Grinding Gear Games’ Chris Wilson and Jonathan Rogers presented back-to-back sessions at the NZ Game Developers Conference last Friday, each devoted in different ways to customer satisfaction.
Wilson’s session, What happens when everything goes wrong? examined a couple of occasions when the Path of Exile game decided to throw a wobbly for – apparently – no good reason. Rogers discussed the search for navigate the (potentially enormous) gap between customer expectation and what’s reasonable or even possible.
Wilson outlined the first of two events, colloquially known as “Kiwi halt”, which had – temporarily at least – brought Path of Exile to a grinding halt. Strange items started to appear in the game, skills that hadn’t yet been released started to appear, items began to appear scrambled, drawing names and appearance and skills from seemingly random sources. There was a helmet which looked like a fish and conveyed the ability to catch exotic fish.
Normally you’d have to spend a lot of money on drugs for such things, but not on 26 March last year. The challenge for Grinding Gear was what to do about it.
“Should we press the big red button?” asked Wilson. Taking down the servers costs money. Path of Exile gameplay is worth thousands of dollars an hour to Grinding Gear, but would the cost of fixing the problem outstrip those earnings? Maybe it would because, despite supposedly impossible things happening, the game hadn’t crashed. After an eternity (or 10 minutes in real time) the button was pressed, the cause determined, fixed. Life, in Wraeclast and Grinding Gear’s Titirangi base, returned to normal.
Wilson stressed that the three most important things were to keep calm (but not necessarily carry on); communicate with the game’s community, clearly and frequently while things were being sorted; and to deliver on what was being promised.
Do it again, Daddy!
“If you say you’ll ensure that something never happens again, you’d better make sure it doesn’t,” suggested Wilson. Unless, of course, it happens by design – which was exactly what some members of the community asked for after the event.
After the storm was over, the idea of having completely random combinations of item names, appearances and abilities suddenly became very appealing.
Wilson also noted that it was important not to establish precedents in the heat of the moment. It was especially important, he noted, not to look at a roll-back as the best option, since that caused negative effects for people who hadn’t been impacted at all by the original issue.
“Jonathan’s really good at writing apologetic posts,” Wilson noted, before sharing that when Grinding Gear had apologised and explained, confidence in the game had been restored and sales went up.
Rogers noted that the problems he was addressing were more of Grinding Gear’s own making – albeit inadvertently. Running low on cash during the development of the game, the company set up its own crowd-funding opportunity. Various levels of purchase and reward were offered. At the top was a US$1000 pack, part of which reward allowed people to design an item that would stay in the game for ever.
“We thought we might sell a few,” Rogers laughed.
The campaign raised US$2.5 million, of which around US$270,000 came from those thousand dollar packs.
“It might cost us more than $1000 to deliver some of those items, but that’s OK,” Rodgers said. Was it worth it? It was, reckoned Rogers, and not simply for the obvious reason that the funds raised allowed the company to get the game into release.
“If it goes right you get a brand ambassador for life,” Rogers explained. “Players will have ideas you wouldn’t have. And we would actually need to make new content anyway.”
However, delivering the rewards on the $1000 packs was causing grief. Grinding Gear is hardly alone in conducting a crowd-funding campaign without a full understanding of the amount of effort required to deliver the rewards but, as offering levels of reward in exchange for cash is an ongoing feature of the game, the company has had the opportunity to refine the scheme.
At its heart, the issue which led to a blow-out in the time consumed to deliver items was the level of input the purchasers wanted to have into the design process. And the fact that lots of people wanted their item to be the most powerful thing in the game.
While working through delivery, Grinding Gear increased the price of the most expensive pack (to US$1500). It reduced the number of sales but not the amount of grief involved in delivery. “We’re still working on some of the original items,” admitted Rogers.
Ensuring benefit for everybody was another change Grinding Gear made. Some of the unique items requested were avatars which, essentially, only the player using them gets to see. However happy that player might be, there’s no love from the rest of the community – either for the player or for Grinding Gear.
“Turn you player into a boss” sounded like a great offer, until it became clear that other changes to the game would affect the boss’s abilities. Ongoing tweaking of individual items to offset wider changes to gameplay was never going to fly.
Now Grinding Gear automates certain elements of the process of designing and approving individual items, to limit user input and prevent it from becoming an ongoing circular discussion.
As with the larger scale the game suffered with “Kiwi halt” and an item corruption event in May this year when things that had been bought or acquired literally vanished from players’ collections of gear, resolving the smaller if more frequent challenges of satisfying top tier contributors required the same approach: keep calm, communicate, and deliver what was promised.
And, added Wilson, don’t laugh when the boss is freaking out.
Where do the arachnids fit in? They were a request for one of the unique items: a barrel which, when opened, would release an unreasonable number of spiders.