@Cetonis: Well, I have never had any intention to use a strict mathematical model, as it
A. Is a pain in the ass
B. is absurd for reasons I mention below in Vivi's part.
That said, you actually got me to thinking over how close to an actual normal distribution my scores were (yeah, I'm a nerd). Using the 68% rule of thumb, I estimated the sd of my scores to be about 1.5 (i.e. about 70% of my scores are higher than 5 and lower than 8). Given that, 10, which is 2.33 sd away from the mean, should appear about 1% of the time. Given that I've completed 290 shows, I can expect about 2.9 shows to be a 10. So having one show as my 10 is not that big of an error - at least not yet.
Moreover, it's nice to have your "1" and "10" be clearly defined as the best and worst show.
@Vivi:
Quote:
|
VIVA: Bell curve, eh? Based on the idea that most anime is average. Interesting, but that might only work if 1) the distribution of good to bad anime ACTUALLY is that way e.g. the people in the industry are like 'oh yes, ****, we've made too many good anime this year, let's make more bad ones to keep the distribution in a bell curve', and 2) if you have watched a wide selection of anime sprinkled representatively across the bad-good spectum.
|
Well, the good news is I've figured out a place to give you general feedback about your reviews. The bad news is I'm about to rant for a while about a topic that I've put a lot of thought into over the years.
For one thing, I don't think the scores I give have
any bearing on the quality of the anime itself. As I've said many times before, objectively measuring artistic quality is pointless. Rather, the scores represent my enjoyment of the anime, which is an entirely different matter altogether. So no, I don't believe that anime makers purposefully make good anime to balance out bad ones (although, whenever Oscar season comes around, it certainly seems that Hollywood is doing it :P).
So do I believe that my enjoyment follows a normal distribution? No, not at all. Enjoyment does not actually follow any distribution. It is not quantifiable. If you ask someone how much they enjoyed their day, they can't spit out a number. They'll just say something general like "it was good."
And yet, we find ourselves assigning numbers to anime enjoyment. In some ways this is absurd, as we find ourselves having discussions like these where there is no clear consensus on what a "7" actually means. The simple truth is, a single score by itself is meaningless - the score only gains value when compared to other scores. Because we are assigning numbers to something that is not quantifiable, EVERY scoring system is arbitrary.
Thus, the importance is not "accuracy" as you seem to believe because that is impossible anyway. Rather, the most important thing is to create an arbitrary system that gives people a clear view of your opinions.
That is why I try to have a normal-ish distribution for my scores - so that the anime I thought were especially horrible or especially good stand out from the rest. More specifically, if I think two anime are "ok," it matters much less whether I think one is better than the other, as they're both largely forgettable. However, if I actively like two anime, it matters very much if think one is "good" while the other is "great." It matters even more if I think one is "great" while the other is "amazing."
That's why scoring everything as an "8" is so bad. If Maelstrom hypothetically makes a bunch of reviews and someone tries to get recommendations from him, then the person will have no idea what to watch from the scores alone. He'll have to read the reviews to get any idea.
I also don't really buy the "I only try to watch good series" bit. For one thing, that's pretty much what I do, with a few exceptions (eww Kanon 2006). Moreover, for the series you've reviewed so far that I've seen, my anidb scores have been:
Beck - 7.5
Black Jack - The Movie - 7.5
Blood The Last Vampire - 6.5
Bokusatsu Tenshi Dokuro-chan - 7.0
Code Geass: Lelouch of the Rebellion - 6.0 (recently raised score, I decided that my opinion was swayed too much by the final episodes)
Escaflowne - 6.5
Grave of the Fireflies - 9.5
Jyu Oh Sei - 6.0
Lovely Complex - 7.5
Major Season 1 - 7.5
Major Season 2 - 7.5
Metropolis - 7.5
Ninja Scroll - 4.5
Rose of Versailles - 7.0
Scrapped Princess - 6.5
Seirei no Moribito - 7.0
Terra e... TV - 6.5
X: The Movie - 2.0
Note that while I generally agree with you on whether a series is at least watchable, my scores are drastically lower than yours for every single series except 3 (Black Jack, Grave of the Fireflies and Metropolis). To me, this suggests that it is not your anime-viewing habits that are resulting in a different distribution of scores. Rather, it seems to be the result of your mindset that something you enjoy has to get at least an 8, but that
nothing deserves to get above a 9. For the reasons I've stated above I don't think this is a good ranking system.
