Uncovering the budget of Breath of the Wild and Tears of the Kingdom
Вставка
- Опубліковано 24 тра 2024
- In this investigative video, we're going to try to reverse engineer the budget of The Legend of Zelda: Breath of the Wild and The Legend of Zelda: Tears of the Kingdom. From the game engine, planning and staffing to the testing, quality assurance, localisation and marketing. We'll be including a look at Nintendo's expectations for sales and the long-term vision Shinya Takahashi set out ten years ago to explain how they budget big projects and show some surprising games that benefitted from the development of these titles. We'll also identify why Shigeru Miyamoto was quoted out of context and review the non-Zelda game that may have been hugely important as a milestone for Tears of the Kingdom's development.
00:00 Aims
01:09 They "know nothing about the business" of games
03:25 Nintendo's approach to budgets
05:15 How Miyamoto was misquoted
07:30 The Takahashi plan
11:42 Staffing
15:42 Year by Year
31:19 The biggest cost of all?
![[柴犬ASMR]曼玉Manyu&小白Bai 毛发护理Spa asmr](http://i.ytimg.com/vi/0TsXQ7z2Dh4/mqdefault.jpg)
After a lot of looking, I've now managed now to figure out a way to make some cuts that preserve the video but remove some of the discussion around game engines that people considered to have problems. Thanks everyone for your feedback!
Maybe check the subtitles, I was really confused reading about Havoc but the audio didn’t mention that anymore…
@joaoparedes1431 Same here. I had to turn off subtitles and it's getting kind of difficult for me to catch up ^^;
@@raeesrichards6786 @joaoparedes1431 Thank you for the heads up - didn't think about these. I'll have a look at the weekend and try to match my edit.
24:00 "I rather doubt that even Nintendo have ever specifically crunched the numbers for individual games" - I think there's an interview out there where someone fairly senior at Nintendo basically said that they don't have the ability to calculate an individual game's budget due to how the company works.
Ah, I'm glad you said that. I felt like I had seen a similar comment but couldn't put my hand to it.
Can you be more specific? Is it because staff and development resources are shared between multiple games, and so can't be consider part of any single game's budget?
@@wishonpleiades6288 Something like that, I suspect - It's a vague memory of something I read years ago so I can't really be more specific, but based on my recollection and nintendoforecast's videos - While most Western companies rely on contractors on a single project, Nintendo - at least in their Japanese offices - seem to move people from project to project fairly fluidly, moving them to where they need to be rather than having specific resources dedicated to specific games. So - based on both what nintendoforecast and my recollection of that interview, instead of having a different cup of resources for each game, they have a more holistic bucket of resources that they dip those game cups into as needed. Basically, Nintendo seems to have an overall budget for their overall game development each year rather than assigning budgets for their individual games.
@@wishonpleiades6288 Nintendo hires people in the long term. By keeping people around even if they work on failed projects that never made any money or even remained only in the planning phase they still keep the knowledge and experience in. This mindset of staying together makes budgeting and allocating people through money less relevant if not almost totally unnecessary.
When money starts to become an issue they will instead first cut the wages of the leadership to be able to keep paying their workers normally. This is one reason why Nintendo's success tends to always come again and again after each failure. Because they let the knowledge of failing stay in the company instead of booting it out. And how do people learn most effectively?
It would have to be. Marketing and admin are pretty easy to allocate in at least a roundabout way. But that tells me that they throw developers around to different projects, whoever or whatever is needed.
Note. Havok in of itself is not a complete game engine, but a tool to simulate physics that plugs in to other engines, including Nintendo’s in-house lunchbox engine
It is likely that botw was built with an earlier iteration of the lunchbox engine.
Thank you for the note - appreciate it.
Its the same engine as wind waker hd
So $315 million spent on both Wild games combined with a minimum of $2.5 billion back in revenue which is over $2 billion in profit (probably closer to $2.5 billion when you include limited editions, console bundles, amiibo and other merch). I think the mega franchises (3D Mario, Smash, Kart, Animal Crossing, Zelda) make Nintendo around $1 billion profit and the secondary franchises (Splatoon, Mario Party, 2D Mario and Luigi’s Mansion) make around $500 million profit. A crazy place to be financially when you look at the absolute state of the video game industry where teams are spending $50 million on a licence and $50 million on cutscenes like Spider-Man 2 for Sony... Long live Nintendo!!
