Volume 2: Books to movies
Do films based on books do better than other films, and does a successful book mean a successful film? Same rules as every volume: public data, questions fixed in advance, every result reported.
Play
- Higher or Lower: adaptations Which of two films based on books did critics score higher?
- Today's adaptation One film a day. Guess the critic score from the book.
- Book or movie? Did viewers like the film more than readers liked the book?
- Practice mode Unlimited adaptations, scored against a book-only guess.
What we found
- Reviews: films based on books score lower by 1.8 points on average, and 1 point higher after allowing for genre.
- Money: among films that grossed $10 million or more, adaptations earned about 1.24 times as much as other films from the same year.
- Audiences: IMDb audiences rate adaptations 2.4 points higher than other films, so critics are 3.3 points harsher on adaptations relative to audiences.
- Was the book good? Does it matter? For the 61% of adaptations with a Goodreads rating, the book's rating and the film's critic score correlate at only 0.14, but at 0.28 with the IMDb audience rating.
- Beloved book, beloved film? A film from a higher-rated book was well reviewed 57% of the time, against 44% for a lower-rated book. A mild edge, not a guarantee.
- Money follows readers: among $10M+ films, how many people rated the book and the film's gross correlate at 0.41.
- Awards: films from award-winning books average 4.4 points higher.
- Can you predict it? Guessing from the book's ratings missed by 19.4 points, against 20.2 for guessing the average. A small gain.
Adaptations versus other films
Do films based on books get better reviews?
Of 16,848 scored films, 1,367 are based on a book on Wikidata. Their average critic score is 60.2, against 62 for the rest, a difference of -1.8 points (-3 to -0.5). Comparing each film with the usual score for its genres, the difference is 1 point (-0.3 to 2.2).
Limits: 'No book on record' is not the same as original. Wikidata misses many adaptations, so some films in that group are adaptations too, which shrinks any gap. Plays, musicals and comics are not counted as books. Wikidata's coverage favors well-known films, and only films with a critic score are included.
Adaptations at the box office
Do films based on books earn more than other films from the same year?
Among 1,616 films that grossed at least $10 million, 342 are based on books. Compared with other films from the same year, adaptations earned about 1.24 times as much (1.07 to 1.44). Median worldwide gross: $151M for adaptations, $111M for the rest.
Limits: Gross is not adjusted for inflation or budget. Franchise adaptations (Harry Potter, Hunger Games) are a large share of the big grossers and are counted as ordinary adaptations here. This shows adaptations earn more, not that being a book causes it. Studios may pick books for bigger productions. Budgets were recorded for too few films to compare profit.
Audiences and adaptations
Do audiences rate adaptations differently, and do critics and audiences split more on them?
Among 1,358 adaptations with at least 500 IMDb votes, the average audience rating is 65 (IMDb scaled to 100), against 62.6 for 13,899 other films, a difference of 2.4 points (1.9 to 2.9). Critics scored 4.9 points below audiences on adaptations and 1.6 points below on other films; the difference in that gap is -3.3 points (-4.3 to -2.3).
Limits: IMDb voters are self-selected and the scales differ, so the comparison between groups means more than either number. IMDb voters include people who read the book and people who did not, so this is not how book readers felt. Only films that have an IMDb ID on Wikidata and at least 500 votes are included.
Did readers like the book? Did critics like the film?
Do films from highly rated books get better reviews?
830 adaptations (61%) have a Goodreads rating from 100 or more readers. The book's average rating and the film's critic score correlate at 0.14 (0.07 to 0.21); against the IMDb audience rating they correlate at 0.28. Average film critic score by the book's Goodreads rating, in four equal-sized groups:
Limits: Goodreads ratings come from the UCSD Book Graph, scraped in 2017-2019. Books published later, and books Wikidata cannot link to a Goodreads ID, are missing. Coverage is the share given above, and it favors well-known books. Goodreads raters are self-selected fans, and ratings run high for nearly every book, so the spread between groups is narrow.
Beloved book, beloved film
When readers rated the book highly, how often did critics rate the film highly?
Splitting these 830 adaptations at the median book rating (3.92 stars) and median film critic score (63): 29% had a higher-rated book and a well-reviewed film, 22% a higher-rated book and a poorly reviewed film, 22% a lower-rated book and a well-reviewed film, and 28% neither. A film from a higher-rated book was well reviewed 57% of the time, against 44% for a lower-rated book.
Book rated higher than the film (relative to other adaptations)
- Diary of a Wimpy Kid: The Long Haul (2017) Book: Diary of a Wimpy Kid: The Long Haul, 4.33 of 5 on Goodreads. Film: 4.3 of 10 on IMDb.
- Battle Royale II: Requiem (2003) Book: Battle Royale, 4.24 of 5 on Goodreads. Film: 4.6 of 10 on IMDb.
- The Cat in the Hat (2004) Book: The Cat in the Hat, 4.15 of 5 on Goodreads. Film: 4.2 of 10 on IMDb.
- The Exorcist: Believer (2023) Book: The Exorcist, 4.14 of 5 on Goodreads. Film: 4.8 of 10 on IMDb.
