Friday, May 11, 2012

Star Ratings: Why I Bother With Them at All

If you've ever visited this site, you've probably noticed those nifty star ratings I put towards the end of each 6-star review (if you haven't, you may want to get your eyes checked).  Let me let you all in on a little secret: I really dislike giving star ratings to stories.  But despite that, I have no intention of stopping.  If you're interested in finding out why I hate them and why I use them anyway, read below the break.

And if you aren't interested in that, don't fret; reviews (complete with star ratings) continue on Monday.
I dislike trying to assign reductivist ratings (star ratings, thumbs up/down, decimal scoring, etc.) to fanfiction.  The main problem with them is, as the name itself states, that they're inherently reductivist: they amalgamate every aspect of a story's perceived quality into a single number, which is somehow supposed to represent the worth of a piece of writing.  This makes that number minimally useful to anyone trying to gauge the merits of a story based solely on that number, because different people weight different aspects of stories differently.  If I say to someone, "Story X has show-accurate and surprisingly nuanced characterizations, but there's no central conflict to tie the piece together and the ending is a hasty attempt to reassert the status quo," then that gives them some useful information to help determine whether or not they'd enjoy reading it.  If I say to them, "I gave Story X two stars," that's only useful insofar as they're able to determine how likely we are to have common taste in writing.

Now, it's true that the first example also requires a certain amount of functional comparison; what I call "a hasty attempt to reassert the status quo," someone else might consider a perfectly valid return to normalcy at the denouement.  But I think it goes without saying that this is still patently more useful than a reductivist rating alone.

Furthermore, reductivist ratings invite comparisons which are often inapt.  Heck, I make it incredibly easy to do exactly that, by having a link at the top of this site to a page listing all my reviews sorted by star rating.  Vengeance and Fashion got four stars, while The Misted Stage only got three?  That must mean Vengeance is a better story, right?  Of course, comparing these two is about as sensible as comparing The Hitchhiker's Guide to the Galaxy with A Wrinkle In Time, or arguing the merits of Airplane! versus Planet of the Apes.  It's true that some people will try to do precisely that anyway, but it's a silly argument no matter what; you're comparing two things which are sufficiently different that they need to be judged on their own merits, not in relation to each other.  

But as I said, reductivist ratings not only allow for these kinds of comparisons, they encourage them.  So why do I bother with star ratings in the first place, if they cause so much trouble?

For one thing, they make my job quite a bit simpler.  My reviews can get a bit disorganized (as you may, just perhaps, have noticed), and putting a star rating at the end gives me a chance to sum up the major successes and faults of a story in a way that makes reasonably clear which I thought were more prevalent.  At the end of a review (as opposed to in place of one), I think they can provide a conclusion more coherently, if probably less effectively overall, than words alone.  Moreover, people like ratings; they're nice, clean, concrete things to hold onto, to argue about or to nod in agreement with.  The list of reviews by star rating?  That wasn't part of the site for the first couple months I was doing reviews.  I added it because visitors kept e-mailing to ask if I'd put one up.  

I don't like trying to assign a numeric value to a story, and I really hate doing it when I give out a low number (especially to stories that I like, or at least that I like more than the low rating seems to me that it indicates).  But despite the flaws any sort of reductivist rating brings to the table, I'm of the opinion that they also provide an excellent jumping-off point for discussions, a convenient shorthand for summary statements, and are fundamentally appealing to most people.  So although they frustrate me at times (especially when I've been staring at an open review doc for twenty minutes, rereading my notes and occasionally pulling the story back up, trying to decide if a given fanfic merits two or three stars, and I suddenly become aware that this is what I do for fun), I plan to continue giving star ratings to the 6-star stories I review.


  1. Hmmm, three stars.
    I kid but I understand how you come to these feelings.

  2. I was going to write something more clever and wordy, but I erased that and simply will write this:

    I know that feel, bro.

    At least you have clear standards set out for what the stars mean. Which is more than most (Is 7.5 out of 10 equivalent to a C grade, or 2.5 "points" above dead average?).

  3. Damn, and I thought it had something to do with ninjas.

    You're so much uncoolerer than I thought you were.

  4. I'm glad you use them. I've found that your ratings are very close to what I'd give a story on EqD, so I find them very useful in gauging how much I'd like it. I still read the full review, of course, and that can lead to my reading a story I wouldn't have read based on the star rating alone, though this is very rare. As a general rule, I'll only read a fanfic if you've rated it 4 or 5 stars here, though I'll make exceptions for newer comedies with an interesting enough premise.