I genuinely and fully believe, that the Switch is largely to thank for the notable revitalization of the AA market that we've seen this gen.
It really is the case that game devs have trouble with arguing for and getting investors to invest in AA projects. To convincingly justify the projects, when the investors see AAA projects worth billions and wondering why you aren't just doing that instead.
But to point at the market leading console, and say that you should do AA games for it. Since AA games are all the console can handle. Then that's an easy lock, because as sexy as AAA shit might appear to investors, the Switch numbers speak for themselves.
And that in turn has had a broader effect on the industry. Because we've gotten more and more successful AA stories. Which gives people who push for AA projects, more to point towards when arguing for how AA projects can pay off. Instead of every single AA game being some odd exception that could have been cast aside as a fluke.
How does this guy still only have around 7800 subscribers? The quality of your analyses is astonishing! Quite the opposite of the uninformed, clickbaity prediction videos I see way too many of nowadays.
Tears of the Kingdom still blows my mind...
In what way?
@@coyy9106In many ways 😅
@@chunkymonkey7983I mean, do you think it’s really good? Or it’s mind blowing that they spent so much money and years on it and barely expanded the world?
@@coyy9106 If I saw this comment a few months ago, I would have disagreed. But now, I have to sadly agree with you.
@@loneranger4282 I don’t hate the game. I actually really enjoy it. There’s a solid gameplay loop and the world is just utterly beautiful.
They just massively underdelivered on new content, whilst fixing barely any of BOTW’s issues, and creating new problems in the process.
But the hype juice really did help in those first few months. I’m still enjoying it, though.
There were lot of mistake on the engine part.
A Game Engine is a mix of lot of different tools that comunicate with each other, a Framework is one of them and are assets specific to one thing that can be used as an easy integretion on an engine (for example Photon is an online framework that can be used on Unity or Unreal projects).
Neither BotW or TotK use Lunch Pack. Botw run on an engine called King System that was made only for that title, TotK runs on Module System a new version pf Lunch Pack that we saw for the first time on Switch Sports, and later for Splatoon 3, Mario Wonder and Mario vs DK and Havok is integrated in all of them because it's implemented in the engine.
The real reason to why they changed the engine is because Nintendo want to unify it's development tool across all epd divisions and relegate on different engine when they have to work with different companies.
While I like the intent of the channel, this is the not the first time that I think you have to do better research on the development process of Nintendo and the teams works
So totk took long because of engine switching? Am I getting right based on what you said?
@@ElJosher Changing a Game Engine is not an easy task, but I'm pretty sure that, giving both engine were made by Nintendo, they were able to transfer most of the logics pretty easly. Nintendo did said that in March 2022 the game was already finished but they took the time to polish the physics of the game and we cant' forget how much the covid impacted the company
Thank you for this. Obviously I use what I can find in terms of research (which did include statements about TotK using Lunchpack) - the nature of the channel - which aims to make "researched, measurable, accountable forecasts" - is that it's about making general estimates based on the best available info I can find and adapting to perfect those judgments. I can't claim to be perfect in any individual area and I don't have the infrastructure to necessarily research everything to a journalistic level. However, I will try to correct where I can.
I appreciate the added information and clarification. If anyone has any sources that would be useful, that would be great.
I've pinned your comment so people can read for themselves and make their judgments.
As to the video, I'll have a think about what to do with it. I may see if I can clip the engine part but that seems tricky. More likely I'll take it down and re-do it at some point; or perhaps just do a separate video correcting/explaining.
A corrections video would be fine, don't worry @@nintendoforecast
@@nintendoforecast I really admire your attitude for taking criticism. @hyperfixatedd may be right but the delivery of this was pedantic and a little rude. This is still a great video and highlights many important pieces of the zelda development process. please don't take it down. Instead maybe a separate video going deeper into Nintendo game engines (and maybe even comparing them to others in the industry) would be a great subsequent video.
Keep up the nice work. Again as a fan and investor of Nintendo I really appreciate the time and effort you put into this channel.
Im very excited to see the BotW engine being employed in more Nintendo games. Pokemon could stand to benefit a lot.
Its never been the BOTW engine. Its been monolith soft's xenoblade X in house engine upgraded the entire time.
