Heads Up: Don't take this too seriously, I think, perhaps, it's primarily nonsense pitted with opinion and anecdote. But hey, it's the 21st Century. So it goes.

The Idea

Do you know who's a better fantasy writer than Tolkien? Terry Pratchett. That's a fact*. Also, H.G. Wells is a terrible writer (see *). There are dozens of (fairly useless) lists online informing us of the greatest science fiction novels, the top 100 horror novels, fantasy you should ensure you read before your untimely, pre-meditated death. There are also books dedicated to these things. These, on the other hand, are important documents and have a bearing on how the genre progresses and is remembered. The published books from the likes of Pringle, Cawthorn, Jones, Rennison, Di Fillipo and Broderick take an academic reader's approach wherein books are chosen from a larger set for inclusion; they’re built from the ground up from expert opinions. Online lists tend to bleed into existence from other lists, popular opinion, awards and their ilk. They kind of grow organically. The published books provide a more viable resource, primarily because they're based on expert rather than popular opinion.

So, it is with total disregard for that little nugget that I decided to put together a quantitatively-derived list of the greatest speculative fiction writers. The first thing to point out is that we're talking about writers, not individual books. A metric based on oeuvres, not stories. And, this is based on public opinion, not some meta-study crunching all the lists together. We'll get to the criteria for inclusion shortly, so be prepared to roll your eyes; this list will have a more caveats than adverbs in a Stephen King novel (yeah, yeah, he uses them, he just pretends not to). Pick up an academic book and you'll be presented with paragraph on paragraph of (almost indecipherable) text outlining the virtues or Wells and Verne, Stapledon, Heinlein (really?**), Gibson, Clarke, Godwin, Le Guin, Orwell, Kepler, Hunt, Jones, Dante, Huxley, Dick, Lem, Milton, Shelley (always Shelley). There's usually a bunch of French writers too, because, let's face it, they pretty much invented the genre. Ballard and Moorcock get dropped in there too because they broke more ground than the molemen of Molopolis. Tolkien too (the ratio of meta-Tolkien published words to actual-Tolkien published words must be the highest of any speculative writer). Lucian, he's always there - he took Asia Minor to the moon. Swift, Dumas, Defoe, More, Shakespeare; not science fiction, but more likely to get a mention than poor old David Gemmell. Is this justified? Well, yes. Yes, it is. Fin. It's justified because these writers formed the bedrock of the genre. They provided for (and provide) the entertainment. Gene Wolfe too. Voltaire. Kafka. Calvino. Borges.

But not all speculative fiction readers are made equally. Some of us are of a superior quality, we (yes, that includes me, certainly; I have many leather-bound books and my house smells of rich mahogany) like to look at the genre for its attributes beyond the aesthetic, emotive or entertainative (not a word, but...academic obfuscation etc.). I for one am a big fan of symbolism (rockets are penises, fuel tanks are testicles, reality symbolises reality, 10x speed of light and anti-gravity boots represent authors in denial). I jest (I'd rather you felt patronised than offended), but I’m sure many readers would accept that their primary requirement is entertainment.

So, H.G. Wells is the greatest writer of speculative fiction? The time machine, tanks, invisibility (with 'science'), chimera (Stevenson gave us the Incredible Hulk [Frankenstein and his Creature in one tidy package], Wells ramped the idea up on a Disney World scale). Well, no. Not according to Goodreads Readers. Wait, wait, Get off your high unicorn and hear me out. I have maybe a dozen Dan Simmons books and I wanted to pick one to read. Ilium (Homer = serious) looked interesting, but it was a thick tome. I wasn't sure I wanted to invest as I knew little about the book. Reviews seemed generally favourable, but I didn’t know the reviewers. Then I saw a link to GoodReads and that helped me decide. Long story short, I figured it might be good to get an average rating for the most popular authors. So, I did.

*   Not a fact. #fakenews
**  I find Heinlein too preachy.

The Criteria (i.e. caveat central)

I took 84 of the best, most established speculative fiction writers. The criteria for choice was fairly loose: authors who have written more than 10 books with the least-reviewed book having more than 500 ratings*. Their oldest book must be no less than 20 years old to allow for some sense of permanence. Repeat titles (omnibuses, best of etc.) were excluded.

Caveat 1. Tendency toward contemporary authors: I'm told that the Internet was only on 4Kbps dial-up when Wells and Verne were writing, so not many people bothered to add ratings to GoodReads. Marketing, hype and newness encourage us to read books as they are published. I'm sure Peter F. Hamilton outstrips Isaac Asimov for sales nowadays. Styles and trends change, writers build on their forebears' work, thus modern work, to the majority, is seen as better than the classics.

2. The series factor: when you've got four million words with which to build your world, readers are going to be more drawn to that world. It takes great skill to keep those readers invested (here's looking at you volume eight of Wheel of Time), but it also gives a distinct advantage. This is why I have in some cases included a secondary, non-series vote (basically excluding Hobb's Realm of the Elderlings, Erikson's Malazan, Gemmell's Drenai, Pratchett's Discworld, Banks' Culture etc.) Nested caveat: I'm talking about writers for whom one series or universe encapsulates the majority of their output.

