ModelCitizen's Blog

Back to ModelCitizen's Blog

The philosophy of ratings

A rating one gives an anime is a stance about it precieved quality and serves information to anyone who looks at it by how a given anime compares to other within a specific genre or theme. Therefore it is necessary for it to indicate a difference as clear as it possibly can. This can't be aciheved if your mean score is well above 7.00, and half of your ratings are 8 or above, because you will send the message that there is miniscule difference between an anime you consider average or below and another you consider to be a true masterpiece. We are provided with a 1 to 10 scale to use, so it just makes sense to do it.

But what to be the 1/10 and 10/10? Thei ideal worst that failes at everything and not just has no value at all, but it was made with malicious intent and ideal best, which is otherworldly flawlass? But that is unachievable, so is it really reasonable to hold back ratings for those? It is naturally not, effort should be rewarded and striving for perfection encouraged. Which leaves me with two obvious choices. 1/10 being the worst thing that I have seen and on the other end, 10/10 being the best thing that I've seen. That leaves the possibility, that at some point I'll see something that underbeats the worst or outshines the best. The ratings can be rescaled accordingly.

Now, the next question that must be dealt with is how does this apply to different genres. Breaking up the scores by genres and saying one is a good comedy but shouldn't compare to a good drama because they are far too different is pointless. Saying that yes, I gave a 7 rating to both, but you still shouldn't expect to expect the same level of quality or enjoyment from them because one is an mecha-action and the other is a harem-ecchi is controvertial. If they are not on similar level of quality, than why the same score? Most anime falls into multiple genres, how do you differenciate? Rather no anime is made without intention. All of them has different artistic and business goals, everything is using or addressing themes that others do, everything sends a message. This can be taken into consideration. I don't think I shoulb be more strict or forgiving because one falls into genre than another. Focus can shift depending on genre, but that doesn't changes the fact that it must be internally consistent.


The mechanics of ratings

Since MAL only lets to rate an anime with a score from 1 to 10, each rating falls into a range around the given number. That range is defined by the rules by which one rounds up or down a decimal fraction to the closest integer. Meaning that something rated with 6 is somewhere in the 5.5 ≤ 'rating' < 6.5 range, while a 10 means everything from 9.5.

As for the scale, it has 3 basepoint. The 1.00, the 10.00 and as for the middle (1+10)÷2=5.5, which is the ideal mean, mode and median for the series of my ratings. But since it is impossible to rate with decimal fractions, it will be distributed between 5 and 6.

Thesis: Everything that is made has a level of quality and most of them is average with increasingly fewer that are below or above that. A good rating system must acknowledge this and the ratings will fall on a bell curve according to what was stated previously. This will provide an outline which ensures not to under or overrate by too much.That eliminates an element of subjectivity and gives a reliable level of ballance to the ratings. If you start to add high or low rated anime to your list, you have to rethink your ratings and take measures to bring it back into ballance by raising or lowering scores.




Increased precission ratings

Even if everything stated befor is being followed, as one's list grows, it'll keep getting harder to indicate the quality difference between any two given anime. Especially if those are fairly close to each other in quality/enjoyment. The number of anime linked to each score will be grow into the double and triple digit numbers and you can't give a dedicated score to every single piece you've seen and even if you could, from a point the difference would be so negligible that there is no point to increase the resolution of the rating scale.

That doesn't mean there is no point at all. That's why I manage a secondary account on another site, which gives me the opportunity to rate with half scores, doubling the resolution (sadly the only way for it to be visible, as far as I know, if you register there and choose advanced rating). The general rules are the same as on MAL, which means a rating of 5.5 is everything that falls between 5.25 ≤ 'rating' < 5.75 and a 10/10 means everything that is 9.75≤.



But it gives an additional gain. If you combine my Kitsu ratings with my MAL ratings, the resolution is further increased. And here is how. If I have a rating of 6 here, on MAL, it means I consider it to be somewhere between 5.5 and 6.5, while if I have a rating of 5.5 on the same anime on Kitsu, it means, I think it is between 5.25 and 5.75. The two range overlaps between 5.5 and 5.75 that eventually will mean that our anime falls in that much smaller range. This lets everyone to see if I think if it's just barely worth the score I gave it to or almost worth the better. This also helps to find out on a lot of occasion that if I rate two anime with the same score, which one I consider to be the better, effectively breaking up all of my MAL scores into 3 tiers (except for 1 and 10), weak, mid and strong, as I illustrated above.


Side note about my drops

As you can see, I drop a lot of anime compared to most anime watchers. Some says that if you start something, you have to finish it, because you can never know how good can it get, some others say it, you must finish what you have started, just because. For some, it's a mater of vanity to finish whatever, but how dumb it is to force yourself to watch something you have no interest in whatsoever or outright hate? Another misconception is that who drops a lot, don't know their taste. Probably the opposite is true. You can't cherrypick everything you will like. Sometimes something that seems to have everything for you to like turns out to be crap, sometimes something that looks bad turns out to be good. I'm willing to move out of my comfort zone and look for hidden gems among obscure or generally misunderstood anime, others just stick with what's popular and the few genres and tropes which appeal to them, then claim they know their taste despite they didn't do anything to discover it.

However I don't consider everything I drop outright bad though none of them are exceptional. I've seen a lot of anime, I've seen a lot of approach of the same subject. Since I look for new experiences, something I haven't seen, I won't stay to finish the 150th textbook example high school romantic comedy about an average boy and the girl who one day unexpectedly transfers into his class. I start it, because maybe it has an interesting twist or it might be executed exceptionally, but if not, I won't force myself to watch another 4 or 9 hours of it.