Toothless, Intelligence, and Agency: Did Toothless Willingly Betray Green Death?

Nine years after How To Train Your Dragon was released, people still like to argue about just how intelligent Toothless (or other draconids of the franchise) really are. Some people argue that they’re intelligent and sapient, other argue against with mostly dumb excuses (the most retarded ones to date are vague fog of “but they aren’t human” and “they don’t believe in gods”).

One of the big ticket items in discussions about how intelligent draconids are — especially Toothless — is the ‘Astrid goes for a spin’ scene, in which Toothless ignores Hiccup and proceeds to give Astrid the Guantanamo Bay package. Right after that, there’s also the incident where Toothless takes Hiccup and Astrid to see Green Death1And yes, Dreamworks. Green. That last minute script change is a load of bullshit. Expense some Enchroma glasses on Dreamworks’ credit card.. The rationale for Toothless’ intelligence in that scene goes like this: When Astrid mentions that Hiccup will “have to kill a dragon,” Toothless gets a very bright idea. He understands what Astrid is saying. He understands the repercussions of what Astrid is saying. So he raises his ears, listens for Green Death’s noises and decides to submit his wishlist item. Green Death must die.

But of course, that point gets contested really quick.

I’ve never put in so much effort into making a reference to a game I neither own nor play. Fuck Overwatch, even, honestly — but that’s a rant for another time.

The two main counters to this theory are “no, Toothless was under the influence — or flat out mind controlled — by Green Death, he doesn’t do that out of his own free will” and “dragons aren’t humans, therefore they aren’t smart enough to do that.” Some may also object with “Toothless got alert before Astrid mentioned killing a dragon,” but we’ll come to that later. Let’s debunk the mind control issue first.

Green Death and Mind Control

Some people really like to point out that Green Death has some sort of control over the draconids, that Green Death can magically make dragons and wyverns return to the nest. Why else would anyone return to a home like that if they weren’t charmed?

The screenshot that proves that Green Death can mind control every dragon and wyvern nearby.

The oft-cited evidence in favour of some sort of mind control is that Toothless’ eyes narrow when he starts hearing the signal, but that’s really not what happens.

First of all: mind control is invention of the second movie. I’m somewhat reluctant about explaining things in the first movie with (or accepting explanations from) the subsequent movies. This is mostly because the second and third movie haven’t even made it to the idea stage when the first one was being produced. There was no plans — let alone specific ones — about what will transpire in sequels, which means that forward compatibility is not guaranteed at all. This means that arguments based on ‘this happened in the sequel, therefore it must hold true for the prequel’ kind of thinking are sorta invalid.

There is one trick that you can use to get around that, though. The sequels generally don’t contradict the original movies (although How To Train Your Dragon: The Hidden World, contradicting the first two movies as well as itself, is a massive exception in that regard). This has one neat effect that we can exploit. Since the no-contradiction rule implies that if something exists in both the sequel and the original, it will work exactly the same way in both movies. With the case of How To Train Your Dragon, theorized mind control does not behave the same way as confirmed mind control in the sequel, therefore we can be sure it didn’t happen.

And we can see that rather clearly:

Mind control from the second movie causes different kind of eye narrowing than „»‘mind control’«“ in the first movie.

Note that if the result was the opposite — if mind control in the second movie exhibited similar properties as something in the first one — that still wouldn’t be a sure confirmation. As it turns out, it’s entirely possible that two different things share the appearance.

There’s further evidence against outright mind control that you can find using this so-called “rule of consistency.” In HTTYD2, Toothless pretty much can’t do anything while being mind-controlled. In HTTYD1, however, Toothless willfully and deliberately ignores Hiccup’s plea to turn back:

But if that’s still not enough to convince you, consider this: during the final confrontation, it takes a grand total of three rocks2Actually, the real number is probably around a dozen or so, but by not exaggerating I would lose some of the entertainment factor and one glowball to make every dragon and wyvern fly away instead of fight. This right here is the clearest evidence against the mind control.

The Nature of Green Death’s Signal

“So what is Green Death’s signal if not mind control, then?” one might ask, “and why do all dragons and wyverns follow it?”

No, it’s not any sort of hypnosis or charm or brainwash either. If the presence of the signal would be forcing Toothless to follow it, he wouldn’t have had the capacity to show any sort of resistance to Hiccup. Toothless wouldn’t reject Hiccup’s plea, he’d ignore it without a reaction.

Green Death’s signal is a beacon.

Nothing more, nothing less. It’s a signal that shows dragons the way home and not much more. Green Death obviously doesn’t treat her draconids too well, but if human victims of domestic violence can rationalize returning home to their abusers for unreasonably long amount of times, so can draconids.

Furthermore, it is entirely possible that dragons find safety in the nest. Sure, they might get eaten by the Green Death — that’s surely a con — but the nest provides many benefits. It’s a damn good shelter, for one, and contrary to what How To Train Your Dragon: The Hidden World and Ensiferum would have you believe, there is, in fact, strength in numbers. Sticking together makes dragons and wyverns safer from human raids — doubly so when you’ve got a dragon the size of a mountain defend the nest against the invaders, be it vikings or other dragon nests looking for new territory.

At this point it’s worth noting that the first movie never explicitly confirms or denies existence of other nests. However, much like there’s not just one anthill, it’s reasonable to expect that there are other nests like the one near Berk that exist in the world.

How To Train Your Dragon 2 seems to mildly agree with this notion by introducing us to Valka’s sanctuary and Drago’s army, both of which can be classified as a nest of sorts; but we said we’re not paying much worth on “evidence” from the second movie.

Meanwhile, if the draconids opted to live outside their nest, they would be vulnerable against vikings and against dragons that weren’t part of their group. Given that Vikings have supposedly killed ‘thousands of them,’ I reckon staying with the nest is a good idea.

Obviously, the movie ever so slightly contradict this point by draconids wussing out as soon as Vikings break out the superior siege engines, but that’s not the problem with this theory. That’s a minor plot issue/self-contradiction on the part of the first movie: namely, we know that dragons and wyverns will defend the island against vikings in the rocky labyrinth surrounding it. Why not once the vikings land?

Congratulations! You found pretty much the only problem the original movie has.