3. GoodReads ratings: GR Ratings sometimes don't make perfect sense. For example, Dune books 1-3 were rated 4.22, 3.87 and 3.92. The Dune Trilogy (which had 55,000 ratings) was rated 4.36. The explanation could be to do with a continued read; the three books make much more sense as a single volume. But ultimately, it shouldn't increase the average by such a margin. There's a lot of noise in there. Also, A Song of Ice and Fire #6 (not yet written) has a rating of 4.41 over 6336 votes.

4. Most people judge books on how much they enjoyed them not on their value to the genre. That's fair enough. Ask me if I'd rather watch The Avengers or Tarkovsky's Solaris (I'll say Solaris every time, because I'm serious [six minutes of spaghetti junction with rocket sounds overdubbed, come on Andrei]). But no, entertainment factor rather than quality is what drives these ratings. That's why the Three-Body Problem has a rating of 4.05 (I gave it a 2.0 because I'm serious).

5. Not every book is used in the criteria. Omnibuses etc. were excluded, anthologies obviously. But also, a couple of authors had books excluded because I wasn't familiar with them - I think they were mostly non-fiction items. So if you see that J.G. Ballard had ratings from 21 books included and you think he wrote more, well he did, particularly when you take into account the format. Some of them won't have met the criteria too.

6. Sort by popularity: this is a tricky one and I don't know how much it skews things, but I sorted the GoodReads ratings by popularity which usually correlates with the actual rating. That means that, for example, Richard Matheson's poorest books will generally be the least popular and less likely to make the list. I've tried to use my bibliographic knowledge to include familiar titles, but my knowledge isn't complete (I know...gasp).

7. The ten books minimum precludes some writers, Olaf Stapledon for example, who has only four books with more than 500 ratings, and many contemporary writers including J.K. Rowling.

8. Brandon Sanderson, John Scalzi, Joe Abercrombie and Alastair Reynolds, all first published within the last 20 years (at least those titles with more than 500 ratings). Their stats therefore don’t count in the results table but are included because they deserve it. Erikson, you just made it.

9. The absent many: yes, many authors who perhaps deserve to be on here are absent. My feeling is that they would be lower on the list anyway. I have included as many as I can think of, such a Simak, to test this idea. I only excluded one writer outright, because his ratings were so bad, and he's a nice guy.

10. Borges has been excluded from the results because some of the lesser-known works can’t really be classed as speculative fiction, meaning he doesn’t meet the criteria. But he stands above all the other writers in this list, more so when accounting for publication dates.

This is just a bit of fun. Don't take it too seriously. Whilst certain interesting points can be gleaned from this, it's verging on anecdotal throughout given the pretty unacceptable use of statistics and the sample provided.

* Aldiss didn't quite make the cut here, but I wanted to know.

The results:

Sample size: 84 writers covering 1995 books

*Anne McCaffrey deserves a mention here too. She maintained a mean of 4.02 with an S.D. of 0.13 over 39 books.

** HPL’s choice was mostly limited to Arkham House publications

*** Sanderson has four, including one above 4.75

Interpretation.

Well, I'm not a scientist, a statistician or a sociologist (I know that's probably a shock), so my analysis can only be treated as anecdotal (it's the 21st Century and this is on the internet so you should really just believe it, quote it and bind it to your cognitive biases, but believe me when I say that that's not enough - and if you think it is, then give me a call and I'll tell you which books you should be buying from our shelves. Don't forget to give us a thumbs up. Don't forget to subscribe and supports us on patreon. Beg, beg, beg (disclaimer: we're not on patreon, we opted to have a business model instead).

Series bias and contemporaneousness play a big factor. I don't want to downplay the achievements of our 'winners', but I'd be happy to dismiss the entire thing given that Ballard has been given such a lowly score. I am confident (remember: I'm a bookseller not an academic) that the intersection of Ballard readers and GoodReads users are missing the point. Ballard was an outstanding writer. And this is the crux of the problem, and really when it comes down to it, it highlights the main value of this list: ratings are problematic.

A reader of 'important' speculative fiction is as much at home with Wells, Verne, Shelley, Shakespeare, Beckford, Byron, Haggard and Kafka as they are with Erikson, Gemmell and Jordan. But the reverse isn't as frequently true. Maybe this comes across as a little snobby, and it probably is. But a new reader, or for that matter an experienced reader, trying to pick between Stapledon's Star Maker (3.93) and Brandon Sanderson's The Way of Kings (4.65) is likely to pick Sanderson. They'll get a damn good ride (I'm told, I haven't read the Stormlight Archive), but they are going to miss such an important part of the genre. It may not be a problem, but the more we place value solely on the entertainment factor the more the genre stagnates. Ask yourself who has really pushed the boundaries this century. Who will be remembered as the 21st Century equivalent of Wells, Verne, Ballard, Moorcock, Kafka, Borges, Tolkien, Macdonald, Lovecraft or Howard. The academic work, the writers who think how they can make their mark on the canon, the J.G. Ballards of this world are the track on which the train runs. If we don't read them and respect their value then the train will go nowhere (not erewhon #serious).

Full Results

Sorted by descending mean