The user review order for http://www.gog.com/gamecard/dracula_4_shadow_of_the_dragon is either really broken or it is not a coincidence that the good ratings are at the top, even thought they have really bad "helpful" ratings.

I was just wondering why the list puts a 1 /13 above a 55/62, when ordered by helpfullness.
And if this is a common occurrence.

in this particular case it could be simply that the ordering hasn't been updated yet, as this is a rather new release?

On the other hand, I have seen some other games with similar situation (better rating reviews wih lower helpful ratio above worse rating reviews with higher helpful ratio). On the other hand, the total number of people clicking on yes/no should also be in the formula deciding he ordering - otherwise the top reviews would be mostly 2/2 or 3/3 or so. Larger denominator might suggest that the review has more information in it, even if the numerator is small, indicating that there are more people disagreeing with this particular review.

I can't tell if the rating it self is in the formula or not, but definitely remember seeing top reviews less than 5 stars, while there are 5 star reviews on the later pages.

The only updating every so often is possible, it does seem to be working now either way.

Better ratings sell games.

And with the software being as weird as it is, they can chalk it up to a bug. Yay for plausible deniability!

Or they can just say "The more accurate reviews are there, you just didn't do the research".

*adjusts his tinfoil hat*

Disclaimer: I think it is neither of those. Considering some of the bugs we've seen on the forums, I'm going to chalk it up to bad coding.

I hope they use a sorting of the ratings like here:
http://www.evanmiller.org/how-not-to-sort-by-average-rating.html

You really need a third party to point it out? Really? Sure? Yes, okay.

They're arranged by how many people click the buttons underneath and say that a review is "Helpful" against how many said it wasn't.