Hold To Reset

Building a new game, a new studio and a new life from the ground up

Alex Hutchinson takes Typhoon’s deb debut into the playtesting pit

As we careen closer to finishing Savage Planet, we’ve moved beyond discussing the features we should include or the type of the challenges, and onto the more distressing question of whether after all that, it’s fun or even understandable to the average person. It’s a normal part of the process, but no less confronting for it, as we check our assumptions against actual player behaviour.

Finding a repeatable way through this part of the dev process has always eluded me. It’s easy to speak at a high level about what you want to achieve, but translating that into concrete tasks is a different matter. I have always felt there must be a more scientific way to approach it than what I’ve experienced over the last 20 years, but a useful example has been hard to find.

While thinking about how we’d tackle it, I was reminded of a quote from artist and Hellboy creator Mike Mignola. “My editor has suggested I say something about my approach to designing covers,” he writes in the foreword to Hellboy: 25 Years Of Covers, “but to a large extent that’s like asking about my approach to writing or laying out a page — I just bang away at it till I get something I like.” Similarly, after marvelling at the tightness of Blizzard’s games, I once asked a table of its designers what their process looked like. I had visions of spreadsheets and graphs and formulae, as their games can be complex, heavy in numbers, and yet feel pretty great. Straight from Rob Pardo, who was deeply involved in the tuning of StarCraft and World Of Warcraft, came the answer. Nope. They just bang away until it feels good.

Ignoring the terrifying leap of faith that implies, we simply do not have the time to run through as many iterations as that would take and worse, feels good to who? The first audience is always the team, and hopefully that’s a good initial test, but you also want to appeal to a much bigger and more diverse group of people. With that in mind, EA and Ubisoft and all the other major publishers have been investing vast resources over the

last 15 to 20 years in building playtest labs, partnering with external playtest groups, and studying usability and player response.

This has given big studios the ability to gather massive amounts of manpower and data, using both focus tests, where you sit a bunch of players down with a build of the game, watch them play and then quiz them about their reactions (you’d be surprised how many people will do one thing and claim they did something else) or ‘Kleenex’ tests, which are the same, but you test the participants once then throw them away (hence the name) to capture their initial responses. At Typhoon we’ve relied so far on a stream of ‘friends and family’ players, from both other studios here in Montreal and the more interested of our friends. It’s not a huge sample size, but it’s an educated assessment more often than not and useful for us to validate fixes or confirm problems. The lack of data volume, though, is not as big a loss as it sounds, as even with a large sample size it is up to the team to sort through what is just a broken, buggy or unfinished feature versus a real issue in your game, which is as much feeling as science. The bigger challenge is prioritisation, especially with a small team. What is a ‘must fix’ versus a ‘nice to have’?

Clarity is our first priority: nothing matters until the player understands what we’re asking them to do. We’ve been trying to find the sweet spot between highly directed gameplay and a freeform experience, which means endless revisions of the mission diary and the in-game markers.

Simultaneously we’re tuning the health of creatures, their damage output and the amount of resources they drop. Continuing our pitch to older players who still love the medium but have less time to play between careers and kids, we want everyone to finish the game, so we’re making sure there’s no grind in the mechanics and that we prioritise novelty and new moments or mechanics over repetition. (My producer calls this the ‘waste of money’ approach, which makes me happy.)

And then we’ll hit celebration. I want the creatures to burst like bubble wrap filled with green goo and a trumpet to play a short solo whenever you get a particularly juicy one. When you return to your habitat after being promoted, you should be showered in confetti while a celebratory tune plays. It feels appropriate for a mid-price game that tells jokes and relies on charm to carve out space beside the bigger, more epic and expensive competition. And besides, if you bought the game, you deserve a little confetti.

We’ve been trying to find the sweet spot between highly directed gameplay and a freeform experience