Anime Taste


Choosing an anime to watch can be an easy process if approached with the right framework: anime and dating are the same thing.
Scores that professional critics or amateur reviewers provide are useless, in most cases
– Use 0-100 scale and interval scores to sharpen and calibrate your own taste and self-knowledge.
– When people agree or disagree about a particular work, it does not mean much. Only after examining the reasons why we agree or disagree can we gain a better understanding and appreciation of different perspectives.


Item choice for consumption
– Will you watch a particular anime?
– What anime will you watch now?
Media discourse
– Why we agree/disagree?
– Purpose of an anime review
Self-knowledge and taste crystallization
– How to rate anime for yourself
– What do you really want?

Anime over the years

Will you watch a particular anime?

Some factors I would consider in deciding to watch a particular anime:

Setting. Fantasy, science fiction, high school, etc. If it is in a high school setting, it better have some other awesome qualities…
Genre. Comedy, drama, horror, action, etc. Most anime are extremely difficult to classify in one genre, but it is helpful to define a set nonetheless. For example, if it is a surreal comendy like Sayonara Zetsubou Sensei, Azumanga or Nichijou, probably I’ll watch it. Also, if it is some kind of a horror/crime/mystery film, it is very unlikely that I’ll watch it.
Director / Screenwriter / Manga. Some directors can make a big impact, but are not sufficient to save an anime (e.g. Space Dandy by Watanabe). If I read the manga and liked it, I am more likely to watch an anime based on it.
Character design. Some people may not like particular designs and will not watch the work no matter how good others make it sound.
Personal recommendation by other people who know me at least somewhat.

What about the plot? Well, we may not know what a particular work is about in the long-term (the short plot synopsis does not count), but most stories are the same – it is the execution that matters the most. A personal recommendation is kind of like a proxy to knowing a plot because people can tell you that things will be awesome (e.g. Hunter x Hunter).

What do you want to watch now?

Let us say you have a list of anime that you decided to watch based on some criteria. Which one would you select for this Sunday? To make the selection easier, think about what mood do you want to induce. Suppose you are somewhat sad and want to see something funny to lift your mood. Still, you have to choose from a set of comedies on your list. You can just select randomly or on some set of arbitrary criteria. Alternatively, if your list is long, consider sqrt(n) rule.

Step 1: Estimate how many anime you have on your list, n.
Step 2: Calculate the square root of that number, sqrt(n).
Step 3: Look at the descriptions and art of the first sqrt(n) anime – the best of them will set your benchmark.
Step 4: Continue looking through descriptions and art, then settle down with the first anime to exceed the benchmark set by the initial sqrt(n) anime.

The same method can be used with dating too! hehe

So where are we so far? Isn’t the above process obvious (except the sqrt(n) part)? Yes, it is, but I would like to point out that throughout the process we could just use a binary system without any continuous rating scale. For example, you just put checkmarks by your criteria:

– Setting (-)
– Genre (+)
– Director (+)
– Character Design (+)
– Recommended (+)

Decision: Yes, I’ll watch it.

You don’t need a complex rating system to choose an anime, and you don’t have to rely on any average score or even a distribution (like ANN) to choose an anime to watch, but a well-written review aimed at helping you make a decision may be useful.

We’ll go back to the question of the rating system in the next section, but, before we proceed, let us establish some basics and examine the issue in the context of discourse.

Belldandy Ah My Goddess


Let us consider utility of watching an anime as a function with n factors as arguments.

Utility = f(X1, X2, … Xn). All other expressions are with respect to this one.

f(.) := arithmetic mean
x1 := Romance
x2 := Plot
x3 := Emotional reaction
x4 := Music

x1 = 80%
x2 = 80%
x3 = 90%
x4 = 85%

Utility = (90%+80%+80%+85%) / 4 = 84%

This can mean that your satisfaction or rating of this anime is 84%. The actual function and factors involved are more complex or much more simple – humans are not perfectly rational beings. In fact, some people can rate an anime just on a general overall feeling, and it is fine. The point here is to make explicit some potentially implicit content and highlight what two identical or different ratings can actually imply.