@@OzzhirothNo, Monolith Soft engine is only used on Xenoblade games. Botw run on King System an engine created specifically for that game while TotK on Module System born to unify their development tools because is used on Mario Wonder and Splatoon 3 for examples
I think it would do the opposite actually, the big problem with open world Pokemon is they gotta be prepared to load any single Pokemon model at any time, instead of the previous games's more controlled approach. Granted I'm not a tech god or anything, but it seems like the lighter-weight the engine is the better in this context.
mario wonder and splatoon 3 use the Botw engine apparently
Perhaps this video is flawed as others have said but it’s the most thoughtful on the topic that I know of.
This channel must take so much work I can’t image how much research you have to do!
Havok is not a graphics engine - it’s a physics engine. It’s not like Unreal, lunchpack, etc.
You can use havok on other engines. It powers the physics in Hedgehog engine, for example.
Huh? Didn't he say that though?
I thought I referred to it as a physics engine but I can from the feedback generally that this whole section needed some work so it's good to have the clarity.
I would also love to see your references in the description to learn more it's a fascinating topic
I remember PhysX, before NVIDIA bought it. Now, it’s rarely used…
This is an extremely well researched and engaging video! I really hope your channel takes off, in depth reporting on the financials of the gaming industry is relatively lacking in the broad scheme of things, so it's wonderful to get a "peek behind the curtain" on that front. Your analysis on the individual staffers that work on each game and how they shifted around over time especially is impressive, these considerations are very rarely taken into account when discussions of the process of video game development are had. Good work, I definitely wanna see more from you!
Thank you!
25:11 Breath of the Wild development started in either late 2011 or early 2012… not 2013. Nintendo repeatedly stated that Breath of the Wild took five years to develop, so 2013 wouldn’t make sense in the slightest.
2013+5=2017
@@tumultuousv 2013+5=2018
@@tumultuousv Yeah thats how it "works"
I was using Fujibayashi's own statement which is in the video clip although I didn't highlight it. It's from the Making of BotW documentary. I think my assumption is that time spent on BotW previously would have been a side-project on Wind Waker HD and the "day job" of the staff would have been that game. If you count 2013 to 2017 inclusive (i.e. from Fujibayashi's "work began in earnest in January 2013" to the last DLC pack dropping in December 2017) you get pretty much exactly 5 years.
@@nintendoforecast Yes, but they’d stated that the game took five years BEFORE the DLC came out. They’d also stated that they started working on it right after Skyward Sword finished development… which is not unprecedented. That’s usually how it works with any Nintendo game. Super Mario Odyssey, for example, started development immediately after Super Mario 3D World finished. That would put the start of Breath of the Wild’s development in late 2011 or early 2012. It being a “side project” is irrelevant, because side project or not, the game is being worked on.
With Tears of the Kingdom, 2018 sounds about right, because they wrapped up the Breath of the Wild DLC in late 2017, and Tears released in 2023.
Just found your channel recently, really love how researched your videos are 🔥
Thank you!
@@nintendoforecast You are very welcome! ☺️
Been putting it off over and over again, and finally last night decided it was time to kill Ganondorf while my friend came over and watched.
This video was truly brilliant! So much work and research, presented in what I felt was a entertaining way!
You also helped confirmed some ideas I have had floating around in my head, with areas where Nintendo would save money - like with staff retention for example.
Easily earned a new sub here 🙂
Thank you!
What an underrated video!! Damn good editing and content.
Thank you so much!
It is interesting to speculate. I don't know how accurate this estimate is, but I hope other companies realize great games can be made with relatively smaller budgets and not every game needs to be $300M plus.
While that would be great it's also the case that it would be much more expensive for anyone else to do this without the benefit of Nintendo's standing developer base, developer experience, established infrastructure and franchises capable of defraying costs across multiple franchises. If you were trying to achieve this from a standing start, it would be orders of magnitude more.
130-185 is both a lot and at the same time less than i was expecting. the simple fact they went for a more low-detail, stylized artstyle probably saved them a lot of time and money
Considering Spider-Man 2 had a budget of 300 and sold less than TotK it having a budget that probably doesn't break 200 makes the game quite a successful one.
Very unlikely any of them had a budget even close to a 100mil they were around 50m for each
Totk doesnt use lunchpack and instead uses ModuleSystem. BotW uses KingSystem
Pretty good video, i like how well researched it is.
very interesting. never seen somebody did budget investigation into botw/totk this thorough.