Actually, it’s not the only one, there’s a second problem that relates to Vikings just blasting the mountain with three rocks. But the point is this: you can trust that proponents of The Hidden World are just gonna take those two problems and throw them around with the zealotry of an anti-vaxx mom who found this one google hit agreeing that vaccines cause autism. Likewise, you can trust that they’ll start to engage in some blatant both-sides-ism, pretending that having two minor problems is the same as having two hundred (with twenty of those pretty major).

The problems with dragons wussing out can be partially handwaved away as “they’re just moving out of the way of the Green Death and allowing it to do it’s job” — but hey, we’ve derailed this post for long enough now.

The point is this: Green Death signal is a beacon, not mind control. Dragons and wyverns live at the nests because they want to. They live there because the nest is a net benefit, not because Green Death would want them to. Believe me: if you had to pick between a shit, run-down apartment and living under a bridge… Let’s just say that you wouldn’t pick a bridge, either.

I Looked Him in the Eyes and Saw …

Yes, I know. The ‘mind control’ counter-theory is pretty much dead by now. But we’ve got further ‘evidence’ that Toothless’ eyes and face don’t indicate mind control. Toothless has had such eyes at various points in the movie. Let’s see some screenshots.

My hosting provider is gonna kill me if I spam all of these as a separate pic, so here’s a collage instead. You can sorta point out the ‘angry’ one and a half out of the collage, as well.

So yeah: it turns out that Toothless’ eyes turn narrow every time he’s scared, guarded or angry.

‘Scared’ sounds just about right.

Long story short: mind control doesn’t seem likely.

“But Toothless isn’t Smart Enough”

I’ve seen some people try to argue that Toothless is closer to being a Jerry than he is to being a Rick (or even at least a Morty) … But that doesn’t hold up all that well, either. There are at least two scenes that require near- or even human-level intelligence.

The Case of Forbidden Friendship

Consider, for a moment, Forbidden Friendship, though this train probably starts some time before that.

This not the first time Hiccup and Toothless meet, but it’s the first time that Toothless notices that Hiccup is a bit of a nosey fucker.

In The Forbidden Friendship, Hiccup initially tries to buy Toothless’ friendship with a measly offering of one (1) fish. Throughout the scene, Hiccup tries to get closer to Toothless, but the dragon isn’t quite thrilled with the prospect. And not only thrilled: he was outright annoyed. But Hiccup stubbornly persisted, until this happens:

Bae caught me trippin’

At this point, Toothless knows that Hiccup is very interested in him. He is also perceptive to notice that of all places that Hiccup could choose unnoticed, he went straight for the tail.

Hmm, interesting, thought Toothless to himself as he noticed the young lad interested in his tail specifically. Thus, he decided to check what exactly the little annoyance is looking for.

He looked and looked, and soon discovered the problem: half his tail was gone.

In all fairness, though, Toothless probably already knew that half his tail is missing by that point. He not knowing that is a very hard sell at this point in the movie. It’s much more likely that Toothless knew that half his tail is missing, but:

  1. he didn’t make the connection between missing tail and his inability to fly until he examined the tail again, or
  2. he made the connection, but hoped he will manage to overcome it, or
  3. he made the connection, but hoped the tail will grow back
  4. he made the connection, knew he can’t overcome it, knew the tail won’t grow back, but he just wondered what the hell does Hiccup see on his tail

The last three seem to make decent amount of sense. After all, the power of denial is immense. It’s not obvious which of the three options is correct, but it’s pretty obvious that this is when Toothless realized that his hopes are unrealistic (assuming b or c are correct) and that there must be a reason for Hiccup’s interest (any option). If you think that the latter option doesn’t require human-level intelligence, just let me remind you that there’s many people who wouldn’t be able to deduce that.

It’s also pretty obvious that during that while inspecting his tail, he realized that — like it or not — Hiccup is the only chance for survival he has.

And just like that, Toothless goes and tries to do something to ensure Hiccup will keep bringing him food, maybe even something extra.

“Hey human, I’m sorry for that fit earlier … pay attention pls?”
“I dunno,” Hiccup replies, “maybe if you challenge me to a draw-off.”

The Case of Spinning Astrid

The second, even better example, is the Astrid Goes for a Spin scene and the events leading to it. To refresh your memory, that’s the point in the movie where Astrid discovers that Toothless exists and then runs off to warn the village.

Fast forward your memory to this point in the movie. I’m aware that this is technically not the ‘Astrid goes for a spin’ scene, but let’s avoid being overly nitpicky.
Yes I recognize the irony of me asking you to not be nitpicky on this post.

Toothless seemingly doesn’t think Astrid running away is a big issue at first, but after some encouragement from Hiccup, he does catch himself an Astrid. This tells us that again, Toothless must understand human language, because that’s the only way he could have known that Hiccup wants him to catch Astrid and coerce her into not tattiling. It’s also very likely (although movie doesn’t provide solid evidence in favour of this) that Toothless actually understood the ramifications of Astrid tattling on them after a brief off-screen chatter with Hiccup.

In any case, Toothless manages to catch Astrid. Hiccup convinces Astrid to give them a chance. Hiccup tells Toothless to give Astrid a gentle ride to the ground. And Toothless …

Hiccup: “This bad boy can fit so many fucking fish in him.”

… Toothless wilfully ignores Hiccup’s words.

Yes, Toothless waterboarded Astrid because he wanted to. It goes against what Hiccup had told him to do, and it ends when Astrid apologizes. While her apology was nothing too specific (🔊 note — this webm has sound):

Excuse the contrast/saturation issues, gfycat is messing that up for some reason.

we can deduce that Toothless very, very likely didn’t go against Hiccup’s words to make Astrid apologize for the axe thingy, but did his best to coerce into not saying a word to the tribe. This is suggested by the lack of Toothless’ eagerness to chase Astrid initially: Toothless seems to see Astrid taking off as a victory enough. She came at him with axe, she ran away, job   d o n e:

Notice how Toothless doesn’t seem to be particularly interested into preventing Astrid from reaching the village at this point.