Skuld Ah My Goddess


Skuld, being an engineering genius, knows that she can do better than an arithmetic average ๐Ÿ˜›

Different function from Belldandy, same factors
g(X1, X2, … Xn)

g(.) := geometric mean
x1 := Romance
x2 := Plot
x3 := Emotional reaction
x4 := Music

x1 = 75%
x2 = 85%
x3 = 82%
x4 = 95%

Utility = (75%+85%+82%+95%)*(1/4) = 84%

Note that the score is the same, but both the function and the scores for each factor are different.

How often you witnessed someone enjoy the same thing as you, especially giving the same score, and – an even more special occasion – do all that on a somewhat obscure anime? What does it mean? Well, obviously it means that we are so similar that we should get married right now! ๐Ÿ˜›

When people disagree, it is more or less clear that they may use different criteria or disagree on a particular aspect, but when people agree, it may be unclear why because you may implicitly assume that you are using the same criteria. People may like the same anime and even give the same overall score, but for different reasons. This is one of the main problems with any recommendation system. The fact that people like the same thing is insufficient – we need to know more about the factors that affected your experience. Determining the function itself would be close to impossible though.

Urd Ah My Goddess


Same function as Belldandy, different factors
f(Y1, Y2, … Yn)

f(.) := arithmetic mean
Y1 := Science
Y2 := Animation
Y3 := Character design
Y4 := Singing

Y1 = 90%
Y2 = 95%
Y3 = 85%
Y4 = 65% (Yes, Enka was there!)

Utility = (95%+90%+85%+65%) / 4 = 84%

Now we have different factors, different scores, but the same function. The score is still the same.

Hild Ah My Goddess


Daimakaichล knows! ๐Ÿ™‚

Different function from Belldandy, different factors
g(Z1, Z2, … Zn)

g(.) := geometric mean
Z1 := Voice acting
Z2 := Editing
Z3 := Color design
Z4 := Minority portrayal

z1 = 79%
z2 = 84%
z3 = 79%
z4 = 60%

Utility = (79%+84%+79%+60%)*(1/4) = 75%

Note that Hild’s score is very different from Billdandy’s, but that is due to a different function, different set of factors, and factor scores. Thus, whenever we discuss our differences in opinion, it would be helpful to know why exactly our perspectives diverge. At least theoretically, it would be possible for Hild to get the same score as Belldandy if both used the same factors and utility function.

Moreover, note that the factors used for the evaluation of an anime post-hoc are different from the ones we use for anime selection. This is a crucial difference that plays a big role in anime review writing and recommendation.

What is the purpose of a review?

Let us consider different review types.

– Commentary, insight, analysis
– Segmental review: Animation, Music, etc.
– Brief emotional summary
– Review that actually helps a consumer to choose whether this type of anime is something that is worth the time investment.

Unfortunately, not many types of reviews fall in the latter category. I might be interested in a formal rhetorical analysis or just some general commentary after seeing a film, but such analysis is not helpful at all, especially those that contain spoilers, in helping me choose a film to watch. If the purpose of your review is to help others make a good choice, consider mentioning previous works that are somewhat similar, and elaborate extensively on the reasons you enjoyed this particular work.

What about the segmental reviews that try to deconstruct a work into parts such as animation, music, etc, and also provide an overall score? Now we are moving into the objectivity and subjectivity topic.

Objective vs Subjective

What is the difference between a professional film critic and an amateur film enthusiast? The difference is in scope. A professional film/art/literature critic can dissect a work into very fine categories, some of which most people are not even aware of. This gives an impression that the rating of a film critic is somehow more “objective”, than the one of a regular spectator. However, the subjectivity persists in both cases, the only difference is that a professional critic is subjective about more aspects of a work than other people.

Moreover, it really does not matter. Certainly, it can be very educational to learn about the evolution of the sentence structure and rhythm within Hesse’s work, and the brilliance of Edgar Allan Poe’s poetry, but each person may have a different purpose in consuming a piece of media.

