If you were a fan of last year’s Horizon Zero Dawn, and happen to be also be attending GDC 2018, good news! Not only are there a billion* talks about the game, there’ll be an entire postmortem of the game’s design process from Guerrilla Games lead designer Eric Boltjes. 

As has been the case for several weeks on Gamasutra, we couldn’t wait to hear Boltjes give his talk, so we invited him onto the Gamasutra Twitch channel for a chat about his work on Horizon Zero Dawn. You should absolutely still make time to see his session at GDC, but while we had him on, we tried to get some helpful insight for game developers hoping to learn from Horizon Zero Dawn’s success. 

You can view our full conversation up above, but in case you’re staring down a Thunderjaw as we speak, we’ve highlighted 3 key takeaways for your perusal below. 

*not actually a billion

Using DLC to subvert player expectations

Our time on the stream was spent in the new content that exists in The Frozen Wilds, Horizon Zero Dawn’s story expansion from late last year. What’s notable about this section of the game, and something Boltjes emphasized about its development, was that it became a space to upend player expectations about the game loop, and create creatures and scenarios that fly in the face of patterns taught in the main game. 

Boltjes admitted that was sort of a risky move, especially with the new Burner machines, dog-like robots that shoot fire and explosive ammunition, and whose attack patterns defy a lot of the “large creature, slow movement” rhythms that make up the rest of the game. Elsewhere, a typical open-world “climb the tower to get the map objectives” encounter turns instead into a fetch quest that introduces one of those difficult new Burner enemies…as we learned to our dismay during our stream… 

For other developers, it’s worth looking at The Frozen Wilds and considering the risks of making DLC content like this, especially when it’s primarily aimed at a fraction of your core player base. 

Tuning stealth indicators was about analyzing a “return” to stealth, not losing stealth

As our conversation rolled along, one viewer in chat asked Boltjes about tuning the game’s stealth mechanics, most noticeably in how enemy characters detect the player when they’re trying to be stealthy. Since Horizon Zero Dawn isn’t primarily a stealth game, Boltjes says, but rather a kind of hunting game, stealth takes on a different role, so designing a “detection” system took a different level of thought. 

According to him, part of the way to achieve this was to give players more leniency in being caught the first time by AI opponents, but making it difficult to fade back into stealth, to encourage players to follow through with whatever battle encounter they started. It’s a kind of stealth design that (hopefully) helps reinforce the sensation of hunting, even though the ‘prey’ is these oversized machines. 

Crafting economies are tough to test

Boltjes’ biggest tease for his upcoming GDC talk was a discussion about Horizon Zero Dawn’s crafting economy, and how it was really difficult to tune and test because it effectively involved measuring run-throughs of the full game. He explained that it was really easy to test how that economy worked within the games’ first 8-10 hours, because a playtester working through that section of the game. But playtesters who were focusing on finishing the story over 2 or 3 days weren’t able to provide feedback on what it would look like if they also focused on side-quests or over-gathered resources in certain areas. 

Boltjes didn’t have a solution for us on stream, but if you’re working on a crafting economy, stories about his struggles may be of use to you when analyzing your own game’s long-term flow. 

For more developer interviews, editor roundtables, and gameplay commentary, be sure to follow the Gamasutra Twitch channel.