How do you make an open world where  the player is completely free to explore - but is also led towards key  locations that will advance the story? This was the biggest challenge  that Nintendo faced when making their very first open-world game, The  Legend of Zelda: Breath of the Wild. They wanted to give players a  sense of freedom and exploration, not seen in the franchise since  the very first Zelda game on NES. But they also wanted to make  sure players were always making progress towards the overarching  goal of saving Princess Zelda. This was no easy task to overcome  - Nintendo had to go through false starts and bad playtests before finally  arriving at the game we all know today. The one that made us rethink how  exploration can work in an open world game. And Nintendo actually shared their experience  of this difficult development process at the 'Computer Entertainment Developers  Conference' in Japan, back in 2017. It was a really interesting lecture  - and a rare act of Nintendo openly discussing the nitty gritty details  of their game design and development. But that information is sadly  very hard to access today. The lecture was never uploaded. All of the reports are in Japanese (and  the only English translation is a tweet thread summary that's been mangled by Twitter). And Nintendo actually got journalists to pull  down their photos of the original slides. So - I thought it was time to right that wrong. To mark the imminent release of the  next Zelda game, Tears of the Kingdom, I decided to resurrect that old talk. I gathered different write-ups of the  talk, and had them translated into English. I used internet archives to  rediscover the lost slides, and used motion graphics to make them come alive. And I used AI upscaling to  bring back old screenshots. So, without further ado, I'm  Mark Brown, this is a GMTK Mini, and here's how Nintendo solved the biggest  problem in Breath of the Wild's design. Okay, so Nintendo had built a humongous world  map for Hyrule, and let players explore in any direction - but they needed to find some  way to lure players towards key locations. The initial idea was to use a  system of "points and lines". The points are the the Sheikah Towers: those  giant, neon-lit spires that soar above the ground. They're clearly visible from a great distance,  and they confer a great benefit to the player: they reveal a massive chunk of  your map when you get to the top. So these 15-or-so towers would be obvious waypoints for the player - which should  effectively lead them around the map. The lines are the routes and  roads between the towers. And so Nintendo could place  various events along those lines. As players walked towards the towers, they would discover characters, enemy  camps and other goodies as they went. But this idea... completely flopped. The guidance worked - but  it actually worked too well. Playtesters felt they were stuck on a linear  path, and forced to follow the towers. Many complained about being trapped on  an invisible, but obvious guide rope. And those who deviated from the line would just  get lost, or find little of interest to explore. And the data bore this out: Nintendo  tracked the playtesters' movements and created a heat map to see  where people had explored. And they discovered that players were  split into two completely different groups: about 80 percent dutifully followed  the main route from tower to tower, and the other 20 percent just  sorta randomly wandered around. Neither play style was close  to what Nintendo was after. So they decided to go for a different approach. Instead of nudging players to always  travel towards Sheikah towers, they could get players to move around the map by luring them towards a larger variety  of landmarks and points of interest. Things like shrines, stables,  and enemy encampments. They just needed to find ways to make  players gravitate towards these places, like moths flapping towards a flame. So, they first made sure each area  would confer obvious benefits. Completing shrines lets you  increase your health or stamina. Enemy bases are filled with weapons to pick up. And while stables were initially just for  registering horses, Nintendo made them much more attractive by adding beds for healing, a shop, and  NPCs who would hand out rumours and sidequests. Other areas would be worth visiting  for the resources contained within. Nintendo purposefully got rid of  simple healing items like hearts, so players would have to go into forests  to get mushrooms or find animals to hunt. And they made rupees extremely rare,  so players would need to go towards mountains and quarries to mine valuable  ore, which can be sold to shopkeepers. For this plan to work, they would  need to make other adjustments, too. You see, the Sheikah towers are enormous and  easy to see - the smaller landmarks less so. So Nintendo had to make them stick out from  a distance, or from a high-up vantage point. Shrines were given a distinctive, lit-up  look; campfires give off a tall tower of smoke; enemy bases are often built  around