Awesome video! Glad this one showed up in the algorithm.
Thank you!
Knowing that Breath of the Wild was on a different engine and toolkit (plus licensing costs associated with it) than Tears of the Kingdom makes that rumored Gamescom demo of Breath of the Wild for Switch 2 makes me think that they'll re-release it with the new engine underneath.
EDIT: I wish Nintendo would go back and fix localization mistakes via patches, particularly in Zelda games.
His discussion on the engines is hugely flawed. The two games likely have the same licensing cost because they both run on different engine yes, but both were in-house but they have some things in common, for example the use of Havok
Those localisation mistakes are usually completely overblown. Mistakes happen, but the localisation also often gets completely misrepresented, like claims that Revali is actually nice to Link or that the Zora and Hylian side quest doesn't have romantic undertones, when both claims are false.
Wow I’ve never been early for a weather forecast!
Now you have
Man that 50min went fast
is there a point in the video that just gives me the numbers and a breakdown. Ie total cost x million, graphic costs is a, marketing is b, etc.
2D Zelda prototype has me thinking when Nintendo had Star Fox Zero after showing the Star Fox Wii U Demo some years ago.
This discussion of Nintendo is my jam. Love the research, context, and insight!
Thank you so much!
Good video but I need to point out that Havok is a Physics Engine specifically, NOT a game engine. Havok is something like a plugin that can be adapted into actual game engines. Nintendo has their own internal engine which they likely initially implemented Havok Physics into and then worked off of that.
Damn a 50 minute video? Bro gonna be dropping entire movies at this rate. On a more serious note keep it up with these more indepth/investigative videos! I haven't watched this one yet but I prefer these types of longer vids rather than the shorter or more speculative ones you've also posted lately. I don't know what your analytics look like, but if you're concerned about having to post regularly to keep up with the 'algorithm', then there are definitely channels like Vsauce that do very well despite putting stuff out incredibly rarely (like once/twice a year). Posting less often would also leave more ideas on the table for future content potentially years down the line.
I'm also wondering. Don't know if this is answered in the video. If Switch 2 is supposed to be at a PS4-ish level graphically, are we expecting Ninty's game budgets to skyrocket accordingly, much like what we've seen with the rest of the AAA industry in the PS4 generation? Nintendo definitely struggled just with the jump to HD. How could they handle this? Or would improved dev tooling/engines, plus Nintendo's massive recent hiring boosts, cushion the blow in this day and age compared to if they were making such a leap back in 2013 alongside Sony and Microsoft?
Thanks for the feedback. I'm open to doing more long-form stuff from the time to time although don't expect them too often at present.
How Nintendo keeps up with budget/time requirements is something I want to address (actually this video kind of grew out of that script which I now need to go back and finish). Digital Foundry had a question on this a few days ago as well which is worth watching.
The soundtrack wasn’t majorly recorded with a live orchestra, most of it was done in software with some soloists and with orchestral pieces reserved for big story moments or as flavor pieces like activating the towers in BotW.
A lot of the soundtrack is still done in MIDI and even features some recognizable stock sounds from Apple’s Logic Pro software.
Oh, very interesting. I'll have to look into that. Thank you!
Not sure if you’re able to do a Forecast video on the yet to be re-revealed Metroid Prime 4, especially since Metroid Prime Remastered last year?
I have a video coming probably in a week or so about how Metroid could make it big that touches on MP4.
@@nintendoforecast cool awesome! Looking very much forward to it!
Game Engines aren't monolithic and the diversity of the games using their new internal engine hints at it being more of a series of interchangeable libraries rather than a complete engine in a box, while the engine on paper would be the new one most likely Breath of the Wild's libraries were integrated into the new engine rather than rewriting everything from scratch.
Good point.
Great video, but you should blend the background music a bit more smoothly ❤
Wait an minute. Does this mean, that there could be Developers at Nintendo, who spend probably 10 years, a whole damn decade of their life, on working on this two Zelda games? That would be wild.
You make an interesting point of Nintendo reusing to be used in other games. It would make games like botw and totk more cost effective if they use the technologies they developed for those game into new games.
You can actually tell that TotK is running on a different engine if you pay close attention. It's very similar in feel, but things move just a little differently. That said, the similarly to Havoc really says a lot about Nintendo's coding teams and devs, and the flexibility of their engine.
