By James Wallace Harris, Wednesday, October 12, 2016
Classics of Science Fiction version 4 is finally finished. I could not have done it without my friend Mike. I had been dreading doing all the work required to update the eleven year-old version 3. In fact, months ago I had given up on the idea, and just moved the v. 3 lists off my hosting site to this blog. I just didn’t want to relearn PHP and MySQL. Mike has been studying new programming languages like Go and Swift, and I asked him if he thought there was an easier way of doing the project. I told him I didn’t want to mess with a hosting site, a SQL database, or creating a data entry system. We finally decided on using .csv files for data entry, and typed everything in Notepad++. Mike programmed the project in Go. He then developed a complete template based system that generated all the .html files, which we copied to WordPress. It took us weeks to enter in the 65 lists used to create the new Classics of Science Fiction list. After all this studying of science fiction books we’re both looking forward to getting back to reading science fiction.
Titles and authors are linked to content on Wikipedia and Goodreads. Percentage ranks are linked to the actual lists each book were on. Unfortunately, this makes the main page rather large and slow to load.
I’m not sure how many science fiction readers care about such lists, but if you study the new site, it’s very revealing. I love studying how books become popular, and then how they are forgotten. The new main table which ranks the books by the percentage of lists each title is on, replaces the old method of just counting total lists. Not all books could have been on all lists, so to be fair, we calculated the percentage of lists they were on of all the lists they could have been on.
I know all this math might put some people off, so we show both percentages and counts on the tables were titles are ranked.
My favorite table is Versions 1-4 compared. Titles in red are those books that didn’t make it to version 4. Titles in blue are those new to version 4. This quickly lets you see what books are being forgotten, and which books are spreading in our memories.
Each time we make a new version of the Classics of Science Fiction list we love to see which newer books stand out in the crowd. This time we were anxious to see how many books by women writers had become popular in the last decade. We also made a separate list showing which books by women writers are the most remembered in all the lists we gathered. This time, we found eight lists devoted completely to women writers, but most of those lists didn’t try to rank titles. The Left Hand of Darkness by Ursula K. Le Guin was on 43 of the 65 total lists. But we figured it could have only been on 48 of those lists at most. This meant that The Left Hand of Darkness was on 90% of the lists it could have been on. 43 of 48.
Using percentages allowed newer books to rank higher than they did in version 3. We figure this is fair because we’re removing the penalty for being newer. Older books have an advantage under the system of just counting total lists. For example, Ancillary Justice by Ann Leckie was on just 11 lists, but because the maximum number of lists it could have been on was just 22 lists, it’s overall ranking was at the 50% rate.
Ultimately, meta-lists like these don’t mean much. It’s just a statistical way to track how books are remembered. But I find that fascinating. I’m sure fans will be outraged if their favorite science fiction novel doesn’t show up on the list. We didn’t choose these titles. Study the 65 lists we used. I’m sure you’ll probably find your favorite novel on some of them. For the main list, it took being on at least 10 lists. To create the list of books by women writers, we wanted it to be around 100 titles. That required using a cutoff of being on at least 4 lists.
We’ve learned that longer lists get boring for readers. I think ideal book list runs to around 100 titles, but for the main list we allowed it to run to 140. Version 3 ran to 193 titles and it was way too long. Probably for many bookworms, Top 25 or Top 50 lists are more appealing. Just scan down the ranked lists and stop when you feel titles are less than classic. The point of this project is to show which books other readers are remembering, and not to make any claims that these are the best books.