massive skull-shaped rocks; and the stable is a gigantic  wooden statue of a horse. Wherever you look, you should  find something interesting to do. Also, while there's usually only one or  two Sheikah towers on screen at once, there could be dozens of other,  smaller landmarks nearby - and that many options can be completely  overwhelming in an open world game. And that was partly the reason behind  Nintendo creating the "triangle rule". You see, Nintendo designed  the terrain and landscape of Hyrule to be mostly made up  of triangles - it's all hills, mountains, and rock formations that  are shaped like pyramids and cones. And this has various benefits  for the world design. Like, whenever you face a giant mountain,  players have to decide whether to scale it, or go around it - creating  decision making during exploration. Also, the player's eye is naturally  guided to the tip of the triangle - so you can place points of interest at  the peek to draw the player closer. But most importantly: these triangles  simply block whatever's behind - meaning that the player is rarely overwhelmed  by a massive field of things to do. There's usually only a couple  attractive places on screen, and the rest is hidden behind hills and mountains. But as you go towards these mountains,  whatever's behind is gradually revealed. Whether you climb the hill  or try to move along side it, more terrain will start to be shown. And this has an interesting outcome: it creates a constant source of surprise and  curiosity as new locations make themselves known. So you might be going towards one landmark, but as you travel there - two or  three new places are revealed. Perhaps a shrine on the horizon,  or an enemy camp around a corner, or a distinctive-looking rock, or a  curious sight on the peek of a mountain. Wherever you go and whatever you do, you'll be given a few new things  to catch your eye and attract you. Perhaps that new landmark will distract you, and you'll ditch your old plan  and go somewhere new instead. When you're finished, you'll remember  where you were supposed to be going and head back there - only to be distracted again. Whatever the case, this creates a chain reaction. An infinite loop of discoveries. A breadcrumb trail of landmarks. All of which makes you slowly move across  the map - in an addictive quest of "ooh, what's that?", "ooh, what's  what?", "ooh, what's that?". And before you know it...  you're at a Sheikah tower! Which is exactly where Nintendo  wanted you to go in the first place. So now, with this system of attractive landmarks, players still went from point to point - but,  this time, instead of following a specific line, they were simply following a breadcrumb trail  of interesting landmarks - one that would, eventually, lead players to Breath  of the Wild's most important locales. And where following the towers made players  feel like they were being forced to travel in a specific way, the littered landmark approach  was much more organic, and player-driven. Players would naturally pick places to go, based on their own curiosity - and  depending on their current goal or mood. Locations might be more or less attractive based  on what you need: if you're looking to increase power, then shrines and enemy camps suddenly  become more attractive than stables and towers. Then when night falls, other locations  become more visible and appealing. So players no longer felt forced  to follow a certain landmark or goal - but they still ended  up where they needed to go. And Nintendo could clearly see  this improvement on the heatmap. There was no longer that awkward 80/20 split in the experiences - instead, all players  fell into Nintendo's vision for the game: they could see that players freely explored  various places, following their curiosity from landmark to landmark - but almost all  players eventually got to the key locations. And I totally found this when  playing Breath of the Wild myself. I never felt particularly guided or led around  the world - I was just following my own curiosity and exploring on my own terms. But I still ended up stumbling into important locations - and was always  making progress through the story. So this was Nintendo's first, proper open-world game. And the team clearly had a lot to learn. In the second half of the talk, Nintendo explained  that to get a sense of scale and density, it initially used data from Google Maps to have  Link run around Nintendo's hometown of Kyoto, and clamber up Japan's famous Himeji Castle. And they shared how they had to make all new tools to allow for a large team to  collaborate on a single map. But through clever design, driven by a  desire to create a specific experience for the player - it solved the biggest  problem in Breath of the Wild's design. Nintendo created an open world game that  beautifully balances guidance and exploration. A feeling of freeform adventure that I've  only really seen since in Elden Ring - and, almost certainly, will see  in Tears of the Kingdom. I look forward to jumping in later this week. Thanks for watching,