- Percy Jackson & The Olympians: The Sea of Monsters (2013) Book: The Sea of Monsters, 4.23 of 5 on Goodreads. Film: 5.7 of 10 on IMDb.
- Sabotage (2014) Book: And Then There Were None, 4.23 of 5 on Goodreads. Film: 5.7 of 10 on IMDb.
Film rated higher than the book (relative to other adaptations)
- Apocalypse Now (2019) Book: Heart of Darkness, 3.42 of 5 on Goodreads. Film: 8.4 of 10 on IMDb.
- Kai Po Che! (2013) Book: The 3 Mistakes of My Life, 2.97 of 5 on Goodreads. Film: 7.8 of 10 on IMDb.
- Drive (2012) Book: Drive, 3.47 of 5 on Goodreads. Film: 7.8 of 10 on IMDb.
- The Curious Case of Benjamin Button (2009) Book: The Curious Case of Benjamin Button, 3.53 of 5 on Goodreads. Film: 7.8 of 10 on IMDb.
- Big Fish (2004) Book: Big Fish: A Novel of Mythic Proportions, 3.64 of 5 on Goodreads. Film: 7.9 of 10 on IMDb.
- Me & Earl & the Dying Girl (2015) Book: Me and Earl and the Dying Girl, 3.58 of 5 on Goodreads. Film: 7.7 of 10 on IMDb.
Limits: Medians split a narrow range of ratings, so 'higher-rated' can mean a few hundredths of a star. Examples rank each adaptation by Goodreads rating and by IMDb rating within this set (the sites rate on different scales) and show the biggest rank gaps. Both come from self-selected fans, and need 5,000+ book ratings and 10,000+ IMDb votes. Sequels and re-releases are matched to the original book, which can make a film look worse than its own source.
Popular books and box office
Do films from books that many readers rated earn more?
Among 248 adaptations that grossed at least $10 million and have a Goodreads rating, the number of Goodreads ratings the book has (a stand-in for how many people read it) and the film's gross, compared with same-year films, correlate at 0.41 (0.3 to 0.51). Median gross by book popularity, in millions of dollars:
Limits: Rating counts measure how many Goodreads users rated the book, not copies sold, and favor books popular with that audience. Franchises like Harry Potter and Hunger Games are a large share of the top group.
Can reader ratings predict the film?
Using the book's Goodreads rating and popularity, can we guess the film's critic score better than guessing the average?
Trained on 620 adaptations through 2017 and tested on 210 later ones, using only the book's Goodreads rating and its number of ratings, the guess missed by 19.4 points. Always guessing the earlier average (58.1) missed by 20.2.
Limits: A small test set. Differences under a point are within noise. Only books with Goodreads ratings, which excludes most recent books.
Award-winning books
Do films from books that won or were nominated for awards get better reviews?
176 adaptations come from a book with at least one recorded award, 1191 from a book with none. The first group averages 64.1, the second 59.6, a difference of 4.4 points (1 to 7.8).
Limits: Wikidata records awards unevenly, and 'award' includes nominations and minor prizes. Award-winning books skew literary and toward drama, a genre that critics score higher anyway.
Novels, short stories and non-fiction
Does the kind of source matter?
Average critic score of the film by the type of source. novel: 59.4 (943); short story or novella: 61.4 (106); other written work: 61.9 (312).
Limits: Types come from how the book is labelled on Wikidata, which is inconsistent. Groups under 30 films are left out.
How long after the book
Do quick adaptations of new books do better or worse than films of old classics?
For 1262 adaptations with a recorded publication year, the years between the book and the film correlate with the film's score at 0 (-0.06 to 0.05). Average film score by gap:
Limits: Publication years are missing for many books and may mark a later edition. Old books are survivors: only those still read get adapted.
Data
Film scores, release dates, box office and "based on" links come from Wikidata (CC0), the same source as Volume 1. A film counts as an adaptation when Wikidata links it to a written work; plays, musicals and comics are excluded. Audience ratings are IMDb averages from IMDb’s non-commercial datasets (500+ votes). Book ratings are Goodreads averages from the UCSD Book Graph (scraped 2017–2019, academic use; 100+ ratings), joined by the Goodreads ID Wikidata records for the book. Awards are those Wikidata records. The number of Wikipedia languages covering a book is used only to pick well-known books for the games, not in any result.
Limits
- The comparison group is films with no book on record, not films known to be original. Wikidata misses adaptations, which understates any gap.
- Goodreads ratings cover only the books Wikidata links to a Goodreads ID and that existed by 2017–2019, about 60% of adaptations here, skewed toward well-known books. Raw data from IMDb and UCSD are not redistributed.
- Wikidata's "based on" also points to folk tales, scripture and book series (Cinderella, Journey to the West, James Bond), so some "books" are not single novels. Films are counted once, against their best-known source.
- These are correlations. Studios choose books for bigger, safer productions, so adaptations differing from other films does not mean the book causes it.
- Only films with a critic score on Wikidata are included, which favors well-known films.