Path of Exile - The Creation and Future of the Royale Mode

Grinding Gear Games' April Fool's event that added an optional battle royale mode to their action-RPG Path of Exile ended up being quite successful - 27,989 free for all matches were played during the 31 hours this event was running. And while there were some balance issues and some usual grumbling, a lot of people enjoyed this new PvP mode. If you're one of them, this developer post shares a few tidbits about the mode's creation and its possible future:

The Creation of PATH OF EXILE: ROYALE

As we're sure you have seen, battle royale games have been taking the game industry by storm over the last year. Recently, online games have been scrambling to add royale modes as quickly as possible to capitalise on the fervour. A couple of weeks before April Fool's day, it was becoming clear that our existing plan for a joke wasn't going to work as well as we thought, so we needed to come up with another idea.

The idea of announcing a battle royale game mode as a joke was pretty obvious (as evidenced by several other companies who had similar ideas), but Jonathan realised that we could actually implement the whole thing very quickly, creating an actual game mode for people to play for the day.

Our plan was to make the following changes to Path of Exile:
  • Create a gigantic island with a bunch of tilesets jammed together.
  • Populate the island with monsters and chests full of items.
  • Use the existing cut-throat rules combined with a shrinking play area.
  • Add support for up to 100 players per game instance. The existing limit was 32, as we used to pack the visibility data for a tile into a 32-bit integer.
  • Make some server optimisations to distribute the large Royale instances over servers uniformly.
  • Add a quick spectator mode for players who have been eliminated. (This is different to the type that would be used for race events in the future)
  • Repurpose an existing but unreleased Rhoa Dinner hideout decoration as a prize.
Each of the above tasks applied to a different developer and looked like it would take less than one day. We decided to set a limit of one workday per person who wanted to work on it, so that our progress on the next content update wasn't disrupted too much. This limit was mostly respected, though the definition of "workday" was stretched considerably for one programmer.

As soon as Royale was playable, it became a huge sensation around our office. Whenever people were playtesting Royale, it drew large crowds of people watching.

Our plan was to deploy it at noon New Zealand time on Sunday April 1, which was around the start of April 1 in Europe and pretty late on the day before in America (so that it was clear it was a joke). We weren't sure how servers were going to perform, especially with instances that were this large, so we overspecced our server hardware just in case. We did not want to take the risk of having the new mode run into capacity problems or affect the regular game, especially on a holiday weekend (though we did end up having one realm issue, despite our efforts!). Now that we have run Royale for a day and a half, we have a lot of useful data that will help us predict server usage if we run it again in the future.

We launched the mode on time and it was a huge success on two levels. Most users loved the joke at face value, but also discovered that it was actually a real game mode they could play, and one that is a lot of fun. We also renamed the game on Steam to "PATH OF EXILE: ROYALE". Over the 31 hours that it was running, 27,989 matches of Royale were played. We had a great time running the event and watching the community's reaction to it. Many laughs were had.

What's the Future of Royale?

We have been inundated with requests to bring the mode back. It's a foregone conclusion at this stage that we'll be running it again in the future, but there's the big question of whether it can sustain itself enough to become a standalone game mode, or whether we should just have it as a cool event that we run occasionally. In either case, we have a lot of balance improvements planned that will make it more fun and less overrun by certain dominant skills.

As a community, how do you feel we should proceed? Is it worth the resources to take this one-day prototype and polish it into a standalone supported game mode with ongoing improvements? We'd love to hear your thoughts.