The Problems with ‘Best Of’ Lists

I like “best of lists” as they are fun, convenient and easy to argue about with friends, but beneath their luster lurks notable problems if you seek great works. It’s far too easy to forget the label “best” is an invention granted by the list maker, whose tastes and opinions of quality might not align with yours. Keep this list below in mind whenever you read a top list of anything:

  • Popular is not necessarily good. For a work to make a best of list it has to be popular enough for the listmakers to have discovered it. This means there is a bias towards popular works, which might not necessarily be the best, or even good (What else did the list maker read or see this year that did not make the list? We are rarely told). For example, the  most popular hamburger in America is (probably) made by McDonald’s, and some of the most popular music is made by Justin Bieber. (See Being Popular vs. Being Good). It’s possible for a work to be both popular and great, but it’s not guaranteed.
  • It can take years for works to earn the respect they deserve. Many great works were not popular or respected in the year they were first released. Moby Dick, The Empire Strikes Back (received mixed reviews on release), It’s A Wonderful Life (mixed reviews), The Shining (earned a nomination for Worst Director for Stanley Kubrik), Fight Club, and many more. And of course many amazing works never get the acclaim they deserve.
  • Works benefit from cumulative advantage. Once a work gets a high profile review, it’s more likely to get other ones. This benefit, called cumulative advantage, means that the most well known movies or books aren’t necessarily the best, but they’re the most well known for being the best. Consider this: have you ever been very disappointed by a movie that all the critics loved? Or found a random unheard-of movie and loved it? Cumulative (dis)advantage may be part of the reason for both experiences. And don’t forget, many best of lists can be cheated by people with enough money or influence (raising how grey the line between ethical and unethical marketing can be).
  • Best for whom? Your personal favorites might not be the works you think are actually the best. We all have different preferences for the kind of art we like. A best of list presumes we all share common sensibilities, which may or may not be true for you. It can be far better to get recommendations from people who know what you like, or who you follow specifically because of their sensibilities and preferences. 
  • The most popular can be the least interesting. A book with a 4 star average might be far less interesting to read than a book with a 3 star average. Averages hide the variance of opinions. For example a 3 star average could mean half the readers gave it 5 stars and half gave it 1 star, which would mean the work was highly polarizing (and possibly very interesting for that reason). A book that exclusively earned 4 stars, with little variance, could be a simpler kind of story that was satisfying but far less challenging or memorable.

I do hope you enjoy this year’s round of best of lists. And I hope my own list above you use them more thoughtfully.

7 Responses to “The Problems with ‘Best Of’ Lists”

  1. Sean Crawford

    Hi Scott,
    (hoping my comment will kickstart others)

    Needless to say, there’s great mainstream, and great art.
    For mainstream movies, say, I find the stars rating fairly good, I can pass it on to near-strangers.
    My definition art movies is: to recommend one, I have to know the person, since art is not nearly so uniform in appeal across the population.

    For example, remembering how Orwell and Jack London said the ruling class is not decadent-weak, I was not surprised to see rich frivolous disco dancers reach for weapons such as morning stars in the movie Snowpiercer, —and I thought, “cool!”— but I could only recommend the show to one guy, an artist. (he liked it too) Meanwhile on Monday morning I was saying to several people at work merely that *I* was impressed by Snowpiercer.

    Speaking of movies, I just had to laugh when Terry M. got his movie Tree of Life marketed as being mainstream. That’s the funniest thing since Kurt V. got his books marketed as mainstream instead of science fiction. Both men made more cash than they would have otherwise, so good for them.

    Note: The way to enjoy Tree of Life is to be willing to sit with it on it’s own terms, as you would with a literary book, or any other art, and not be demanding a mainstream movie.

    1. Scott Berkun

      Tree of Life is a great example of how your tastes and preconceptions matter – I’m a Terrence Malick fan, and I liked the movie, but when several people stood up and left 1/3rd of the way through I wasn’t surprised.

      Regarding individual films, and to some extent, books, I’m convinced expectations play a huge role in how much you enjoy it or not, and we are generally terrible at managing our own expectations (often resulting in us being disappointed at good works merely because we expected far more).

      > The way to enjoy Tree of Life is to be willing to sit with it on it’s own terms

      I’m not sure most of us are good at this – certainly for films the way previews are created now, where much of the plot is given away (which I’m sure is done with evidence it supports ticket sales rather than hurts them), we start consuming the actual art knowing a great deal or at least think we do.

      1. Sean Crawford

        As for those infuriating movie previews, I am pleased to learn I’m not the only one who closes his eyes and puts his hands over his ears until they are over. And (exaggeration) I don’t care who knows it.

        I suppose some people can’t handle suspense, finding it soothing to know nearly all the story before they go into the theatre. Such folk would be the lowest common denominator for the previews.

  2. Marc

    Suggestion for an added bullet point …

    Where a list is published and by whom. Some ‘best of lists’ are simply tools for analysis to enable a comparison where comparisons are difficult, for example in music or art. Yes, tastes differ but what is good is different than what is liked. Someone earnestly attempting to tease out an objective measure of good might legitimately use a best of list. Billboard charts are in of themselves ‘best of lists’ although the method by which they are compiled is as opaque as it has ever been. The argument here might be that as long as they are truthful reflections of purchases or streams or concert tickets sold, then maybe what is popular must by definition be good (if we believe in the wisdom of crowds). And maybe the passage of time is a useful objectifying instrument. For example movies and songs and novels that survive the passage of time and remain popular might just be benchmarks of what we should consider to be good, even if we don’t like time. In summary, we might be blue to resolve the controversy by changing the labels … instead of ‘best of’ use ‘most popular’, ‘most liked’, etc. Then at least there is a chance that the title of the list and its purpose are in sync. After all, few ‘best of lists’ are intended as objective measures of goodness.

    1. Scott Berkun

      Good point – the distinction between a list that reflects sales or clicks vs. one that is chosen editorially where the editor can pick any work they like is important.

  3. Fendy Heryanto

    It is indeed a problem. However do we get any good solution or replacement for this “best of” lists that we can use?

    Nowadays, I think asking people in each respective places is better recommendation for me than “best of”. ex: asking for similar or recommended music in “Hotel California” video at youtube can bring you a lists of music nearly similar like that.

    And I wonder whether this problem also occurs for scientific books and historical books (non art).


Leave a Reply

* Required