This suggests that Hiccup did or said something off-screen which made Toothless realize the true danger of Astrid tattling on the tribe — an obvious sign of intelligence. Then, he ignored Toothless in favour of his own plan to keep Astrid quiet, which likewise indicates intelligence (and not just intelligence: it’s clear that Toothless both has his own agency and can plan at least to some extent).

All in all, preliminary data shows that Toothless is indeed smart enough in order to plan the entire thing.

What does the movie say?

Now that we know that Toothless is smart enough to have reacted to Astrid’s words and that mind control bit is a bunch of bullshit, that surely confirms the theory that Hiccup did, in fact, react to Astrid?

Well, no.

The movie makes that pretty clear (🔊):

So what happened?

Let’s take a closer look. When Toothless hears the Green Death’s beacon, he becomes afraid. And that’s probably exactly why Toothless took Hiccup to see her. Because he recognizes the danger she presents to Berk, and because he’s very concerned for Hiccup’s well being (if you need evidence that Toothless cares for Hiccup’s well-being, see his Leeroy Jenkins impression during Hiccup’s final exam the next day).

Toothless wasn’t using Hiccup as his personal Death Note. The trip to Green Death was a warning.

And when Toothless ignores Hiccup’s plea to return back to Berk here:

 

… he doesn’t do that because he was under Green Death’s influence. He refused because he deemed it important for Hiccup to see that.

Whether this means Toothless betrayed Green Death or not depends on your definition of betrayal.

TL;DR:

  • Toothless does seem to exibit near- or even human-level intelligence
  • Toothless wasn’t mind-controlled by Green Death into visiting the nest
  • Toothless likewise didn’t show Astrid and Hiccup the nest because he wanted to submit his suggestion to Astrid’s “you’ll have to kill a dragon.”
  • Toothless showed Hiccup and Astrid the way because he knew that Green Death was a danger to Berk and wanted to warn them.

Did Toothless Really Have No Choice?

Sorry, but Toothless definitely had a choice, no matter how you slice it.

So there’s been this tumblr blog that got shared on one of my discords. For those who don’t want to follow, here’s a TL;DR.

Remember this scene from How To Train Your Dragon 3, where Toothless basically orders dragons into cages?

  • OP got their jimmies rustled by people saying Toothless shouldn’t have pretended to be French
  • And in the follow-up post, they basically stated that Toothless had no other options but to tell dragons to back off even if you ignore the Thotfury issue.

So let’s take a look at why this is wrong, no matter how you slice it.

If Grimmel Killed the Light Fury

… then he wouldn’t live for long. He wouldn’t even live for short. He’d immediately get blasted apart by every single draconid on Berk.

Contrary to the popular belief, the weapon specimen that Grimmel wields in that scene is not an M4A1. It’s a standard-issue medieval ballista. Its fire rate is much, much lower than that of M4A1 and reload times are pretty awful to the point getting more than one shot out of ballista in any given minute would likely be a massive achievement. In any case, once Grimmel fired his first shot, dragons and wyverns would have forever to blast him.

Obviously, we’re going to face a minor issue here. First of all, you can’t damage the dragon-powered quadcopter Grimmel flies on. That problem is easy enough to solve (just grab him and/or knock him off his stand). The other problem is that precision-blasting Grimmel off the quadcopter leaves the dragon-quadcopter without its pilot. This is a slightly awkward situation, because there’s no telling how deathgrippers would react. However, all the possible reactions generally fall into three groups:

  1. They hover in place
  2. They continue moving towards the ships at the unchanged pace
  3. Panic like my XCOM2 squad the moment they catch a whiff of aliens.

Option 1 is ideal and makes rescuing Toothless easy. Option 2 is a bit worse, but still ideal. There’s six death grippers (four on quadcopter + two solo) facing Toothless’s entire army, which means that Toothless would be returned to the island before he made it halfway to the ships.

Option 3 is where things get a bit more problematic. Panicked death grippers translate into problems for the quadcopter. Best case, quadcopter will become unstable. Worst case, death grippers end up knocking themselves out and freefall to their death. The chance of anyone falling to their death is still borderline negligible in this case — if four death grippers were enough to hold the quadcopter in the air, dozens of Berk’s dragons shouldn’t have a problem holding up the damn thing in the air.

Here we can see that the quadcopter isn’t hovering above the land at this point — it’s decisively over the cliff edge. But keep in mind that New Berk is three kilometers tall (from sea to the lake). If the quadcopter starts losing altitude, team Berk has lots of time to catch it.

If Thotfury gets killed, then it’s game over for Grimmel while team Toothless wins with zero additional causalities.

And if you wondered which of the three options is going to happen if you remove Grimmel from his seat without damaging the quadcopter, here’s your answer. Death grippers continue hovering in place.

The quadcopter only started crashing once Toothless dropped his mixtape.

Could Grimmel be Disposed of Before He Gets a Chance to Kill Thotfury?

Let’s just put things this way. Grimmel doesn’t have eyes on his ass. Have some dragons go around and attack him from the back.

The obvious issue with this plan is that two solo-flying death grippers could notice them and alert Grimmel. That still doesn’t mean Grimmel would get a chance to kill the light fury. The moment he turns around to check what the ruckus behind him is the moment he’s not aiming for the light fury. And the moment he’s not aiming at the light fury (or just being generally distracted) is an opportunity to get rid of him without thotfury getting it.

Would Grimmel Kill the Light Fury

Given everything said so far: lol.

Of course he wouldn’t. Not only would he no longer have a bargaining token for Toothless — once Light Fury is dead, there’s literally no reason for draconids to take it easy on Grimmel; and as first section shows us: Grimmel would get absolutely annihilated in this case.

His threats were empty.

Coulda called his bluff with little consequences.

“And the Flock was Never in Danger”

One of the reblogs of this post states:

And the flock wasnt really in danger. The hunters left the cages unlocked bc they thought Hiccup wasnt coming, and Toothless obviously knew he would bc he then immediately commanded the three like, second in command dragons to break out and go fight

What? That’s patently false.

First of all, the trappers didn’t leave the cages unlocked because they thought Hiccup wasn’t coming. They left them unlocked because “plot convenience.”

If you take a closer look at the cages, there’s two things one can notice.

