When Destiny 2 "weekly active users dropped lower and faster than we'd seen since 2018," Bungie assembled an A-Team to put out some fires: "We needed to do something"

In fall 2022, Bungie had a problem on its hands. The Destiny 2 developer hasn't gone a single day without some kind of problem lurking somewhere, but this was a big one. As principal technical designer Alan Blaine explained at a Game Developers Conference talk titled "Rescuing a Playerbase from the Doldrums: ‘Destiny 2’ in 2022,'" the game was dipping and needed a solution.

Weekly active users fall between expansion and season releases, "that's pretty standard for most live services, but starting in September 2022 and continuing through November, our weekly active users dropped lower and faster than we'd seen since 2018 with no known cause," Blaine said. "Similarly, player sentiment was also falling lower and faster than we'd seen." (The perennial boogeyman in Destiny 2 is seasonal fatigue, and I don't think Blaine is putting his head in the sand here, but rather saying there was no smoking gun, hot button issue driving players away.)

"Leadership got worried. At the same time, likely due to this falling sentiment, pre-orders for our yearly expansion were well under expectations too," he added. "And strong expansion sales are the majority of our revenue." Needless to say, "we needed to do something really fast."

The response was to "spin up a PRT," or player retention team, a small and agile strike team of developers laser focused on ways to "improve player sentiment first, and then improve player retention." Devs from several other teams and disciplines came together to hash out comparatively quick ways to improve the game, asking forgiveness rather than permission if necessary to get a few things done fast. "If you cannot explain the player impact in a sentence or two, it's going to be very hard for players to understand why they should care in the first place," Blaine said of the messaging at the time, stressing punchy, impactful changes with a fast turnaround.

Sometimes this literally came down to PRT members looking for "things that they could get into the game in the next week, like actually get into the live game in the next week, which meant, in reality, the work had to be done in the next day or two to get it tested in time," Blaine said.

"Many were longstanding player requests that we'd had a hard time prioritizing over new expansion items, but the PRT was able to prioritize some of these really player-friendly features," he explained. The list of target features formed with player support and community teams proved to be valuable to "the entire Destiny 2 project over the next few years as they were planning new player-informed quality-of-life features."

Agility was the focus of development – "don't add any detail until you absolutely need to," with a team working with "just the right amount of structure, and no more" – and the goal was to improve player sentiment. And "if the goal is to improve player sentiment, sometimes telling players about an upcoming quality-of-life feature is as important as the feature itself," Blaine said, which led to the uptick of communication on daily and weekly changes at the time.

That communication goes both ways. "If your players are grumpy and not playing your game, I can guarantee the morale inside your team is falling too," Blaine said. "Your devs are going to be players of your game, as well as working on it, and they want it to do well. So to fight this, I put together weekly emails that went out to the stakeholders list that highlighted new and meaningful work along with any fun screenshots or videos of that, any new high impact [key performance indicators] that we'd discovered, and then a regularly updated shipping soon list of features." (Stakeholders, in this case, seems to refer to team leaders of various disciplines.)

Going into Lightfall – and for now, we're not going to talk about the mess that was Lightfall – this coordinated strike was working. Using Bungie's "rate last week" player survey data to evaluate the changes – an end total of "17 unique items" shipped in the six weeks prior to Lightfall – Blaine said "the numbers for what they thought about them, and the numbers for popularity, were surprisingly high. We knew we were doing a good thing, but the response of players is exactly what we were aiming for."

"We've never seen a season like season 19 where we're basically just climbing gradually the entire time in terms of sentiment, so we knew we were having the effect that we wanted, which was really great news," he noted.

This push for player sentiment and retention continued post-Lightfall. "Some were changes to systems that showed up in Lightfall that we only realized in final play testing needed something extra, but you don't have time to do that if it's in the final round of play testing," Blaine said. Elsewhere, "we had some things that we were requested that were essentially tech debt, which are great to clean up, but that's not this team. This was not our mandate, to solve tech debt."

There was one big oopsie along the way: "We accidentally shipped a nasty character corruption bug that caused us to take the game down for 24 hours and issue only our second rollback of a character database in Destiny history." You know, in case you were wondering how that happened. But despite "one big mistake," it was "a really big success."

The postmortem playbook for this team was "designed to help future PRT efforts within Bungie as well as within the larger PlayStation family." Bungie more recently assembled a PvP strike team to juice one of the game's weaker areas. Which begs the question: why not do this all the time?

A permanent PRT is an encumbered PRT, to sum it all up. Different teams have different needs and priorities – this team had no artists, while the PvP team needed artists for new maps – and Blaine reckons it's "better to actually look at each situation with fresh eyes and find the team that's right for it."

"Once a PRT becomes a permanent team, it'll naturally develop a backlog," and then suddenly "you have a two-year process and you've lost all the agility you've ever had."

As for Lightfall's reception, with our own Destiny 2 Lightfall review skewering "too many well-worn mistakes," Blaine acknowledged that "it was intense" and there were "some rough moments." But much of this wasn't the realm of the PRT.

"If players had a problem with the narrative, let's say at Lightfall, that wasn't something my team was equipped to deal with, but tuning numbers instantly, my team was perfect for that," he said in response to a question from PC Gamer's Tim Clark. "And so we did a bunch of work on Guardian Ranks to smooth out the objective progression, things like that, to help some of those issues."

It's an interesting look behind the curtain of live service triage, and a rare peek into Destiny 2's tougher moments. This troubled season came at the tail end of a year that began with The Witch Queen, then widely regarded as the best expansion the game's ever had, and tragically ran smack into Lightfall, widely regarded as the worst. The game was demonstrably struggling – just as the community is once again in doldrums today, seeking the next big thing after some waffly Episodes – but meaningful changes were still happening somewhere, somehow, and it's fascinating to hear how a dedicated A-Team was assembled to push many of them through.

"A fun example of a bug becoming a feature": Destiny 2 accidentally made Exotic Glaives free to all classes, and Bungie says "we're going to let this ride."