expect from an open story. I'm not dismissing the other explanation, either: M/M is a very large category on AO3, with 7,605,276 works against 3,893,175 for F/M, though a work can carry more than one category, and that surely helps. But Red, White & Royal Blue is M/M too, and it didn't come close. The Deal had 21 works before its adaptation and 450 after. My guess is the show pried a closed book open a crack by putting its world on screen, but the story itself stayed closed. Red, White & Royal Blue probably sits somewhere in between. I'm sorting these three after seeing their numbers, which is exactly what the first question below is meant to fix.

None of this makes open better than closed. The Deal tells a tightly resolved story, and well over a million ratings say readers love exactly that. Closed stories clearly work. I don't know what open ones do for readers, and the usual counts don't show it.

As a writer, this is the part I can't stop thinking about. If openness is something a story can have on purpose, then leaving room might be a choice I get to make. Which doors would I leave ajar for the readers who can't let go?

What this can't tell me

• Three books are three books. All three were picked because they were adapted, so they can't show what predicts an adaptation. They only describe what these three looked like.
• Two archives aren't all of fandom. AO3 and Wattpad are two places readers write, and they disagree about how big the gap is.
• The tags don't match the books exactly. The Shane Hollander/Ilya Rozanov tag covers every story about that couple, across the whole Game Changers series and the show. The ratings are for one novel.
• Tags get renamed and merged. AO3 volunteers rename and merge tags. The Off-Campus couple tag I counted in August 2025 now opens a renamed tag. AO3 doesn't show when tags were merged, so the rise from 21 works to 450 may include works that moved in rather than works written since.
• Announcement dates aren't decision dates. Playback reports that Crave greenlit Heated Rivalry for development in March 2024 and for production by July 2024, before Off Campus was announced. The dates I use are the public ones.
• Every maker count is a floor. Logged out, I can't see works limited to signed-in users.
• Ratings aren't readers. A Goodreads rating is one person choosing to rate. Plenty of readers never do.
• The Wattpad card read is rough. A story can be about Shane and Ilya without saying so on its card.
• Logged out isn't the whole picture. Signed in, Wattpad showed between 1.7 and 2.3 times as many stories on each tag: 315 against 146 for thedeal, 268 against 159 for redwhiteandroyalblue, 783 against 344 for heatedrivalry. On AO3, the signed-in count was about 8% higher: 49,517 against 45,712. I read the tag logged out at 9:38 pm Central on September 28 and signed in later that same evening. A tag doesn't gain thousands of works in one evening, so the gap is about what each view shows, not about timing. I use the logged-out counts because they're the ones any reader can check.
• Open and closed is a theory. Nothing in these counts tests it yet.

Questions I'm chasing next

• Why do some books draw so many writers? I'd like to take a few dozen romances with a high maker rate before any adaptation, and a few dozen with a low one, and read them for how much room they leave. "Open" is very subjective, so the first job is a set of criteria I can score each book against before I know its maker rate: things like time skipped on the page, side characters with unfinished arcs, how tightly the ending is tied off. If the high-rate books score as more open, the theory has legs.
• Can a signal like this exist without being extractive? Counting public works on a tag is one thing. Mining the people who wrote them is another. I counted totals, I didn't look at who wrote anything, and I never will. What I want to know is whether counts like these are useful at all, and what fan writers think is fair.
• Does a small, intense readership tell me anything about how a book will do? If I follow the high maker rate books from the first question, adapted or not, does the Heated Rivalry shape repeat? If it doesn't, I'll publish that too.

The numbers

What I counted. Readers: Goodreads ratings, pooled across editions. Makers: public AO3 works on each book's couple tag, read logged out. Maker rate: fan works for every 1,000 ratings. It compares two kinds of visible activity. It isn't a count of readers who went on to write. The three couple tags are Garrett Graham/Hannah Wells (Off-Campus Series), Alex Claremont-Diaz/Henry Fox-Mountchristen-Windsor, and Shane Hollander/Ilya Rozanov. The Off-Campus tag was called Garrett Graham/Hannah Wells in the August 2025 capture; the old address now opens the same tag under its longer name.

The "before" captures sit at different distances from each premiere: nine months for The Deal's AO3 count, a month for Heated Rivalry's, five days for Red, White & Royal Blue's. Within a book, the fan-works and ratings captures can also be weeks or months apart (August and November 2025 for The Deal).

How I counted. I read every count by hand, one public page at a time. The "now" figures were read on September 28, 2026. A few reference checks, such as AO3's terms and category totals, were done on October 4, 2026. Every "before" figure comes from an Internet Archive capture of the page as it stood that day.

I don't rebuild the past from today's AO3. Writers can backdate a work. Volunteers attach tags after works are posted, so older works can join a tag later. Deleted or locked works drop out. The site's own history keeps shifting, so a capture of the page on the day is the only record I trust for a past count. These counts move daily, a recount on another day will come out a little different, and every rate here is computed from the two counts printed beside it.

Other platforms. Goodreads and Wattpad are commercial sites with their own sign-ins and business models, and both print these totals publicly. I treated them the same way as AO3: public pages, a few at a time, and the totals printed on them. I also signed in once on AO3 and Wattpad to compare counts, and I've labeled those.

Corrections to this piece are logged on the Corrections page. Want the next one? Get updates.