LastPerson Wrote:Assassinator Wrote:U got to be by far the most optimistic person I've ever met.
I'll post some ratings myself sometime soon.
seriously?
look over the series I've watched say theres one that's essentially not worth its rating D:
I would almost never give ratings like 9.8, unless it is pretty much perfect, flawless in every sence, and has pretty perfect storyline, animation, music, everything. And nothing is perfect and flawless, so nothing is likely to get the rating u gave, except for the one that got 8, (and the other ones that are close to 9, if they are really good.)
Look at Zinga and BluDevil's ratings, that'll give an idea about how i would rate. Maybe less sharsh than Zinga, but nowhere close to a mean of ~9.3.
Out of the ones u rated, I've watched Elfen Lied, Love Hina, Rumbling Hearts and part of Zero no Tsukaima. And out of these, Rumbling hearts (which u gave 8) would be the only 1 rating like something i would give. Elfen Lied has a really cool storyline, but gets dragged down by the enormous amount of flaws present. (had a 3+ page discussion with Zinga about all the flaws on email), Love hina isn't worth 9.5. Zero no Tsukaima isn't likely to be close to 9.8 from what I've seen so far.
To me, most stuff fall within "average" and "good" and that'll be something like between 5 and 8. There aren't many series i would classify "excellent/near perfect" (only a very few would ever get 9+), and some would be poo poo enough to get like 3.
ZiNgA BuRgA Wrote:Actually, your standard deviation is surprisingly large and mean surprisingly low. This is probably because of your outlayer (the 8 score). I do think you can even get a normalised score above 1...
I'm not trying to start anything, or diss your style of rating, but ZiNgA has a point.
Statistically speaking, ratings such as anime ratings should somewhat fit the normal distribution.
The mean is a measure of central tendancy. The standard deviation is a measure of dispertion. In other words, the means is the aveage, where most stuff lie, the standard deviation measures how much the ratings are spread. The best scores should sit around 2 standard deviations. Yours is less than 1.
Lets draw some pretty pictures to explain this ^.^
A normal distribution curve is usually something like this, the standard bell curve. With the average at the middle of the "bell", 1 standard deviation at the red markers, and 2 standard deviation at the blue. It does go beyond 2 standard deviations, but = signs are not the best things to represent smoth curves, so i can't show that. 1SD covers about 66% of all scores. 2SD covers around 95%. 3SD covers around 99.5% if i remember correctly.
= =
= =
= =
= =
= =
-----------------------------------------------
The nature of there being generally more anime above the median than below (mean=arithmetic average. median=middle score.) would make the standard theoretical anime curve be something like this:
=
= =
=
= =
=
= =
Zinga's current one is something like this:
= = = =
= =
=
= ===
And you're one is like this:
==
=
====================
Ok, I'm not saying everything has to fit the bell-curve exactly. But don't think yours is a bit extreme? That's not a curve, it's a cliff.
The purpose of having a rating out of 10 is so there could be a bit of spread, so u could really show the differentiation between the good from the bad. If everything most things are to be rated like 9.3, then that kinda defies the purpose of being out of 10. U may as well just minus 8 from everything, and leave it out of 2... Unless if u really are that optimistic, and think everything is perfect.
Don't mistake this as me not appreciating your work. Although i think you're ratings are a bit unrealistic, I do appreciate your work nevertheless. ^.^