[EM] Guardian 100 Best Novels

[Ross Hyman]

Matthew Aldridge comes to the same conclusion for the voting method
and that My Antonia is an error:
https://mpaldridge.github.io/blog/guardian-novels.html He also found
some books with variant titles that were not properly counted.

On Wed, May 20, 2026 at 2:23 PM Kristofer Munsterhjelm
<km-elmet at munsterhjelm.no> wrote:
>
> On 2026-05-19 22:03, Ross Hyman wrote:
> > I thought the score would be of the form approval - weight*(rank sum)
> > as well but this is incompatible with My Antonia (approvals=4, total
> > rank = 4+5+10+5=24) at #100 and Left Hand of Darkness (approvals=4,
> > total rank = 8+7+10+9 = 34) at #84.
> > Even rank dependent weights can't help because each individual rank on
> > Darkness is lower or equal to a corresponding rank of Antonia.
> > If one removes books not in the top 100, the total rank for Antonia is
> > 2+4+9+5 = 20, and for Darkness it is 7+5+5+7= 24. Darkness is still
> > lower ranked. But now there is the possibility that they choose
> > weights on the individual ranks in such a way as to make Antonia
> > lower, such as making the score for 9 count as 14. But that would be
> > weird.
>
> The really weird part is that the results are so close. If you create a
> custom method (e.g. a Nanson hybrid or something with non-Borda
> positional weights), you'd expect it to disagree more than once, but...
>
> I ran the full data set through the solver and it found that
>         31 * approvals - rank sum
> (unless there's a bug) matches the order for every book except for My
> Antonia up to tie-breaking. However, if My Antonia's Approval count is
> lowered to three, then it, too, matches.
>
> More specifically, near the end, after excluding ties, we have
>
> Vote count      Rank sum        Desired rank
> 4               34              89      // The Left Hand of Darkness
> 3               7               90      // Jacob's Room
> 3               10              91      // Life and Fate
> 3               11              92      // Sentimental Education
> 3               12              95      // The Return of the Native
> 3               13              96      // Pedro Páramo
> 3               14              99      // The Go-Between
> 4               24              100     // My Ántonia
>
> These appear to have decreasing vote counts and increasing rank sums
> within the same vote count bracket, except for My Antonia at the very
> end. But 24 > 14, so if it were counted as 3 instead, we would get the
> proper order.
>
> This is quite odd. Maybe there's a bug in their software? It's hard to
> imagine a method that's not weighted Approval,Borda but is *that* close.
>
> About the only thing I can think of is if their tie-breaker is defined
> as some ranking of every candidate submitted, that they add some
> multiple of the tie-breaker rank to the score. Then My Antonia could be
> at the very end, so that its rank sum pushes it down; but this feels
> kind of arbitrary -- why would it affect *just* My Antonia and not
> overturn any of the other non-tied results?
>
> The tie-breaker doesn't seem to be strictly alphabetical either.
> Catch-22, The Road, and The Go-Between all have the same Approval and
> vote counts (namely (3, 14)), but they're listed in that order, whereas
> a tie-breaker would put The Go-Between second and The Road third of the
> three.
>
> -km