The Game They Want to Watch

How the study worked

Six boards of twelve images each, one per dimension of the film’s production design: what Zelda should look like, what Link should look like, what Ganon should look like, which supporting heroes and villains should appear, and what styles should guide the movie. The images span every era of the franchise, official art, game captures and fan illustration, from the 1986 original to Tears of the Kingdom.

Thirty participants completed all six boards, out of 138 who opened the study, each placing green dots on the images they wanted to see in the film and red dots on the ones they did not. The platform sorts the result at a 71% threshold: an image is resonant when at least 71% of its dots were green, resistant when at least 71% were red, and contested in between. See the full report.

Every number on this page is a share of that image’s own dots, printed beside the share of the board’s attention it drew. Green and red are the study’s own encoding, so the verdict is written next to each score.

The finding: Ocarina of Time, grown up

Across all six boards, participants converge on one visual thesis. The film should look and feel like Ocarina of Time grew up: the characters, locations and narrative grammar of the N64 era, rendered in a painterly realism that reads as mature without tipping into grimdark. Not Breath of the Wild’s open-world ambiguity. Not Twilight Princess’s cinematic posturing. The specific story of Ocarina, painted like a Frazetta or a Ghibli background in oil.

Figure 1 · Six boards, one film

The convergent signal is not darkness versus light or realism versus stylisation. It is narrative specificity. The resonance arrays map almost perfectly onto a single Ocarina playthrough: young Link at the Deku Tree, adult Zelda before the Triforce, Ganondorf as the Gerudo King, Saria in the forest, Majora’s Mask as the sequel hook. Participants were not picking a vibe. They were casting a film.

Zelda: a woman who could address a war council

The highest-scoring Zelda is not from Ocarina at all. It is a Twilight Princess portrait at 92.31%, and the apparent contradiction reveals the real logic. What unites the four resonant Zeldas is not game era but visual maturity: detailed armour and ornament, composed posture, a palette of golds and warm earth tones. Every winning Zelda is rendered as a woman who carries political weight, not a girl on an adventure. The tiara with the red gem is constant across all of them.

The rejections tell the sharper story. Wind Waker’s Tetra is the single most rejected image in the entire study at 95% red, and it also drew the most attention on its board, 13.07% of every dot placed. Bright, rounded, chibi-proportioned, winking. Participants are saying this is not a children’s movie. The dragon-form Zelda from Breath of the Wild went 82.35% red: nominally from the most popular game, but abstract, non-human, and unrecognisable as Zelda without the lore.

The unspoken rule is a recognisability threshold. Zelda must look like Zelda: pointed ears visible, tiara present, regal bearing, human proportions detailed enough to exist in a live-action frame. Anything that pushes her toward cartoon, abstraction or childishness is rejected, however beloved its source game. The action-capable warrior Zelda with light arrows narrowly missed the bar at 68.75%, and the young Ocarina Zelda drew the most even split on the board at 57.14%, which is a warning about child-Zelda flashbacks.

Link: the moment before the hero

The only image in the entire study to take 100% green is a painterly young Link standing before the Deku Tree’s entrance, Navi glowing ahead of him. Not one participant placed a red dot. And the image is not heroic. His back is to the viewer. He is small against the forest. The palette is deep green and fog. The mood is uncertain, even lonely, and Navi’s glow is the only warmth in it.

Participants chose that over every battle-ready, sword-drawn alternative on the board. They do not want a Link who is already a hero. They want the moment before he becomes one. For a film, that is a gift: permission to start quiet and earn the spectacle. It is also the most Ocarina-specific image in the set, green tunic, Kokiri Sword, Navi, the Deku Tree’s cavern. The audience is not choosing a vibe. They are choosing a scene, and telling the production where the film should begin.

The rejections are categorical. The Hero’s Shade at 88.24% red and Dark Link at 86.67% red. Participants want Link, not anti-Link, however loud the fan discourse about a dark Link has been. And every official Breath of the Wild and Tears of the Kingdom Link landed in the contested middle, with the Zonai-armoured Link nearly resistant at 64.29% red. The franchise’s most commercially successful design splits the room when the question is a film.

The photorealistic Links sit tantalisingly close to the line at 62.5% and 66.67% without crossing it. Participants are intrigued by realistic rendering but not yet convinced by any execution they have seen. That is an opening for the production, not a reference.

Ganondorf: the man, not the pig

Ganondorf’s board produced the clearest consensus of any character. The top two images, a Twilight Princess portrait at 95.45% and the Tears of the Kingdom Demon King at 93.33%, are the study’s highest-scoring character images outside the unanimous young Link, and they share one feature: Ganondorf as a recognisably human antagonist. Green-grey skin, red hair, jewelled crown, yellow eyes, ornate Gerudo armour. A person you could cast an actor to play.