  1. All cages come equipped with bolt locks and nothing else. Nobody is using padlocks or keys, because why would you? Literally no benefit. Keylocks are mighty expensive and time consuming to manufacture and offer no extra security when you’re trying to keep a dragon in the cage.
  2. Doorhandles are also nowhere to be seen, which means that if the door is not locked with the deadbolt lock, it can be easily opened from the inside of the cell.

Let’s take a look at some screencaps:

No keyholes, no padlocks.
95% chance this is just a regular deadbolt lock.
Hey look, it’s deadbolt locks.

The second thing worth mentioning is that trappers were, in fact, locking the cages. However, since there was more cages than there was trappers, the trappers couldn’t lock everything at once. I was a bit unfair when I called it a ‘plot convenience’ earlier, I know.

Trappers in the process of closing and locking Cloudjumper’s cage.

Just because the trappers haven’t managed to lock all the cages in 60 seconds flat, that doesn’t mean they haven’t been locking them.

Snotlout is unlocking a locked cage.

So that statement is a bunch of ballooney. But then again, this second reblog is so dumb that it barely warrants a response.

Anyway, that’s it for the night.

Safety of New Berk and The Hidden World: an analysis

If shadiversity can do a whole series analyzing defendability of various locations from Game of Thrones, I reckon I can do one for How To Train Your Dragon 3 as well.

If Shadiversity can do a whole series analysing the defendability of various locations from Game of Thrones … boy, let’s do one for HTTYD3 as well.

Fair warning: I’ve already written this post on reddit, but never really cared to port it over to my blog as well.

The warning about excessive musical references still applies here.

Just surrender here tonight: Safety of Berk

By back-of-the-envelope calculations, the settlement of New Berk lies about 3 kilometers above the sea level — and we aren’t talking about 3 kilometers of a gentle Sunday walk slope. We’re talking straight up, vertical wall. The writers stated that Minecraft Chunk Error is accessible only with the help of dragons, and everything — from the visuals to the height estimates — confirm that.

Back in the medieval ages, height advantage was everything. If a hostile army came to the doorstep of your castle and your castle wasn’t high enough, you’d be in danger of getting flattened by a 90 kilogram rock fired from over 300 meters away when busy with your evening number two1Note: this is not how Erasmus of Lueg actually died, but it does make for a more entertaining story.. New Berk, however, is far too high for trebuchets to reach — or even cannon balls — to reach it, let alone do significant damage.

99 Luftballons

Conquering Berk only becomes technically possible with the advent of aircraft in the late 18th century as even early hot air balloons were reportedly capable of floating at around or even over two kilometers above ground — even though their relatively slow speed and susceptibility to adverse weather doesn’t make them too feasible. In practice, conquering the New Berk only becomes realistic prospect in the 20th century with the invention of an airplane, and subsequent improvements thereof.

Alternatively, you could just conquer New Berk with dragons and wyverns that you’ve captured ahead of time. The trappers still have them. Just saying.

Trapped on an island, lost at sea

A man has need, but those needs aren’t just booze. It’s food as well.

Impenetrable geography means very little if the hostile force can just starve you out to death. On the first glance, this seems likely: encounter with Grimmel’s armada certainly put New Berk on the map. With likely rumors that Berk is hiding dragons for themselves, they could easily send another armada that would surround the island, capture every inbound or outbound ship and keep an eye on the island. With no dragons to help them out and carry them over blockade, the vikings can only wait on the island and hope that their food supplies last longer than invaders’ patience.

But it turns out that Berk may have lucked out in this aspect as well. The island is pretty big — and although there doesn’t seem to be much farmland there as most of the island is either a lake or steep slopes, the situation is not completely hopeless. After all, if Inca could figure it out, then Berk should be able to as well.

That is, of course, assumes Berk doesn’t mostly or majorly rely on fish for their food as in that case, their chances of siege survival would dwindle to single digits. This isn’t as unlikely as one would think: while the lifts are very slow the way the movie presents them, using a counterweight and water power from the nearby waterfalls would be a quick way to improve the lifting speeds massively — to the point a ship could be lifted in a reasonable amount of time that wouldn’t render fishing ineffective.

There is no strength in numbers

Have no such misconceptions..

I won’t cover why separation is bad from safety perspective because this was covered elsewhere in great detail already. Just giving a shout-out.

Now that the safety of Berk is dealt with, let’s move to take a closer look at The Hidden World.

Here at The End Of The World: Safety of The Hidden World

Alternative title: Through dungeons deep, and caverns old but I outright refuse to give that hot trash the honor of having a title feature not once, but twice.

The Drakkars were surprised

Let’s be honest for a moment: the water around The Hidden World entrance is very shallow, likely on the order of magnitude of a meter or two. The screencaps seem to confirm that:

The water here looks pretty shallow. No way you could dunk a full Hiccup into that stream. Some parts of the pic seem to even violate the physics in a very Cities: Skylines way, with water flowing uphill.
Judging by photos of sea shores, the water around the hole looks shallow enough for a man to stand in with no problems (assuming the water was still)

In all those close-up shots, you can see that the rocks poking above the water are more numerous than the rocks a disgruntled DM keeps behind their screen (and they’re bigger, too):

If you design a ship with enough draft and a hull that doesn’t roll once you hit those rocks, you’ve got yourself an island. Load the ship with rocks and you quickly discover that you don’t even care if the ship is seaworthy after hitting said rocks: you’ve just made yourself a small island and a mooring point for future ships.

This would allow the dragon trappers to just make a big ship and run it aground — be it on the rocks that poke out of the water or on the seafloor. Ensure additional anchoring if necessary. Congrats, you have a platform.

And when you took a look at how mighty the waterfall is at it’s bottom, saying that waterfalls are deep on the top becomes a very tough sell. We’re at the bottom of this hole where the waterfall is at its narrowest — yet there’s barely any water?

The waterfalls here amount to not much more than a drizzle.

Assuming actually competent villains, I don’t think anyone is in any danger of getting swept off here.

We lift together

The trappers won’t be making peace to build our future, though. That would be hiccup, and he just threw that chance away. Alternative title: Are you coming in so we can carry this on?

