By the Numbers: 5,433 Movies Across 11 Years Say Ratings Went Up 0.32 Points — But Only Horror and Sci-Fi Earned It
Are movies actually getting better in 2026, or are we just rating them higher? I ran the question against every movie in the spameri.cz database with at least 100 Trakt votes — 5,433 films from 2015 through 2026 — and the answer is yes to both, but the math is more interesting than that.
Are movies actually getting better in 2026, or are we just rating them higher?
It's the kind of question that gets asked a lot and answered with vibes. So I ran it against the database. Every movie title we index on spameri.cz with a Trakt rating and at least 100 votes — 5,433 of them, spanning 2015 through 2026. Then I asked the boring questions. What's the average score per year? What does it look like broken down by genre? And does the picture support the lazy "everything is inflating" take, or is something more specific happening?
Let me show you what the numbers reveal.
The Headline Number
Here's the year-by-year average Trakt rating for every movie in the database with 100+ votes:
| Year | Movies | Avg Rating | |------|-------:|----------:| | 2015 | 810 | 6.33 | | 2016 | 871 | 6.32 | | 2017 | 915 | 6.39 | | 2018 | 961 | 6.43 | | 2019 | 880 | 6.55 | | 2020 | 677 | 6.52 | | 2021 | 752 | 6.58 | | 2022 | 751 | 6.51 | | 2023 | 664 | 6.56 | | 2024 | 540 | 6.60 | | 2025 | 390 | 6.58 | | 2026 | 66 | 6.65 |
From 2015 (6.33) to 2026 (6.65), the average movie rating is up +0.32 points. The steepest gains happen between 2017 and 2019 (+0.16 in two years), then things basically plateau in the 6.55–6.60 band from 2019 onward.
Note the 2026 number — 6.65 — comes from only 66 films so far this year. That's a real signal but a small sample, so I wouldn't bet the rent on the exact year-only figure until the rest of 2026 lands. The trend across the eleven-year window, though, is unambiguous: today's average movie scores higher than 2015's average movie.
So case closed, ratings inflate, right? Hold on.
Where the Inflation Lives
A real, neutral inflation effect would lift every genre roughly the same amount. A quality shift would lift some and leave others flat. So I split the database into two eras — Late 2010s (2015–2019, 4,437 films) versus the current era (2024–2026, 996 films) — and asked each genre to show its work.
| Genre | Late 2010s | 2024–26 | Δ | |---|---:|---:|---:| | Horror | 5.77 | 6.18 | +0.41 | | Science Fiction | 6.11 | 6.52 | +0.40 | | Thriller | 6.07 | 6.32 | +0.25 | | Family | 6.80 | 7.02 | +0.22 | | Comedy | 6.26 | 6.47 | +0.21 | | Action | 6.32 | 6.52 | +0.20 | | Music | 7.09 | 7.23 | +0.14 | | Adventure | 6.61 | 6.74 | +0.13 | | Documentary | 7.18 | 7.29 | +0.11 | | Drama | 6.58 | 6.65 | +0.08 | | Animation | 7.04 | 7.02 | −0.02 | | Mystery | 6.20 | 6.16 | −0.04 | | Romance | 6.57 | 6.48 | −0.09 |
There's the shape. Horror and Sci-Fi are the engines — both up roughly 0.40 points, more than the overall lift. Thriller (+0.25) and the comedy / action / family cluster (+0.20-ish) do the next-tier work. Drama, the genre that should be most resistant to taste-cycle effects because it's the biggest catch-all bucket, barely budges (+0.08).
And then there's Romance: the only major genre that actually rates lower in the 2024-26 era than it did in 2015-19. Not flat. Down 0.09.
That's the finding that breaks the inflation story.
Real Quality vs. Real Inflation
If the lift were pure rating inflation — voters being kinder over time, scoring scales drifting upward, social effects pulling everyone above 7 — then the effect would hit Romance the same way it hits Horror. It would lift Animation the same way it lifts Thriller. The shape would be roughly flat across genres.
It isn't. The shape is selective.
The two biggest gainers — Horror at +0.41 and Sci-Fi at +0.40 — both correspond to documented critical and audience renaissances. Horror spent the second half of the 2010s rebuilding around the A24 / Blumhouse model: smaller budgets, sharper concepts, directors with vision instead of franchise mandates. The numbers say audiences noticed. Sci-Fi has its own post-Dune, post-Everything-Everywhere recovery happening — a genre that used to mean either superhero or Transformers-shaped action figure is now where some of the most ambitious filmmaking lives.
Thriller's +0.25 fits the same story — the elevated genre era pulled its closest neighbor up with it. Family at +0.22 plausibly reflects streamers actually competing on family-coded prestige instead of dumping content. Action at +0.20 lines up with the John Wick / mission-statement-action trend pulling the average up off the floor.
And Romance going negative fits too — it's the genre whose theatrical pipeline collapsed hardest after 2019. The big screen rom-com largely became a streaming bet, then a Hallmark bet, and the median film in the bucket is rated by a smaller, more polarized audience. The numbers reflect that.
That is not the signature of a number-line drifting up under everyone's feet. That is the signature of which kinds of movies got better and which didn't.
The Honest Caveats
A few things to keep in the box where they belong:
- These are Trakt ratings, mirrored into the spameri.cz database. They're a real signal from a real audience, but they're not native spameri.cz user votes — we don't have enough native voters yet to make that a meaningful aggregate. (I called this out in the Cannes-vs-Oscars piece a few weeks back and it's worth re-flagging.)
- The 100-vote filter matters. Without it, half the sample is direct-to-VOD films with seven votes and a 9.4 average from the cast's friends. The cutoff is the same one I used for the mid-budget ROI piece earlier this week.
- 2026 is in progress. 66 movies is enough to know the year isn't catastrophically off-trend, not enough to put a stake in the ground for the year alone. That's why I bucket 2024–2026 together for the genre comparison — about a thousand films, which is a real era-sized sample.
- Selection bias is real. What enters the database with 100+ Trakt votes is itself filtered toward movies people heard of. But that filter is roughly consistent across years, so era-to-era comparisons are still meaningful — it just means the absolute number is "average rating among the discoverable" rather than "average rating among everything ever shot."
None of that changes the genre shape.
The Data-Backed Conclusion
So: is rating inflation real?
Kind of. The average movie does score 0.32 points higher in 2026 than in 2015. That's a real number. If you stop reading there, the inflation story works.
But the lift is not flat. It's concentrated in Horror and Sci-Fi — the two genres that actually changed during this window — with Thriller and Family riding the same wave. Drama, the biggest and most representative bucket, barely moves. Animation is flat. Romance is down.
That shape says the right answer is closer to: a few genres actually got better, the rest mostly didn't, and the headline average is half a quality story and half a composition story. The films we're rating in 2024–2026 lean more toward the genres that had a renaissance, and within those genres, the median film is genuinely stronger than it was a decade ago.
Which is a much more interesting answer than "everyone gives everything a 7 now." Because it isn't true that everyone gives everything a 7 now. The Romance audience is still out there giving 2026 movies 6.48s, just like the database always knew they would.
Data sourced from the spameri.cz title database — Trakt ratings, movies with ≥100 votes, 2015–2026 (n = 5,433).
Comments (0)
Log in to leave a comment.