April 06, 2006

My friend (a 2nd generation Japanese) who has seen far more anime than all other people I know (and myself) combined, has come up with an heuristic for judging whether something is worth getting an actual viewing. There are far too many series around competing for attention, even more so for him since he can watch region 2 releases (no need for English). Time and resources are finite, so some sort of prioritization must be used. Basically it is this: Watch the opening and the closing.It doesn't always work, of course. The opening theme to Aishiteruze Baby was some appallingly dreary Suzanne Vega-ish thing, but I liked the show a great deal. And they can be terribly misleading (Narutaru, I'm looking at you). But it works - for me - far more often than not. And that's why I'm archiving them here. A reviewer can tell you that the animation is fluid (or stilted, as it may be); the music joyous or inspired or achingly beautiful or, well, none of the above. But in ninety seconds, you can determine that for yourself. * Most anime series that aren't adaptations of manga are adaptations of computer games. But they almost invariably suck, so I'll ignore them. ** Ranma ½ notwithstanding. *** And if you thought that the closing themes for Popotan or Happy Lesson were ear worms, just writing the name of Urusei Yatsura has got Lum no Love Song stuck in my head. Hoshitachi ga kagayaku yofuke
- If both are good, then the series is, more likely than not, also good and worth watching.
- If one is good but the other so-so, then the final opinion could go either way. (Openings are worth more points.)
- If both are bad, don't watch.
Yumemiru no anata no subete.
Aishite mo anata wa shiramburi de.
Imagoro wa dare ka ni muchuu. Oh yes, here, in case you were wondering.
Posted by: Pixy Misa at 10:07 PM | Comments (20) | Add Comment | Trackbacks (Suck)
Hey! don't be hating on Nerf Herder.
You apparently liked the UY opening/closing animations more than
I. Aside from "Rock the Planet", I don't think I've seen a UY
op/ed that didn't make me want to put a fist through drywall.
That horrible giggle from the first TV opening... shudder
Posted by: Mitch H. at April 07, 2006 01:20 AM (iTVQj)
Actually, I can find out pretty quickly. Where's my iPod? Right. Now w-a-i-t for it to boot. Oookay, the giggle is a bit much, but it's over with soon enough.
(The second anime I watched was Dominion: Tank Police, which was also great. The first one was a forgotten bit of fluff called Ultimate Teacher. This was pre-web, so I have an excuse.)
Posted by: Pixy Misa at April 07, 2006 01:37 AM (zZVLb)
I have a difference of opinion with Steven about the OP and ED of Banner of the Stars. I thought that the ED was great (as a music at least), and I have it on my iPod by J.Greely's method. But the funny thing is... it took me a few months to figure it out. It starts a little weak, so I always skipped it until some chance intervened. So, for me Banner is actually a matching hit for the heuristic.
Posted by: Pete Zaitcev at April 07, 2006 04:36 AM (9imyF)
I notice that no one is disagreeing with me about Card Captor Sakura.
The single most important difference between American episodic animationand Japanese episodic animation is that American shows are always intended to be of indefinite length, just in case they're hits and may get renewed for another season.
The majority of Japanese episodic shows are expected to be only a single season, if not a half season, and as such they can actually tell a real story over the course of the show. When that's handled well (e.g. Noir) you can have a completely riveting series where every single episode contributes significantly to the story being told.
Of course, not every series manages to do that, but the majority of them at least try to. There still are series which are 24-26 episodes long which really should have been 12 (e.g. Chobits) or which have a substantial number of filler episodes (e.g. El Hazard, The Lost World) or which are 12 episodes but only really had 6 episodes of story to tell (e.g. Popotan) but at least they usually try to tell a long term story.
I would propose a different heuristic: generally speaking, half-season series (12-14 eps) average much better than full season or two season series.
Of course, there will be exceptions. Card Captor Sakura ran 70 episodes and was excellent, and Happy Lesson TV was mediocre at best. But when I go back and look at it, the majority of the series I really like the most were half-season: Haibane Renmei, Someday's Dreamers, Banner of the Stars, Magic Users Club, Hand Maid May, Serial Experiments Lain. The half-season format seems to force directors into an economy of story telling which generally results in a tighter story. (Not always, though; Key, the Metal Idol got through its half season and hadn't even begun to tell its story yet.)
It's another heuristic, of course, and like all heuristics it's not perfect. (As the old joke goes, if a heuristic was never wrong it would be an algorithm.) Thisheuristic has an added benefit, though: if it turns out that the half-season series you bought did suck, you're out less money for fewer DVDs.
Posted by: Steven Den Beste at April 07, 2006 05:10 AM (+rSRq)
Posted by: Steven Den Beste at April 07, 2006 05:12 AM (+rSRq)
Posted by: Steven Den Beste at April 07, 2006 05:27 AM (+rSRq)
Posted by: Steven Den Beste at April 07, 2006 07:08 AM (+rSRq)
Let's see if I've fixed it...
