Ostranauts: A Unity case study
After founding Blue Bottle Games, Dan Fedor’s community-focused approach to game development made his 2012 title NEO Scavenger a cult classic. Its dedicated fanbase kept the survival RPG in Steam’s top-100 best-rated games for over two years. The unofficial sequel, Ostranauts, launched on Steam early access in September 2020 with the same player focus, so that Fedor could “build it together” with the community.
To leverage analytics for gathering player feedback at scale in order to give players a voice during the development process
1 lead developer, 1 publishing strategist, 1 community manager, 20 contributors
Seattle, U.S. and London, UK
Developing with the players, for the players
When NEO Scavenger first launched, it was nothing like the game that it eventually became. The final game was born of BBG’s dedication to dialogue with its players: countless hours of developers’ time spent listening to the community across message boards and Discord servers. This time, London-based publisher Modern Wolf came on board and leveraged deltaDNA to get similar feedback from Ostranauts players in-game, at scale, without interrupting the game flow.
By implementing an in-game voting system using deltaDNA, Blue Bottle Games was able to prioritize feature development.
- 1,000 unique players participated over two weeks
- Streamlined process for collating and analyzing feedback
- Working on player-chosen initiatives helped increase Steam approval rating by 30%
Shifting from qualitative to quantitative
With player satisfaction the number-one priority, Fedor wanted to make sure he was getting as much feedback as possible – in an efficient and non-intrusive way. While scouring forums and interacting with fans individually is a great way to get qualitative anecdotal feedback, it doesn’t work at scale. Furthermore, there are thousands of players who won’t ever visit the forums or Discord servers so he needed a way of canvassing their opinions too. If changes were to be made in-game, he decided that’s where feedback should be gathered.
For a game like Ostranauts, which totally immerses its players in a rich gameworld, metagaming is tough to execute properly. Most games send polls and surveys via in-game messages from third-party tools but this was not a viable option. Any mid-session prompt needed to be baked into the lore itself, consistent with other in-game systems and mechanics, so as not to disrupt the player flow.
After considering the options, Fedor decided to give players access to an in-game “cinema” service via their personal digital assistant (PDA) and designated locations. Subtly embedded within the user interface (UI), players could vote for one of several features that they’d most like to see next at the K-LEG Filmhouse. Presented as film posters, they actually represented a range of specific gameplay features that the development team was planning to introduce. At the backend, this vote was to be linked directly to deltaDNA so that the team could quantitatively evaluate player priorities and schedule their work accordingly.
A new deltaDNA use case: cinemas in space
As with all developers, Fedor’s focus is building the core game and improving player experience. Developing, maintaining, or even just implementing analytics tools takes time that could be better spent on features and improvements. That time is especially precious for early-access games, where out-of-the-box tools make a massive difference. With basic data flow and reporting set up in deltaDNA on day one, Fedor’s attention turned to more specific and powerful use cases.
Instead of building extra mechanics on top of gameplay, the deltaDNA support team worked with BBG’s developers to set up event-tracking on player interactions with the UI itself. Player journeys are rarely simple and users typically learn by navigating around the game. Unfortunately, ingesting all that data as individual events creates an extremely complex dataset that is tough to analyze. In order to avoid data overload and ensure every player has an equal voice, they used aggregates to ensure that only each player’s final vote / UI action was included for analysis.
The people have spoken!
Within days, it became clear that the voting system was popular and effective. Over the course of two weeks, 1,000 players voted on imminent feature development. For context, that’s more than half of the total membership of Blue Bottle Games’ Discord server. As predicted, data aggregates proved particularly important – the total votes numbered roughly 10x the number of total unique voters.
Blue Bottle Games remained very active on Discord throughout the process with deltaDNA. Early on they noticed a significant and interesting disparity. The features that seemed most popular on Discord differed from those actually being voted for in-game. Charis Reid, the BBG community manager, says “this was great for us, as we were able to know ahead of time which feature was clearly going to win. That meant Dan was able to get in a few days work on the feature before it was announced as the winner.”
The voting concluded at the beginning of November 2020, with data in deltaDNA showing that players had a clear preference for “That’s an Order!” – improving the AI and performance of non-playable characters (NPCs) so that they weren’t so difficult to manage.
Powering the feedback loop
With initial bug-fixing priorities dealt with, AI improvements began to surface as early as three days after voting closed. Developers continued to release changes through routine patches in the weeks and months afterwards.
In the first devlog released after the feature vote and subsequent development work, feedback was universally positive. Above is a selection of comments that illustrate the overall reaction of a community thrilled with their developers’ continued dedication to improving the game they want to play.
With NEO Scavenger, Blue Bottle Games took an approach to game development that exemplifies the best possible attitude towards making games. However, the resulting anticipation for Ostranauts required the implementation of a more scalable and data-focused strategy. This ensured the developer/player relationship could be maintained throughout an ambitious and groundbreaking early-access process. By implementing analytics from deltaDNA, Dan Fedor simplified and increased the scope of community-driven development enormously, enabling his team to give players more of what they want, quickly.
“We’re using data events to make a live voting system so players can vote on what they want the developers to work on next as a priority. It’s a perfect solution for the democratic development model that we’re going for.”
“Community focus is extremely important to us and we still interact with players across multiple out-game platforms, but gathering quantitative feedback from within the game itself has been key to scaling our approach.”
“Implementation of deltaDNA was fairly trivial on the dev side and the fact that the reporting ‘just works’ has been a big benefit. I have little bandwidth to spare on non-core development, so quick ramp-up and reliable results are important.”
“If it weren’t for the voting system and being able to see the data updating frequently, we would have been making decisions based on the opinions of a vocal few, instead of what the majority of Ostranauts players actually wanted to see.”