Once you get yourself a platform or a dozen on the edge of The Hidden World, you can start working on descending into the hole. This turns out to be fairly easy: we know that world tech level is sophisticated enough to build lifts if there’s a need. Not only has New Berk managed to achieve that feat — and to top it off, they seem to be using some special elven rope that’s both light and unbreakable. Powering said lifts would also be fairly easy, thanks to the near infinite water supply all around you.

Sure: The Hidden World entrance resembles an inverted wedding cake, but that’s not a problem: you can build multiple lifts: one to reach the first stage, another to reach the bottom. This will be mildly annoying thanks to all the water, but that’s nothing that can’t be worked around.

Judging by the movie stills, the hole is pretty shallow as well. During the shot where Astrid and Hiccup enter The Hidden World, you can see that the water behind them is lit directly by sunlight, which means that the hole probably can’t be deeper than a kilometer if we are really, really generous:

There’s approximately zero chance this isn’t sunlight on the upper portion of the waterfall.

But if you find the second stretch of the way down too challenging, you can do something you can’t do on New Berk (without getting noticed):

Welcome to my mine, we are mining dragons

The part about fighting mobs is up to discussion. Alternative title: I dig my hole, you build a wall

Brothers of the mine, rejoice!2I’d link the original version by Yogscast, but once you hear the Wind Rose cover the original one kinda sounds kinda meh. No offence, I know it’s hard to compete with people who sing for a living. Raise your pick and raise your voice! Since The Hidden World is completely surrounded by rock, you can dig a a dinky dinky hole3Jesus fucking Christ, is that a third Minecraft reference in a span of a paragraph? Tam, you need to put a damper on this shit. parallel to the tunnel. It’s gonna take a while, but it has some advantages:

  1. It’s safer. Dragons can’t get you there.
  2. You get free rock

You can use that free rock to build a dragon-proof fort (stone and water don’t burn, usually) or build a massive dike around the entrance so you don’t have this pesky water to deal with anymore. And when the water’s no longer a problem … well you could just build a couple more lifts, really.

Now we’re (not) defenseless in a land where dragons rule

You think dragons will just wreck anyone who comes near The Hidden World with such fury that doritos would start flying everywhere? Not gonna happen.

Next time someone tells you that US has the biggest navy in the world, show them this picture.

The armada has a massive numbers advantage. For every ship dragons would potentially destroy, three would take its place. They could easily surround The Hidden World to the point — if not because they’d run aground, they’d manage because they’d get stuck trying to fall in all at the same time. They will also have a height advantage by the virtue of entrance to The Hidden World being below the sea level. Once the trappers have made their base, they can use ballistas that shoot spears, bolas and nets in order to down anything that comes their way.

But it gets worse for dragons and wyverns: not only are they at a massive disadvantage both due to height and number of the attackers — they’re also being massively bottlenecked by the size of the entrance acting as a three-lane toll on a dozen lane highway.

All in all, this fight looks like a reasonably safe win for humans, but with one caveat: you need army leaders whose IQ wasn’t reduced to single digits for plot reasons.

Moral of the story: If you want an impenetrable location with maximum safety, you should be looking at New Berk (or distribute the dragons around the original archipelago)

Thanks for coming to my Ted talk.

 

But we’re not done yet. Bonus round.

High to the skies, across the seas: Safety of the Old Berk

I’ve already did write about this once on reddit, but it’s somewhat relevant here as well so we’ll address it.

It’s been established that Hiccup isn’t particularly wise in the second movie already, that he makes emotional decisions in a heat of the moment rather than listen to reason. Had Hiccup started expanding dragons and Berk’s influence around the archipelago, Berk could become a force to be reckoned with — and while the dispersed islands seem vulnerable to attacks at first, it turns out that they really aren’t. It’s possible to work that to your advantage.

Dragons and wyverns are generally faster than ships (unless they’re given invisible engines made of handwavium and the same stuff plot armor is usually made of) — especially adult dragons that don’t have to carry an entire Berk with them. You could use them to patrol seas around the archipelago, spotting hostile ships way before they become a problem.

When the ships are spotted, main settlements can be alerted, coming to aid in very little time. And when other settlements come to aid … well, that gets the attackers stuck between the rock and a hard place.

Expanding all over archipelago makes plenty sense as a solution to the overpopulation problem — way more so than cramming everyone into a single space.

How Tall is New Berk — a Follow Up: Unofficial Official Height

There is some long-standing things I wanted to address about my estimates of the height of New Berk in How To Train Your Dragon 3, but I could never be bothered to actually follow through with it.

I’ve been thinking about making another round of “how tall is berk” calculations almost ever since the original blog post went up, as I figured I could make those eyeball estimates a bit more accurate — plus, the comment section on reddit did cause some commotion. Eventually I realized that I really can’t be bothered spending a weekend crunching through numbers, so I’ve decided to abandon the plan.

But there’s one thing I really wanted to address still. If you’ve been following my paper napkin calculations regarding aspects of New Berk, you might have noticed that I’ve been using two figures for the height of Berk: the pi kilometers one, which was calculated by pixel licking, and a two kilometer one. I guess it’s high time I explain where the two kilometer figure comes from.

There is this one redditor who claims to work for Dreamworks, and he claims that he can measure the heights of the island. The redditor was kind enough to give us the height of the lake on the New Berk (43,677 ft) and the height of the “overlook” (we can only assume they meant the rock the lift is placed on; ‘about 44,097 ft’), we get a difference of 420 ft. This gives us about 21 ft per meter. Converting those measurements back to meters using this ratio gives us the height of just shy of 2100m for the lift (and ~2080m for the lake).

That number is to be taken with a huge grain of salt, though, as the redditor claimed that ‘somewhere in the village’ was ‘about 53k feet above the sea.’ 53k ft is 2.5 km above the sea level, which makes these numbers anywhere between ‘sus af’ to ‘load of bullshit’ — judging by the screencaps, no part of the village appears to be over 400 m above the lake. Even 200m seems a bit much by eyeball estimates.

You can give the redditor a benefit of the doubt that he forgot where his initial measurement was taken from, but I’ll be sitting in the corner, smashing that X.

Un-disputing the total height of Berk

A different redditor has pointed out ‘a flaw’ in my calculations. The comment says that I can’t really determine how tall the island is, because objects further away from the camera appear smaller.