The other 100% in the study is the pig-form Ganon from the original Legend of Zelda, and it is total rejection. The painterly pig Ganon went 71.43% red. The Tears of the Kingdom dragon form went 85.71% red. Non-human transformations of main characters are categorically rejected for a film. Beast Ganon has been the franchise’s final boss since 1986; the audience is saying the climactic confrontation should be with the man. The one beast image that cleared the bar, at 76.47%, reads as a brief climactic event, not the form audiences spend time with.

The Wind Waker Ganondorf is the cautionary plate. It is one of the most critically praised characterisations in the series, a weary king who lost everything, and it sits at 70.59% red. Narrative complexity is welcome. The visual treatment still has to meet the mature, painterly standard the winners set.

The supporting cast: Saria, Marin, the Deku Tree, and not the Champions

The supporting-hero winners are strikingly intimate. No Champions. No warriors. No one with a weapon. Saria, a Kokiri child with a fairy, at 84.62%. Marin, the gentle islander from Link’s Awakening, at 80%. The Great Deku Tree, a paternal guardian, at 77.78%. All three are defined by their emotional relationship to Link rather than their usefulness in a fight.

Saria at the top of this board, combined with the unanimous young Link at the Deku Tree, is the study’s strongest single narrative signal. The audience wants the Kokiri Forest prologue: the childhood friendship, the bittersweet departure. Marin is not even from Ocarina, and her inclusion sharpens the point. Participants are choosing emotional specificity over lore accuracy.

Tingle is the only character to reach the resistance threshold, at 84.62% red, and the note extends past him: comic relief that breaks the tonal register is unwelcome. And every Breath of the Wild and Tears of the Kingdom character lands in the contested middle or below. Daruk at 55.56%, Rauru at 46.67%, Purah at 33.33%. Urbosa came closest at 70%, one point under the line. The ensemble from the franchise’s biggest-selling games does not belong in this film’s first chapter.

Villains and style: the mask, the crest and the Zonai problem

Supporting villains: a sequel hook and atmospheric threats

Majora’s Mask is the highest-scoring villain at 86.96% and the most-attended-to item in the villain set, 13.86% of every dot on the board. It is not from Ocarina. It is the sequel’s antagonist, and its presence in the winners suggests participants are already thinking about franchise continuity. The mask reads as an artifact, not a character, which makes it a cinematic tease that costs nothing to set up. The Darknut at 80% and the Poe at 75% are enemy types, not named villains: atmospheric threats that belong in the world.

Vaati from The Minish Cap took 100% red. Villains from the franchise’s obscure entries have no place here. Zant, from an otherwise well-scoring Twilight Princess, went 80% red: the tall helmet, the insectoid posture and a convoluted plot read as too esoteric. Participants want villains who are immediately legible or iconically mysterious. Zant is neither.

Style: heraldry over technology

Neither style winner depicts a character, a scene or a place. Both are heraldic symbols: the elaborate Hyrule crest at 87.5%, gold on dark green, ringed with lore, almost an illuminated manuscript; and the Sheikah Eye at 80%, the one Breath of the Wild element that consistently scores well because it reads as rune and glyph, part of Hyrule’s deep history. Participants want a film that feels like it has its own mythology, predating the story being told.

The Divine Beast symbols went 90.91% red. Flat, graphic, closer to corporate icons than mythological artifacts. Participants want the mystical Sheikah, not the technological Sheikah. The Wind Waker tapestry, one of the most accomplished pieces of visual storytelling in the franchise, went 75% red for the same reason Tetra did: too flat, too naive, too childlike for a film. And the Zonai aesthetic, Tears of the Kingdom’s defining visual contribution, was the most-attended-to image on the style board and 70% red.

The distinction the production needs: a glowing rune carved in stone reads as mythological. A glowing circuit board reads as science fiction. This world should run on belief, sacrifice and song, not on machines.

The whole study in two images

The two unanimous verdicts on the character boards are the study in miniature. One is a boy who has not become anything yet. The other is a monster the franchise has kept for forty years.

One is the moment before heroism. The other is a creature. The audience wants the man behind Ganon and the boy before Link, and it wants both painted, not rendered.

The decision logic translates directly into production notes, and the report sets them out as a synthesis rather than a single-source adaptation: Ocarina’s narrative, Twilight Princess’s visual maturity, Breath of the Wild’s sense of scale, and none of its story.

One caution from the report. Thirty completions from 138 opens is a 21.7% completion rate, and at n=30 the findings are emerging patterns, not verdicts. The convergence across six independent boards is what gives them weight; individual scores near the 71% line should be validated with a larger sample before a production decision hangs on them. The mechanism itself, an audience sorting into a specific answer it could never have written as a brief, is the same one that ran through the orange juice study and the make-it-pop study.

Frequently asked questions