Yes, some work may be wonderful theoretically, but I may not enjoy it for a variety of reasons. Sure, I may consume some work for the learning purposes, but, for most people, we are in a business of engineering our own mood. We want to feel better, relaxed, inspired, intrigued, puzzled, surprised, thrilled, excited by the experience. The score that a critic has assigned to a work may not be useful in assessing how much joy you’ll obtain from the experience. Thus, the film scores assigned by other people are mostly useless.

Instead, what matters more is knowing the probability that you’ll enjoy this work, given your previous experience and your preferences. Now, there are many sophisticated methods to do this type of prediction, but it is beyond the scope of this article and most of them are terrible anyway.

Instead of focusing on the scores, describe why you enjoy or didn’t enjoy the work in as much detail as possible, but without spoilers if you are writing a review trying to help other people make a choice.

However, if you are trying to put a score on a film for the purposes of quantifying your own impression, it can be a very rewarding and insightful experience. By carefully assigning a score to the films you watch, you are sharpening your ever-evolving taste and understanding your preferences better.

Bitter Coffee

How to quantify your impression

Provide both a point estimate and an interval associated with your uncertainty: 93 (91-94). Why should you bother with such a precision when A, B, C, D / stars / 1-5 will do just fine? Yes, all those systems are fine, but the point of 100-point scale is to force you to think about the films very carefully, deeply consider what you really value in a film, what brings you joy, and what factors are more important to you.

When setting up a new system, start with a few fundamental titles that define major categories. For example, list a few anime in 90-100, 80-90, 70-80 range. Then assign a point and an interval score for each film. Iterate a few times comparing each title with ALL others. For example, if you assigned a score of 93 for one anime, is it really better than ALL anime with a score below it on your list? If that is not the case, adjust your scores, possibly for several anime.

This way you are continuously calibrating your rating scale and the interval captures your uncertainty about the rating that may capture your true feeling about an anime over the years, but some big fluctuations are possible in transitional periods in life. It is possible to “outgrow” some anime for example. Whatever happens though, you can always update your scores at any time.

What do you really want and why?

We stated some obvious things in this article, but also highlighted some aspects that may help you understand yourself and others better. Moreover, we can go deeper and even question the very process of consumption itself.

Why are you watching anime? Do you enjoy the process? Think of watching anime or any other activity as an investment of time. Do you get any benefit from this investment or is this just a maladaptive coping with the issues you have?

Now, this does not mean that we have to apply the same mentality for everything in life. There are many things that we do because it is the right thing to do, regardless of a potential benefit and not expecting anything in return, but being mindful of the time we have left, and how to spend this time to make the world a better place or reach whatever goals you have set for yourself, is important.

Time is all

Further Reading

THEM Anime Reviews (one of the oldest and best review websites)
The Secretary Problem. An algorithm for deciding who to marry, and other tough choices.
Your Taste is Bad and So Are You


5 thoughts on “Why Belldandy, Skuld, Urd, and Hild Can (Not) Agree on Anime Taste

  1. Makes me realize that I don’t really wanna write reviews, because other people can do it better than me. (Although I had that desire to write reviews to just develop my own evaluating skills on anime, and other medium I happen to consume as well.)

    1. The more reviews, the better. The only thing to keep in mind is what type of a review you are writing, and what are your objectives. Everyone can write a general impression detailing your thoughts, reaction, and some observations. Also, you may uncover an interesting perspective or interpretation that other people may never thought of. Or some people have a similar taste as you, so your opinion is very important to them. Thus, do write reviews if you want! ๐Ÿ™‚

      1. I’ve been thinking, though. If I ever want to write reviews, I want them to help people decide whether or not they’ll like the anime. I’ve been thinking a lot these days how I will structure them, just so all my ‘future’ reviews will be more or less consistent. But before that, I still have to refine my writing. Anyway, thanks Kitsu-nee! You sure know how to make me reconsider stuff.

        1. Don’t feel too much pressure to set and keep a format from the very beginning! It is not some kind of long-term expensive project that you can not change! Usually ,it is a good idea, but things evolve and change, so you may adjust your format – it is perfectly fine! ๐Ÿ™‚

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s