So I decided to double-check where mountains are located.

This isn’t very scientific, I know.

Quick guesstimate says that village and the highest peak should be roughly the same distance from the camera, so our quick guesstimates for total height of New Berk (not just sea-to-settlement) shouldn’t be too far off.

How Fast Is the Lift on New Berk (HTTYD: The Hidden World)

I was very bored, so I decided to calculate how long would a ride on New Berk’s lifts be. It would be … very long.

There’s very few things to like about How To Train Your Dragon: The Hidden World, but that doesn’t mean they exist. In fact, the lift at the end of the movie is legitimately my favourite thing about it (and I feel a bit bored today). This, of course, means we’re gonna analyze it a bit.

Do you even lift?

We’re gonna pretend that we’re in high school again and that things like ‘friction’ and ‘drag’ don’t exist where we don’t need them to simplify our calculations.

The first question is: how much power do those windmills produce? To start things off, let’s re-cap some already known calculations that should put us somewhere in the ballpark.

The image doesn’t state the height of the hut outright, but re-calculating hut height from the pixel counts give us 10 meters.

I’m not gonna go into this level of detail today, I’ll just do the eyeballing. If you take the right lowermost blade of the big windmill and measure its length, you’ll get about 110px. Using same conversion factor as we did for the viking, we get about 6.6m for the long side of the blade on average. The width of the blade seems to be about 30px, or about viking-sized — and we’re running off the 180cm estimate for that. That gives us 11.88 m² for each blade. Since there’s 8 of them, that brings our total surface to about 95 m².

But surface area is only one of the things to see how much we could lift by the power generated by said windmill: the other two are wind speed and density. For density, we’ll just take the density at 2000m from engineering toolbox, which comes out at 1 kg/m³.

Wind speed is trickier. In the movie, wind speed doesn’t exist at all, so we’ll have to use our own. 5 m/s seems to be fairly reasonable constant speed. 10 m/s feels slightly generous, but given that the island sticks two to three and a half kilometers out of the ocean … we’ll allow it. 15 m/s would already get into risky conditions, probably and is almost ridiculous. We’ll do all of these three.

Fortunately for us, some people have already made online calculators that calculate wind power based on our parameters. We’ll take this one because it has units that we want. Results are out: ~6 kW for 5m/s, 47.5 kW for 10 m/s, ~160 kW for 15 m/s.

Boy these numbers surely rise up fast, don’t they? But while we’re cheering at those very high numbers, Albert Betz is standing in the corner with a baseball bat, ready to beat some harsh reality in our results. Turns out that you can only turn 59% of that wind energy into something else, with modern turbines being only ~40% efficient. Berk isn’t a modern wind turbine, though — it’s more like those Dutch windmills. Internet says ~15% or less for those, so we’ll just operate with 10% just because the Dutch know their shit better than Berk.

That leaves us with … 600W, 4.75 kW and 16 kW.

But that’s just the biggest windmills. What about the second and the third one?

The second windmill (#2) probably doesn’t produce anything even under full load. The blades are parallel to the wind. The third one is just a smaller version of the first. Eyeballs say blades are ~8 m² (6×2), and there’s 6 of them. This gives us a surface area of ~48 m², and power outputs of 360W, 2.9 kW and 9.7 kW.

Combined outputs are 960W, 7.65 kW and 25.7kw.

If we factored losses from pulleys that run the lift, rather than just the losses in the windmill portion of the lift, this amount of power probably couldn’t lift shit.

Speed of the lift

See this speeds?

Not only does the lift move way too fast, the wind isn’t blowing at all, either. Take a look at the trees.

It’s nowhere this fast.

To brush up on high school physics, everything that you lift has a potential energy. The amount of potential energy depends on the mass of the object, how far above the ground it is and the acceleration of the free fall:

U = m*g*h

Free fall acceleration is 9.8 m/s², height is 2-3.35 km (note: ‘2 km’ number came up after the linked post was written, but I haven’t written the third revision quite yet). We’ll be using 2 km and 3.2 km for the calculations.

And now we get to the weight. We’ll assume that the lift has to lift a standard viking warship, about ten warriors and about their weight in cargo. For weight of a viking ship, we’re surely in luck. After all, Draken Harald Hårfagre — a modern reconstruction of a viking ship, constructed using traditional materials and somewhat traditional methods — is a thing that exists. With displacement of 95 tons (presumably empty), it’s pretty heavy.

It’s also 35m long, which the ships in HTTYD aren’t. Eyeballing the lift, those ships appear to be more like 15 meters long. Maybe 20m, but even 15 might be a stretch. We’ll assume uniform scale in all three dimensions. Scaling down to 15 m gives us the weight of about 8 tons. The 10 warriors at the average ideal weight of 75kg and their weight in cargo add up another tonne and a half. We’ll round the ship weight to 10 tons.

And then there are the ropes. Given this is middle ages, they’re probably using hempen ropes. And boy do we have data for that. Breaking strengths, safe load factors, weight per meter. Long story short, it quickly turns out that 2-3 km is too far for a single run of rope as its own weight blows the standard safe load factor in under a kilometer. However, even at three kilometers (and two hundred meters) we don’t blow the minimum breaking strength yet. While transfer stations would be preferred, we’ll just use a single rope and a meager safety factor instead.

Two kilometer run of 2″ rope would weight about three tons. 3200m run of rope would weight about 4.9 tons. Load at the top of the rope would be thus ~29.8 kN and ~48 kN, respectively. Minimum breaking strength is 120 kN. We want to maintain a safety factor of a measly 2, which gives us 60 kN. Rope takes out 29.8-48 kN of our 60 kN capacity, which leaves us with 30-12 kN to work with.

10 tons is 98 kN, which means we need 4-10 (It’s actually 9, but we’ll round up to keep rope counts on both ends of the ship the same).

This gives us another 12-49 tons that we have to lift. However, since weight of the rope changes with how far up the lift is, so does the weight we have to lift. In the end, ropes effectively add only 6-24.5 tons to the mass we have to lift.

Potential energy at the top of the lift comes out to be anywhere between ~315 MJ (2km) to ~1.1 GJ (3.2 km).

