Bungie outlined eight distinct design pillars when it set out to create what would become Destiny 2’s player-versus-everything-versus-player mode Gambit, but the only reason the mode became what it is today is because the dev team knew that breaking rules can sometimes be as important as making them.

That was the big takeaway from Peter Sarrett’s talk on the creation of Gambit during the Game Developers Conference this year, a talk that highlighted how many of the elements that make Gambit work as well as it does wouldn’t have come to be without a great deal of iteration and some healthy disregard for the rules.

As Gambit is today, two teams of 4 players are dropped into two identical (but separate!) arenas, fight through waves of enemies, collect the “motes” enemies drop, and stockpile enough motes to summon the round’s final boss. Dying with motes means they’re lost for good, but depositing a large amount at once can send additional complications over to the enemy team. It’s a sort of strategic race between the two sides to see who can pull this off first, and there are a number of ways teams can complicate matters for their competitors throughout the game as well.

The elevator pitch for Gambit “was Halo Firefight meets Super Puzzle Fighter,” explains Sarrett. They wanted to make an “end game ritual” for PvE players, equivalent to the Trials of the Nine competitive mode that experienced PvP players had to turn to. Gambit was meant to have elements of both halves of Destiny’s gameplay but “not enough PvP to scare off PvE players was our hope.”

It had to be something with enough variability to be replayable and fit into Destiny 2’s other weekly rituals. Compared to the PvE activities Bungie had already worked on, it needed to feel more replayable than Strike missions and have longer legs than the wave-based Prison of Elders activity from the first Destiny. 

Knowing when to back away from “lofty design goals”

The team laid out that they’d like two teams to compete against each other in two separate but identical arenas, but the prototypes they built by sticking to those pillars struggled to capture the competitiveness that they were going for in the activity. 

“If I can’t see the other team why do I care about them? We hadn’t come up with an answer to this and it was a really critical question.”

Some early versions had view screens or windows players could use to see the other team, but nothing quite worked. Looking back over their design pillars, Sarrett says it became apparent that “we didn’t have a good answer for [‘why can’t I shoot the other team?'] outside of lofty design goals,” so the ‘no direct player interaction’ pillar was cut and playtest feedback showed that Gambit was instantly better for it

Because of how the two maps were set up, letting a player from one team briefly invade the enemy side as a PvP threat was easy to implement and Sarrett says that adding a human threat radically changed the drama and tension. It took some balancing to find out how often players should be able to invade and work out the risk/reward balance for doing so, but once that balance was struck invasions became one of Gambit's defining features.

“We wanted to encourage PvE players who we knew we’d be courting with this mode to try invasion,” but it was unfortunately easy to make invading feel too risky for inexperienced PvP-ers. After some tweaking, it was decided that an invader would drop 3 motes if they died mid-invade because, while the team wanted taking down an enemy combatant to feel like an accomplishment, setting the mote reward too high made it too intimidating for the players they were making Gambit for.

“Ironically, we succeeded by failing,” says Sarrett. “We didn’t find a way to make the players care about the other team without direct player interaction, and it turns out that was the secret sauce that really makes Gambit stand out.”

Finding solutions in plain sight

With the PvP spirit of the PvEvP mode captured, the next challenge was to finetune Gambit’s core loop. It was a mode meant for PvE players, so the objective of killing waves of enemies was a given. Introducing motes added an extra level of complexity on top of just basic wave clear. Collecting and banking a set amount of motes before the other team was how games would be won, and adding a timer to how long those motes would stick around in the level helped Bungie force players to “un-turtle” and stay on the move. 

“Instead of just killing now you’re constantly cycling through [that loop], It makes for a much more interesting mode.”

Increasing or decreasing the number of motes an enemy would drop gave designers more balancing tools to play with (and, as an added bonus, the little explosion of bright-rewards from a freshly downed enemy gave players a rewarding dopamine rush!).

Depositing motes in a central bank let teams send ‘blocker’ enemies over to the other side, a concept that was essentially borrowed directly from Super Puzzle Fighter and prevents players from depositing motes until the enemies have been cleared. But, the simple goal of just quickly depositing motes didn’t quite feel as satisfying and fell short of the “comebacks are always possible” pillar.

“Something was still missing,” said Sarrett. “Ending a round by filling a bank often felt anti-climatic, the end snuck up on you […] and it was hard for a trailing team to make a comeback.”

In the final version of Gambit, disputing motes is a means to an end. Players bank 75 motes to summon a boss-level ‘primeval’ enemy that requires teamwork and strategy to take down, making for an intense end to a match. Originally however primevals were only a tie-breaking mechanic that popped up when both sides failed to fill their banks in time.

Making that battle the climax of each round gave players more situations to prepare for outside of just mob clear, brought the whole team together at the very end of the round to all around made Gambit matches more exciting and watchable.

“We realized that the answer to our problem had been hidden in the game all along and we just hadn’t played with it enough.”

By the end of it all, the Gambit team had created something that Bungie leadership dubbed “too much fun” to lock away as an end game activity, leading the team to knock down two more of its original design pillars. So, outside of knowing when to back away from your original design decisions, Sarrett offers two more pieces of advice: 

“Focus on the things players are already doing.” Simple is always better and using existing player habits to craft new experiences makes for something that is simple to pick up and understand subconsciously. 

“Players don’t always know what they want. No one was asking Bungie to make a PvEvP mode,” he says. “Sometimes its good to surprise and delight them with something they didn’t even know they wanted.”