Path of Exile: Designed to Be Played Forever

Watched Chris Wilson’s GDC2019 talk the other day:

A lot of eye-opening insights in here:

  • the standard population decline of any online game when they first launched and how they got it to spike consistently and even grow over time through their league seasons
  • a quick look at their custom tool for procedurally generating interesting map levels in really short time frames
  • the importance of marketing and having enough content to market to different subsets of players to make a big enough impact to prompt returners to return
  • the importance of consistency and predictability to cultivate a customerbase (or else they will look elsewhere and get distracted and then you’ve lost them.)

A couple of his points I don’t necessary agree with, or think might work for -every- game out there, but perhaps are more game and population specific:

  1. The idea of investing time to design aspirational content for the 5-10%, knowing full well the majority of their customers will not reach it, but creating this content for the 5-10% to feel special because no one else can get there, and those 5-10% tending to be the more hardcore influencer types who stream and thus draw in hopefuls and additional player numbers
  2. Economy resets so players can start on a fresh playing field periodically
  3. Layers and layers of randomness to create interesting variability
  4. Avoiding day-night cycles so that assets can be re-used
  5. Designing spare assets to sit around in a warehouse/library so that they can be pulled out when there is a need
  6. Avoiding pipelining releases so that people aren’t distracted working on two things at once, or tempted to avoid dealing with a tough problem in favor of something easier

Point 1 always raises my hackles. My opinion is that it works for games that start out designed that way, so they attract a playerbase that accepts that premise from the get go.

Something like Warframe apparently attempted large group raids and later removed them because apparently too few of their playerbase was interested, they seem to be doing better investing effort into content that both groups and soloists can do.

As for GW2, well, their “little” u-turn and about-face during Heart of Thorns introducing aspirational raid content lost them the better part of their initial playerbase, and attracted a newer, more competitive, and hostile sort of player in lieu. Hopefully they pay more. Else it was a really really bad strategic decision, no?

Path of Exile on the other hand is built around the idea of competition, of races, of getting to level 100 and feats of getting somewhere “first” broadcast to all and sundry. It has a hardcore permadeath league mechanic for the challenge seekers. So yes, logically aspiration works for a primarily competitive, challenge-seeking, numbers-crunching playerbase that can deal with that PoE skill tree. Somehow, I don’t think playing PoE to “relax” is a majority motivation here.

The solo self-found playstyle was more of an underground subset of players who chose to remove themselves from this competitive economy and create their own fun – it’s only recently they gave a nod towards it by delineating a separate group to declare oneself that way. The stated rationale is for bragging rights, and they are very careful to assure players that you can jump back into the economy any time you want; separately I suppose it is also a way for them to keep tabs on just how large or small this hermit-like player subset is. (SSF all the way, huzzah. Fuck yo’ aspirational content.)

In theory, I really like the idea of Point 2. I was first introduced to the broad principles of resetting in MUDs that had something called ‘remort.’ You reach max level (ie. near immortality), then you ‘remort’ (become mortal once again) to level 1 and get to level up again, but with some bonuses for choosing to reset yourself that way.

For some games, this works and comes as part of the game. Kingdom of Loathing is a browser based game that uses the remort mechanic. A Tale in the Desert has an extended long reset with new Tellings. There’s that One Hour, One Life game I never tried, but the reset concept is right there in its title. You can choose to reset almost every single piece of gear in Warframe with forma and level it up again so you can cram in more and better mods to make it even stronger.

For other games, I don’t know if their playerbase would recoil in garlicky vampiric horror at the concept of being set back to square one and starting anew. I understand that World of Warcraft tries to reset gear every expansion – from an outsider’s POV, it seems to be a 50/50 mix of acceptance and frustration among its populace. GW2 resets WvW in varied intervals and it seems most players have gotten numb to the resets over time, as winning means very little. Still other games are all about the collection and character/account progress, and I doubt those players would be happy with a reset – does Monster Hunter World or Final Fantasy 14 reset anything?

Point 3 I also like on a personal level, it’s a very roguelike foundational concept, and I love me a whole bunch of roguelikes that can offer me procedurally generated layouts that allows me to have a different and strategically interesting time each playthrough. Playing through City of Heroes near identical and unvarying tilesets and fixed predictable spawn size for 4 years will do that to you.

But not every game can be a roguelike/sandbox type of game where the player is expected to react with the resources available and create their own story. Some games are more linear, more dev-created story-oriented, and handcrafted, hand-placed content still has that level of uniqueness that can break the pattern recognition of players reacting to procedurally generated stuff. It’s just that handcrafted stuff takes a lot longer time to create.

Some games do try to mix the best of both worlds. Don’t Starve has handcrafted set pieces mixed in with procedural generation, and a bunch of Minecraft mods also do the same thing, sprinkling in handcrafted stand-out pieces and allowing the general landscape to be procedurally generated.

Which I suppose point 3 also covers, this idea of mixing and overlaying random stuff atop of random stuff, so that it is harder for players to discern predictable patterns.

Point 4-6 sound very much studio-specific and game-specific decisions, so I won’t comment there.

Still, it is interesting to learn what he feels works well for Path of Exile.

And I really want to sit in on a three hour talk to hear what he thinks about loot and itemization.