Why The Rankings Just Changed

You may have noticed that the main list looks different this week. Here's what changed, why I changed it.

How Voter Counts Are Weighted

Why voter counts (and names) matter
When a source list has lots of identifiable people behind it—think a national library poll or the NY Times List, we can trust that its choices aren't one person's whim or, worse, a bot's.

Old Approach

  • 1 voter → heavy penalty
  • 2 voters → strong penalty
  • 3-5 voters → medium penalty
  • 6-10 voters → light penalty

Anything above 10 voters counted the same as a list with hundreds. A Goodreads list with 11 votes carried as much weight as a jury of 500 critics.

New Approach

  • Lists with no data or just 1 voter are heavily discounted.
  • Lists that reach the typical crowd size (the median of all lists) get full weight.
  • Everything in between is shaded in: the closer a list is to the median, the less penalty it takes.
  • Transparency bonus: Lists that hide or omit voter names take a small extra penalty—because accountability matters.

Result: small, niche lists still count, but mega-lists finally carry the clout they deserve—and anonymous, zero-context lists lose influence.

Why now?

With AI-generated lists multiplying online, tracking both voter counts and names gives us an extra safeguard. It's our way of making sure we're importing rankings judged by real people, not AI slop.

Automatic Western-Tilt Adjustment

Any list that claims to be global but turns out to be ≥ 90 % Western (North America, Europe, Australasia) now gets dinged automatically. Region-specific lists—"Best Latin-American Novels," "Japanese Classics," etc.—are exempt, because their focus is declared up-front.

Why the Aggregate List Still Skews West

  1. Translation flows are one-way. English-language books are translated into other languages far more than the reverse, so they pop up on more source lists.
  2. Anglophone markets under-translate. Only a sliver of books published each year in the U.S. or U.K. are translations, limiting what most voters have even seen.
  3. The canon has historical inertia. Big "best of" lists, prize juries, and school syllabi inherited a Western core decades ago and tend to reinforce it unless they go out of their way to correct.

When you weight lists by crowd size, you inevitably magnify those structural facts.

What You (and I) Can Do

  1. Use the dedicated non-Western views. Global Canon and Top Non-Western Books let you jump straight to a broad, regionally balanced selection.
  2. Send me better data. Know a high-participation reading poll, library list, or prize long-list from outside the usual Western circuits? Please share it.

Compare Versions

Questions, critiques, or fresh list suggestions? Please contact us