Energy is also power over time. If we divide the energy we need to put in with the rate at which we’re putting it in (so, power), we get the amount of time for which we need to keep putting the power in.

So let’s see how the wind speed translates into lift speed.

One lift, windmill #1
Wind speed2 km3.2 km
5 m/s 6 days (14 m/h) 21d 5h (6.3 m/h)
10 m/s 18h 24m (1.8 m/min) 2d 15h (50 m/h)
15 m/s 5h 30m (6 m/min) 19h 15m (2.8 m/min)
One lift, windmill #3
Wind speed2 km3.2 km
5 m/s 10 days (8.3 m/h) 35 days (3.8 m/h)
10 m/s 1d 6h (66.6 m/h) 4d 9h (30 m/h)
15 m/s 9h (3.7 m/min) 1d 7h 30m (1.69 m/min)
One lift, both windmills
Wind speed2 km3.2 km
5 m/s 3d 19h (21 m/h) 13d 6h 30m (10 m/h)
10 m/s 11h 25m (2.9 m/min) 1d 15h 30m (1.35 m/min)
15 m/s 3h 24m (9.8 m/min) 12h (4 m/min)
Both lifts, both windmills
Wind speed2 km3.2 km
5 m/s 7d 14h (11 m/h) 26d 13h (5 m/h)
10 m/s 22h 50m (1.4 m/min) 3d 7h (40 m/h)
15 m/s 6h 24m (5.2 m/min) 1d (2.22 m/min)

If you convert those numbers to actual speed, you’ll quickly notice that this is borderline snails pace. And mind: those are the average speeds. Speeds are going to change depending on the rope length.

Of course, that assumes they don’t use counterweights (and in the movie, it’s clear that they don’t) and that they don’t have transfer stations (cannot be determined from movie). Just by adding counterweights (about as heavy as the ship), you could drastically reduce the time you need for a trip pretty drastically — and by adding transfer stations (hopefully located near a waterfall for free water power) you could also reduce the weight associated with ropes.

But as things are in the movie … well. Better take the stairs.

How tall is New Berk, really? (Revisited)

My previous island put the settlement of New Berk at about 5 kilometers above the sea. This is obviously ridiculous, so I decided to revisit my calculations …

Not too long ago, I made a reddit post in which I calculated how tall is New Berk in How To Train Your Dragon: The Hidden World. Certain discord server ended up expressing some doubts that I misidentified the location of the lift on the full view of New Berk. While it turns out that I didn’t misidentify the location … let’s be completely fair and honest, 5 kilometers is a ridiculous height estimate. With 4K BluRay in my hands, the time has come to do a second round of calculations.

Seeing the forest for the trees

Another way to eyeball the size of Berk is to take a look at the trees. There’s spruces growing all over the island. Setting-wise, it would be reasonable to expect that these are the European kind of spruces — and these can grow up to 35-55 meters tall. Do spruces on New Berk grow this tall? Quick sanity check says they do:

Note that I didn’t bother correcting calculations for the fact that trees farther away from the camera appear smaller. We’re only doing a quick sanity check

Now that we know that our assumptions aren’t too out of the line, we can start counting pixels. Time to get our hands on some 4K footage for best results. The scene where vikings arrive to the New Berk seems to be a nice candidate for that:

Imagus users can hover over the image for fullview. The rest of you will need to click image to open it in a new tab.

That doesn’t seem to unreasonable, although it may not bode well for our initial “Berk is 5 km above the sea” estimate. Yikes. Let’s find a complete shot of New Berk.

Click image for full view

Well. There’s not much to say, really. While the original estimate was within the same order of magnitude, the 4K screenshots seem to suggest that the height of the new island is between 3,3 and 5,6 km total, with settlement being between 1.7 and 2.8 kilometers above the sea.

But we aren’t done yet.

Autism intensifies

There’s still further confirmations to be had, and chapter title like that certainly doesn’t inspire much confidence in what’s to come.

Let’s try to guess the FoV of this shot.

Can we do it? Turns out yes … but we’re gonna need math. For starters, we need four points on the image that we know real-world coordinates of. Since we know how tall the boathouse is, we can quickly pull four points out of thin air.

Now, I haven’t watched enough Rick&Morty to understand the sort of math needed to calculate the FoV and direction in which the camera is pointing, but someone over at stackoverflow has — and they were nice enough to write their answer in Python — a language that I know a very tiny bit. That answer doesn’t give us everything, though, but let’s not worry. There’s also this math stackexchange post, which pretty much seems to cover the rest.

I know what you’re thinking. That’s a lot of math, and I really want to avoid having to do said math. Especially since I haven’t watched enough Rick&Morty to even figure out how my datapoints translate to solutions described in that post. That, and I’m also lazy.

There must be an easier way, right?

There is. Behind Berk, there’s a lake. Lakes have one very nice property: unless the lake is particularly big, its surface is going to be flat. With that in mind, the New Berk shot was picked very carefully: when moving through the scene frame-by-frame, I found a frame where the lake that surrounds New Berk is aligned with horizon. With horizon known, we only need another point, for which we know the between horizon and the direction in which the camera faces towards said point.

Fortunately for us, that will turn out to be easy. We can trust that support pillars for the lifts (as well as the walkways and terraces around the lifthouse are roughly perpendicular. This means we can easily determine the directions in which the axes of the 3D coordinate system point.

Since support trusses are perpendicular to each other and parallel or perpendicular to the horizontal plane, we can use a few strokes of white to highlight the coordinate system. Note that the “vertical” axis is incorrect, as it should not be completely vertical, but we can approximate and pretend that it is.

And not only we have that: we also have Minecraft.

Welcome to my mine. We are mining diamonds. We are not a strip mine. We don’t have to fight mobs. (Haters go away.)

Turns out that in Minecraft, pressing F3 gets you some nifty developer tools. Among said tools is a reticle that’s shows the direction of the main axes of coordinate system. The tools also tell us which direction we’re facing, with second number telling us the angle relative to the x-z (horizontal) plane.

