Shaman King is not currently available for streaming.
Anime adaptations of light novels and especially manga have a nasty inherent problem to them. Most are made with the intention of promoting sales of their source material – usually a still-ongoing series – which will inevitably have a slower release cycle than the week-to-week episodes of a TV show. Left to progress indefinitely, most of these adaptations will handily overtake the progress of the content they’re based on.
This is how you end up with the infamous poor pacing of shonen action shows like Dragon Ball Z and One Piece, which at times endeavor to stretch a 19-page manga chapter into as many minutes of screen time, or stall for time with consequence-light “filler” material while waiting for the original story to gain some ground. There are some cleaner ways of getting around this problem, such as the Monogatari series, JoJo’s Bizarre Adventure and My Hero Academia’s strategy of taking substantial breaks between seasons to better match the original author’s output. And if a series isn’t as concerned with having a long-form, over-arching plot, sometimes a show will just pick a convenient stopping point and leave the rest open-ended (saying in essence that Haruhi Suzumiya and Baccano aren’t getting another season, so just go read the novels).
But just as common is the “Gecko Ending” – the analogy being that much like a gecko cutting off its own tail and growing a new one, adaptations of ongoing stories are apt to diverge sharply in their third acts and invent their own original endings so as to create what should be a satisfying conclusion. The most famous anime-based example of this is possibly Fullmetal Alchemist, which was completely re-made near the manga’s conclusion to address the original series’ deviations. As of writing, we’re currently getting a high-profile example with the TV adaptation of A Song of Ice and Fire, with reception of the two “off-book” seasons appearing more mixed and a floating question of how deeply the remaining novels will diverge from the show despite the author’s involvement in both.
Shaman King makes for a unique example in that the series has three endings: one for the show, an “original” ending to the manga, and a “perfect” ending.
Naturally, since we’re talking about the ending of the series, there will be some light and very broad spoilers on the other side of the image:
There’s a common conception that when a piece of fiction and its adaptation diverge, the source material tends to weigh more heavily as the “correct” (or canon) ending, while the adaptation-original elements are considered as a secondary or alternate telling of that scenario. This is in no small part because of that tired (and lightly gate-keep-y) maxim “the book is always better”, which is backed by the fact that the majority of Western film and TV adaptations are of already-finished works, providing a pre-existing, concrete, and “intended” ending that the adaptation has to live up to.
Manga and light novel adaptations tend to work in the opposite direction, making surrogate endings necessary rather than elective, and frequently leaving screenwriters to come up with a continuation in the original story’s voice, but without direct control from the story’s original author. Naturally, getting the tone right is a tricky business here, and it’s one of the most common failings aside from issues of pacing.
Shaman King distinguishes itself here in that the first-run manga’s ending was not received especially well. It’s arguably not even what was planned by the author from the beginning, as the series’ premature cancellation resulted in him cutting the climactic confrontation short and wrapping up character arcs in what many considered to be a haphazard way. In its own weird fashion, the ending of the original work became divergent from itself with somewhat dissatisfying results.
The anime, on the other hand, finished two years prior with a more typical action-series finale, cashing in on a climactic battle and the heroic defeat of the Big Bad. It didn’t completely resolve the central conflict of the series by its own logic, but it still reads as much more satisfying and on-brand for a series that amounts to one very long Tournament Arc. This one seemed better-accepted despite its shortcomings, resulting in the uncommon case where viewers and watchers were split on whether the adaptation was a better representation of the source material.
Then, five years after both the manga and anime were complete, the manga was re-released with what the author clearly communicated as the “true ending”, expanding on the story and substituting the last few confusing chapters with a more conclusive and satisfying resolution that better matches the tone, pacing, and logic of the rest of the series. Of the three, this feels the most congruent with the story told so far, which is almost certainly why it’s the version most fans will point to.
So Shaman King is in a bit of luck here, as even in the meta-context of interviews, it’s been pretty clearly communicated to the readership that the second manga ending is intended to be the most “correct”, and that the series did well enough in the long term to even come back and tell that ending. But what about series where that isn’t the case?
Rebuild of Evangelion, for example, splits its story off quite strongly from the original series (which itself has two endings), all being produced by the same general creative team; the main reason we talk about End of Evangelion as the “true” ending is that it feels the most natural, despite director Hideaki Anno defending the series’ confusing final episodes. Nausicaä of the Valley of the Wind was written with the express purpose of being adapted into an animated film, yet goes on to conclude many years later and after a much longer, somewhat divergent story. And on the American side of the pond, you have series like Harry Potter where the books are considered to be the “default” version of the story, yet the prologue movies continue to expand on the films’ setting beyond the books’ scope; or the eternally-messy Star Wars, which explicitly has two separate canons (essentially, pre- and post-Disney-acquisition).
Having a degree of confidence and pride in its own direction certainly helped in this case – as did planning ahead. The TV adaptation for Shaman King saw itself running out of track and made the conscious decision to start splitting off early, slightly before the halfway point in its story (and before the one-third mark in the manga). The resulting space gave the writing team plenty of space to plan ahead and develop their version of the story in a way that led to a more natural conclusion. As it turns out, giving yourself over 35 episodes to steer toward your ending feels much smoother than taking a hard right in the last five (here’s lookin’ at you, Claymore). It’s that kind of thinking that may or may not pan out well for the ongoing A Song of Ice and Fire adaptation.
But that kind of foresight is a luxury that many shows don’t get. Some have their series renewals up in the air until the last minute, resulting in either a wide-open non-conclusion (such as Ranma ½’s “and the adventure continues” ending) or a situation where the finale of a story arc has to potentially double as the finale for that representations of the series (such as in both seasons of Dollhouse, or in both the anime and print iterations of Haruhi Suzumiya). And the alternative – planning ahead for an independent conclusion – seriously undercuts the chances of the publishers and producers commissioning a continuation, much less a full remake.
But having the confidence to let an adaptation build its own legs can give it the identity to stand as an accepted alternative, both in and out of the context of anime. You even get this with the current breed of comic-book-based films, which virtually never opt for wholesale adaptations of existing stories, but instead tell their own “inspired-by” tales. Build your own ending, and it can stand or fall on its own. Scramble to find something at the last minute, and it becomes clear that the ending is a last-minute solution.
The key, in the end, seems to be giving yourself space. It may lead somewhere quite different, of course, but different can be preferable.