Posted by: Pixy Misa at April 07, 2006 09:16 AM (zZVLb)
Posted by: Pixy Misa at April 07, 2006 09:16 AM (zZVLb)
Also, a lot of viewers nowadays like to hear a bit of the dub, though I don't much go for it myself. (Whaddaya want? I was/am a subtitler, I don't watch dubs for a living! ;p)
The real reason why the OP and ED garner a lot of attention on the Japanese production side is simple - it's work that only has to be done ONCE. Yes, yes, good opening and ending and it enhances the show and helps sell it and yadda yadda, but fundamentally, it's 3 minutes of animation that you don't have to do over again every friggin' week. Essentially, because it's going to be run 13 or 26 or god-knows-how-many times, it's very efficient to pour extra work into that bit of animation.
Heck, some recent shows (Chrono Crusade springs to mind) have gone to air the first few episodes before the "real" opening was animated - they ran on air with a montage of shots from the show, then went to DVD with the "real" opening. Heh, Japanese animation production is very much a just-in-time industry, and it occasionally gets them in trouble...
Posted by: Avatar at April 07, 2006 09:30 AM (mELpt)
It LOOKS like somesurreal conglomerate of clips thrown together over an oddly catchy song... but if you pay attention, it's a 90 second retelling of the ENTIRE SERIES! I didn't catch it until I watched the series for the second time, and then it surprised and delighted me.
Scary.
Posted by: Wonderduck at April 07, 2006 10:21 AM (7+BNY)
Chobits strikes me this way. The two of them are standing apart, alone. They bump into each other. They look. They regard each other for a moment. He is quite happy, but she gives no reaction.
Then there is this crazy pattern of digital-esque light around Chi's eye and she suddenly begins moving. She reaches out to him.
He accepted her, but she chose him.
Posted by: Shamus at April 07, 2006 12:00 PM (GDT1x)
Posted by: Steven Den Beste at April 07, 2006 02:29 PM (+rSRq)
Anyway, while I intuitively agree with Heuristic 2 (H.1 := "Size") and Heuristic 1 (H.1 := "compute_suck(OP+ED)"), they seem weak even in conjunction when examined closer. The data which I have is quite inconclusive.
I pulled 3 boxes full of DVDs from the closet and compiled the list of Like factor (corresponds to Rewatchability), H.1 factor, H.2 in eposode units. The table has movies (from same boxes) and non-DVD thrown in too (for illustrative purposes), sorted by R.
S R H.1 H.2
Azumanga 5 3.5 26
Sp.Away 5 - M
Haibane 4.5 4 13
[Windy T.] 4.5 5 13
Kamichu 4.5 4 16
-- Naruto 3.7 5 180
Banner 3.3 3.5 13
Porco 3 - M
Dai-Guard 3 4.5 26
Stellvia 3 5 26
Chobits 3 3 26
Tenchi OVA 2.5 3.5 13
Furuba 2.5 3.3 26
Stratos 4 2 4 13
Excel Saga 1 3 26
Jubei-chan 1 3 13
Ga.Angel 1 3 Many
Drag.Half 1 3 2
ROD 1 4 3
[FLCL] 1 3 6
I was unable to calculate sensible corelation values, because of normalization difficulties. In case of H.1, there's a lot of randomness and self-selection, because my shelf has better shows than store shelf. See, no ones or twoes. Basically I never saw a complete stinker of an opener. For H.2, the trouble is in the scale. I mean, I do not even have CCS (70) or Maison Ikokku (96). But Naruto pushes 180! The other trouble for H.2 is how a bunch of stinkers has very small number of episodes.
But even so, it's obvious that H.2 is weak, in the sense that it gives good guidance for 13 and 26 group, but not for any oddballs. The way to apply it is to tabulate: 13 is factor 4.5, 26 is 0.5, everything else is 2.5.
H.1 is weak because truly horribly OPs/EDs are rare, so it fluctuates wildly and thus throws many exceptions.
It was a good fun to think about it, but I'm afraid we may be better without either H.1 or H.2.
Posted by: Pete Zaitcev at April 08, 2006 07:31 PM (9imyF)
I don't consider quality/likeability to be the same as rewatchability.
I think Cowboy Bebop is a brilliant show - and I have no intention of rewatching it.
(I think Connie Willis's Doomsday Book is brilliant, and I have no intention of ever rereading that either.)
I do think we need a bigger sample set, though. I just unpacked one, but I don't know when I'll get to analyse it...
Posted by: Pixy Misa at April 08, 2006 07:52 PM (lLRbG)
Also, H.2 gives a good guidance indeed, I was mistaken calling it "weak". I got carried away with calculations which were futile. But if we simply split the table in three between Porco Rosso and Dai-Guard, and between Jubei-Chan and Galaxy Angel, the upper half is almost all 13s, middle part is almost all 26. Spooky, huh. Should've applied common sense earlier.
A bigger data set would be nice, but I'm lazy.
Posted by: Pete Zaitcev at April 09, 2006 05:46 AM (9imyF)
Posted by: Pixy Misa at April 09, 2006 12:13 PM (oyvZL)
Posted by: Caty Tota at July 08, 2006 05:05 AM (EfiW2)
Processing 0.01, elapsed 0.0106 seconds.
16 queries taking 0.0048 seconds, 40 records returned.
Page size 27 kb.
Powered by Minx 0.8 beta.