Through the magic of Linux, we can make GIMP transparent and ensure it’ll always stay on top of any other windows:

That allows us to overlay GIMP over Minecraft. We carefully align the point where all axes cross with the corner of Minecraft’s reticle and wiggle the mouse around until the minecraft cursor covers the lines we drew in GIMP:

Aligning that on a ~160 PPI monitor was pretty tough.

Let’s see whether the reticle truly fully overlaps our lines — it does. We also get to read out the angle that we need:

We have our angle. -7.7° (that’s 18.14° for our American friends).

And now we can finally calculate the FoV of the camera.

This seems to be on the edge between “fine” and “too wide.”

The results are in: the shot seems to have vertical fov of 38.1°and a horizontal FoV just shy of 90°. This feels somewhere between just about right and too wide. Angle isn’t not unreasonably wide, though: assuming this were shot on 35mm film, the lens equivalent for that would be a 18mm lens.

Now that we know the FoV, we can try verifying the island height — and not only that: we can also calculate other dimensions.

Mathrix: reloaded

Back when I was making the initial calculations, I only had a shit (low-bitrate) 1080p version of the movie, so the details weren’t really clear while pixel-licking. Fortunately, though, I now have 4K bluray, which has all the bitrate and all the detail. And on 4K version, we are actually able to make out some impressive detail. What used to be a small blob of pixels is now indeed a viking:

Resemblance is uncanny.

What is more, we can reasonably guess where the floor ends, his legs begin and where his head ends. Your garmin is about to start making noises.

[Click for full view] — The viking strat we used the other day — combined with the FoV we calculated — seems to be giving off familiar results this time around. Keep in mind that “distance to the mountain” doesn’t account for the camera being roughly 145 meters away from the viking by the lifthouse.

Would you look at that. Our optimistic estimates from earlier seem to be roughly confirmed: it’s safe to say that the island is about 3-5 kilometers tall, with Berk being 1.8 to 2.8 kilometers above the sea.

The Final Interpolation

Up until this point, all calculations have been done by estimating the island size based on some assumptions regarding tree sizes. But now that we know the distance between horizon and the bottom floor of the lifthouse, we can actually get a bit more accurate measurement.

Since this shelf is on the horizon line, we know it’s level with the lake

So we know that this bit of rock is level with the lake, and we know that the bottom of the lifthouse is roughly at the top of the rock (maybe sunk a bit deeper thanks to foundations). In theory, this means we should be quickly able to determine the height of that arch, right?

Of course not.

The top line is probably too high as it is. I’ve moved it lower in order to account for vikings standing on it. However, it seems that they removed quite a bit of the top when placing the lifthouse. I haven’t accounted for that.

It’s apparent that the shape of this rock has been at least somewhat changed when building the lift. The top is much flatter with the lift on. However, by eyeballing some things we quickly get how tall the part of the arch located under lake level is: anywhere between 17.5 and 20 meters.

As the last step, we can now pull up the screenshot that shows the Berk in its full flory, from the lift rock to the sea and re-calculate our heights:

[Click for full view] — Turns out you can’t trust trees because they’re all high and shit.

So that’s just over three kilometers for the settlement and six to six and a half for the entire island.

This seems to be the definitive, most accurate answer possible.

Follow-up maths: The Lift and transfer stations

Since the New Berk is just over three kilometers above the sea, there’s one more thing that we need to keep in mind: the ropes have a maximum possible length before they break under their own weight. Let’s see if the lift we see at the end of the movie is even possible.

I will use this as a baseline when calculating my loads. 2″ (48mm) rope would weigh about a kilo and a half per meter, and it’s gonna break when you hit 120 kN (about 12 tons). This means that once the rope is 8 kilometers long, it won’t be able to support itself. But while this number is significantly over the 3.35 kilometer worst case scenario, it includes only the weight of the rope. It doesn’t account for the force of the ship being lifted, additional friction created by the wheels — both on top of the lift, as well as wheels that keep ships on tracks (you can see guiding tracks on the screenshot) — the lateral force that the wind exerts on the rope (as rope gets longer, those forces can get very significant), and any additional stress/force exerted by wheel/rope slipping and then quickly tightening.

If you don’t want to run the risk of rope snapping halfway up, you really want to treat maximum safe load as the maximum weight you can hang from that rope. The site suggests that maximum safe load is 1/12 of the breaking point, so we’ll take that as a gospel. 1/12 out of 120 kN is 10 kN or about 1 ton, which means you’ll need to plant a lift station once every 650-700 meters along the way if you want to lift nothing at all, and 300-500 meters (depending on how much rope can you afford) if you actually want your lift to be useful.

This means we’re probably looking at 8-10 transfer stations along the way. One could argue that, vikings being vikings, they had no OSHA and other pesky safety bureaus and organizations. On the other hand, though, they probably knew their ropes and how much they can carry, and knew their limits. Even if they initially felt a bit more adventurous and cut corners on the transfer stations (having as few as three or two, maybe none) they’d eventually come to the conclusions that transfer stations are not optional and built a proper number of them.

Follow up (2019-07-31): Unofficial official height

So I’ve been thinking about making another round of “how tall is berk” calculations almost ever since this blog post went up, as I figured I could make those eyeball estimates a bit more accurate. Eventually I realised that I really can’t be bothered, though, but there’s one thing I really wanted to address — so I’ve decided to update this post instead.

There is this one redditor who claims to work for Dreamworks, and he claims that he can measure the heights of the island. The redditor was kind enough to give us the height of the lake on the New Berk (43,677 ft) and the height of the “overlook” (we can only assume they meant the rock the lift is placed on; ‘about 44,097 ft’), we get a difference of 420 ft. This gives us about 21 ft per meter. Converting those measurements back to meters using this ratio gives us the height of just shy of 2100m for the lift (and ~2080m for the lake).

That number is to be taken with a huge grain of salt, though, as the redditor claimed that ‘somewhere in the village’ was ‘about 53k feet above the sea.’ 53k ft is 2.5 km above the sea level, which makes these numbers anywhere between ‘sus af’ to ‘load of bullshit’ — judging by the screencaps, no part of the village appears to be over 400 m above the lake. Even 200m seems a bit much by eyeball estimates.

You can give the redditor a benefit of the doubt that he forgot where his initial measurement was taken from, but I’ll be sitting in the corner, smashing that X.