Definitely the same engine, just some adjusted movements
This was a good video. Thanks
These videos are surprisingly well researched for a small channel...! Do you have an economics degree or an MBA by any chance?
I have a masters but not in economics/business.
Whatever it was, it made a tidy profit
What is the music at the beginning?
From the final sequence of TotK.
"The final fall" from TotK
I know alot of localisation was done by Side UK, they're on my LinkedIn and sometimes post about thier latest Nintendo collab.
Very extensive and interesting video! Came to watch it after @Commonwealthrealm mentioned it in one of his latest video, great content!
Thank you so much! Love @Commonwealthrealm but haven't seen that video yet. Appreciate the heads up!
Wait Scarlet Violet uses the Havok physics engine?
well it doesnt mean it used it well😂
I will only play Totk at 60fps because after playing Botw for 100 hours I have had enough of how slow and bad it felt, hopefully next year with the switch 2!
I hope Nintendo does not give up on Advanced Wars. My guess is you have seen Mings video on why it did not succeed. My big 3 switch franchises for time played are Splatoon, Zelda and Advanced Wars.
Yeah, me too with a bit of Mario Maker and ACNH up there too. Although personally I think the toy aesthetic was the absolute correct move for AW.
32:33 Wait, are you inplying that there is another Hyrule Warriors game to be released relatively soon?
Just imagine....
Hyrule Warriors: The Imprisoning War
Depends on the "relatively". I don't think I've done a video formally forecasting this but I think another Hyrule Warriors mid 2025 to early 2027 is 70%ish. Maybe higher. (Lots of reasons: cheap to reuse assets, the Warriors team put out lots of games, Zelda's 40th, it's a good-looking Zelda for a new console).
Advance Wars 1+2 Re-boot Camp, my beloved. 💙
I think my low-key aim on UA-cam is to do in the next decade for Advance Wars what Arlo has done with Paper Mario TTYD and Pikmin.
@@nintendoforecast I fully support this endeavor.
Let’s gooooo Nintendo forecast hour essayyyyy let’s goooooooooo
Moevius battle is insane
43:55 I think you were meaning to cut this out
Thanks - have done so now.
@@nintendoforecast epic! I’m glad UA-cam allows that fairly easily
42:46
Nintendo spend 638 dollars on marketing😱
Yamauchi "Yama - Uchi"
Yes, great note - apologies!
42:45 Nintendo spend 638 dollars on advertising? How cheap! 😀
I expected more than a mere mention of Monolithsoft, they cut half of their workers to go working on BotW, and having all that information... Welp
It’s weird how games don’t publicly reveal their budget like film productions do at this point. Mind you, I don’t believe games should strive to be like movies in ariistic style, like Nobuo Uematsu said “Hollywood-style scores are ruining video game music”. I mean in an economic legitimacy they should, as video games are the highest grossing media/entertainment industry in the world.
Did he really say that?
These games cost more than the budget of the best movies
If you pay 50 developers 200,000 dollars working straight for 5 to 6 years and pay 1 million dollars per year for renting the installation and 10 million total for hardware and research you still won’t get to 100 milllion and that is if they get 200,000 a year wich most don’t.
And heck we know they dont work with that many people at once and the whole time.
It costed a billion trillion yen, which is equivalent of 1500 dollars
am i allowed to cry?
for people in the netherlands, breath of the wild and tears of the kingdom were both 70 euros
I’m sure Nintendo is not blowing 300 million $ like western dev lol
The way Nintendo makes games is sustainable and that’s what I want to see more from Sony in the future
Less graphics more gameplay.
Spider-man 2 budget is ridiculous lol.
Great video 👍
Stewart from family guy loves Nintendo apparently
Yama-oochi is probably the correct pronunciation 😅
Yes, good call!!
The remakes fund the respective new titles partially.
You're mixing physics engines and game engines up a lot and seem to be using the terms interchangeably even though they are very different things. Lunchpack could very well still use Havok as its physics engine. Many game engines do such as Source while others like Unity and Unreal use Nvidia's PhysX.
The physics engine is just the code that handles collisions between moving objects and their colliders and not much more. A physics engine will do no rendering or taking user input of any sort. It won't have a visual editor usually. It will be entirely command based. This is because physics engines like graphics libraries are just a programming tool meant to be integrated into a larger project and don't do much on their own.