    I also love how your ratings, at least the last time I checked, follow a Gaussian distribution

  5. Don't be too hard on yourself. Your star ratings have a purpose, a meaning, as other have said. Not to mention, if I didn't want to read your whole review, I could check out the star rating and the blurb beneath it for a quick "what did he think of it and would I like this?" That's more than a star rating by itself would ever accomplish.

  6. Any numerical rating system is going to be suspect like any other rating system. There's certainly an advantage to using one and I think that individuals should take advantage of both the numerical and lexical (essentially the plot of the "Phantom Tollbooth").

    Now, I watch a lot of old cartoons (almost all from the 30's-50's, although I don’t think this would be a surprise to anyone), and since at least 2008 when I put a top fifty list together, I needed an easy way to mark off what I have seen and where they came from. I also wanted to see where the one's I liked the best came from. That's why I implemented a tier system when creating the list (although it’s still subject to human error). It gives me a quick guide to look at and it can also be helpful in recommending cartoons to others, or which one to watch when I bored. And I don’t have time to write a review on all of them (a good detailed review that is). I’ve since expanded beyond to other mediums because I wanted to do the same (and I have given them meanings). I find it useful for myself.

    Let me add another reason why they can be helpful because words can be (and often are) ambiguous. I think this bit from “The Looking glass” sums it up well.

    'When I use a word,' Humpty Dumpty said, in rather a scornful tone, 'it means just what I choose it to mean — neither more nor less.'
    'The question is,' said Alice, 'whether you can make words mean so many different things.'

    Even if our dictionaries say the same meaning, what a word means to us can differ widely, so use and interpretation will as well. To you, the word ‘terrible’ may be harsher than the word ‘abysmal’ but it isn’t for me. But a number is not (for the most part), three is always greater than two, two is greater than one, five is greater than four and three, and two equals two (at least when both share the same base). If I give a book, a film, a comic, or a painting a five and give another a four, I am saying that the former in some way shape or form that it is better than the latter.

    Now all that said, I still find such ratings useless for the most part by themselves. Why is that? Well, because most ratings these days are massed together from a bunch of people and quite frankly that’s an average without system (I have similar opinions when the same is done for rankings in a list). We all rate things differently and thus find value in separate things. Grouping say ten different people with ten different opinions in some number form is just messy and says little; it’s why I don’t use the star system on EQD. Even individual ones can be questionable; my own system is useful, but only for myself. I have doubts that anyone else from my friends to some future historian that somehow discovers my files will find much merit in them.

    That why it's better to use both words and numbers, to take advantage of the explanation of the former and the clarity of the latter. And at the end of the day all reviews from a thumbs up or down to a long essay are simply that, an opinion that we use to explain to others our thoughts on something. As my favorite critic said, there is fundamentally no difference between a regular moviegoer who leaves a theater thinking “What a piece of …” and a critic who tries to explain what he or she found lacking in the same film. To me, the best (as in most accurate) way to tell someone whether or not he or she will enjoy something is very simple, “Try it for yourself and form your own opinion”.

  7. My word, how HAVE I only just discovered this blog? I'm up to review 32 and I don't think I've been dissatisfied with a single post. You are possibly the most consistently interesting person I have encountered in the fandom.

    As for the numerical/text reviews issue, I'm wholeheartedly in the middle camp, with the proviso that I'm approaching this an author and reader. A star gives you an impression - a review reinforces or tempers that impression.

    As a reader there are any number of fics that I've dismissed at first because, despite being 6-star'd, the summmary didn't grasp me, then gone back to after someone has pointed out that "Yes, Fallout: Equestria is a crossover that looks patently ridiculous, but it also has a mean eye for a moment and is a fantastic example of how to take an established universe and integrate it into another seamlessly."

    As an author, I see my star rating and go "Okay, it's 6-star'd, so people like it. But what do they like? What do they hate? Where am I falling down?" A comment saying "THIS IS AWESUM" is nice, but unhelpful. Which is where reviews (and, I suppose, critical commenters, but if I'm brutally honest I've seen very few of those on ANY story) come in. It takes the huge cloud of averaged opinion, all too easily gutted in usefulness by one star-bomb, and focuses it into something that I and presumably every other author wants - something to help them improve.

    And, in the end, it's all opinion. There are fics I've (unwisely, most likely) ignored despite having a high rating AND having been reviewed well. Your "Who'll like this" segment is great for that.

    ... It's too early in the morning for a satisfactory conclusion. Reviews are cool and star ratings are cool, when combined, but in the end the only way to know for sure is to give it a shot. THE END.