The game engine is the set of tools you work with to create a game. You can keep a game engine mostly the same and just swap out the physics engine under the hood and it can be entirely possible that a developer using the engine might not even notice.
Nintendo as far as I can tell has been using the same Lunchpack game engine for both Breath of the Wild and Tears of the Kingdom. A complete change of game engine would make the game feel very different to play and likely look noticeably different and would be kind of a waste. The effort that would be needed to make two games made in two different engines feel the exact same to play is not small. Hyrule Worriors clearly put a lot of effort into making the game look as much like BotW as possible.
It's entirely possible that Nintendo may have ditched Havok at some point, perhaps for some technical reason, but that's not really as big of a headline as you make it out to be. Little work by the rest of their team is going to need to be redone with a change like that. It's just going to be technical work behind the scenes to make the new physics engine slot in and work with the rest of the game engine as smoothly as possible.
Think of a physics engine in a similar way to graphics libraries. It's not really a big deal when a game developer switches from OpenGL to DirectX or Vulkan or something like that. On PC, some games even ask you which one you would like to use and just switch on start up, and neither will significantly change the player's experience. They'll just each have their own technical strengths and limitations.
Thank you - one of the clearest explanations so far.
Hyrulean -> Hylian
100+millions USD for the amazing world like Hyrule is still too cheap comparing to fast-food Hollywood movies.
I wonder if they built upon Botw to create TOTK.
Or did they implement all new assets into a new game engine, and then started making the changes where relevant? To account for 6 years later and the Upheaval changes. I’m curious
imagine GTA vi budget
It's probably close to a billion.
Totk made a billion dollars by itself
Havok is not a game engine. I was just thinking about how much of what is being said in videos like this can be trusted; when I come across something I know as a programmer/game tech nerd to be so plainly wrong I begin to question the legitimacy of anything else presented about subjects I am not so intimately familiar with. Thus, I will not watch the rest of the video.
Botw and totk don't share engines dude... And ring fit don't share engine with totk either
Bespoke
ai is going to change game development forever in so many ways
I dont think these games r that expensive.remember tgese games r made in japan and making profits all over the world.
$18
2:33
Ah, so budget reasons were the most likely reason Rare was sold :(
Ah
I feel 60€ for a game now is almost a gift with those numbers
we know Nintendo games are made on a shoe string budget
We don't "know" anything.
I hope many news channels cite you as a source for successfully calculating the budget of both games, that would bring a lot of viewers over to this video
Ha, nice idea - although we can't say for sure I have been "successful" since I doubt even Nintendo will crunch the numbers.
Idk who needs to hear this but, Tears of the kingdom was infinitely better than Breath of the wild. It’s not even close. Breath of the wild is the worst 3D Zelda game ever made.
No way totk was that expensive compared to botw. Same engine, same base map, underground was automatically inverted. Almost no new assets. Meanwhile cheaper titles have worlds were everything is hand made.
Thank you!
Have you ever watched the GDC? They literally redo the physical mechanics in all assets of BotW, and developed a new sound engine that no company had ever done (literally audio version of ray tracing). There's a reason why engineers were "wowed" by a chain in TotK but not in the entire BotW playthrough. The cost of this supposed-buggish game can't be lower than BotW tbh.
I'd say TotK had the biggest software tech leap for Nintendo since OoT (BotW is for them catching up and optimising existing methods
Same engine seems to be wrong. What's "automatically inverted" even supposed to mean?
"We made two of the greatest games of all time."
"What did it cost?"
" _Everything._ "
Tbh I find to add Marketing costs pointless.
It has nothing to do with the game development.
I mean, in what way does Marketing make a game better?
They just add it to show bigger Numbers.
Because, you know, bigger numbers sound better.
But at the end of the day Marketing as absolutely nothing with the games development to do.
Well I have the pre-marketing budget identified as well but most often - rightly or wrongly - when you see these figures quoted they're inclusive of marketing so I think people might have thought the video incomplete without it.
Totk sucks if you’ve played botw
Nope
@@tumultuousv I just couldn’t get past the shrines, koroks, memory flashback storyline, dungeons were eh . But just my opinion
It's certainly no Gothic 2 in terms of sequels, that's for sure.
Or even Majora's Mask, for that matter.
I feel like Gothic 1&2 already surpassed BotW & TotK in terms of open world design in the early 2000s.