Update 9 Test Report


  • Shift from board gamey roll-to-move and heavily random events to a more static, dungeony, linear approach to map design (including removing the deck based event system). New map style features days that get drained by resting/dying rather than turn timers.
  • Completely dropped support for a PVP mode to focus entirely on co-op.
  • Switched back to a diverse set of classes that fill very specific roles (partially based on previous versions of co-op classes)
  • Complete redesign of user interface more in line with PC-style hotbars rather than vaguely console-styled submenus, as well as including information menus accessible even when it isn’t your turn. Switched types to Rock Paper Scissors and a more visual display of types breaking other types during attack animations.
  • Equipment Point progression system changed to primarily feature upgraded versions of core class abilities rather than entirely new abilities.
  • New equipment with more synergistic abilities for the classes.
  • Completely new art style (including new map tile style).
  • Honestly since this update took over a year, probably things I’m forgetting.

Test Results

It’s hard to even process the response at this point, I guess. Testers who play rarely were still initially confused. Some of them seemed to get over it far quicker than they used to (whether from new UI or experience is another question). That said, I’m not drawing too many conclusions from it since part of the problem was the test scenario’s difficulty being cranked up fairly high since I wasn’t originally planning on wider testing, which makes the learning curve frustrating. Most confusion problems can really be placed into one of 3 categories: UI needs an artist pass to be pretty and also emphasize what’s actually important, missing effects (sounds, visuals, etc) create confusion, and no learning curve in the game’s content. These are all things that just need done, rather than requiring an overhaul.

While the response wasn’t overwhelming, the smaller details of how testers played were far more encouraging than previous tests. Players discussing the equipment they wanted, actively seeking out EP abilities without prompting, refining party builds between playthroughs, hunting for secrets in the choices, reacting to some of the bigger surprises, wanting to play again until they beat the demo, etc. It was a lot like watching people play a game. It’s pretty clear the co-op focus was the right choice. Timing wise the games consistently hit 1hr 20mins for experienced players, which is still higher than the target time of an hour. Additions like simultaneous turns will probably speed it up, and beyond that I don’t care quite care as much about perfect lengths now that co-op is the focus.


Reading the last update report’s next section is distressingly like looking in a mirror. A year later we’re still miles away from finished, yet still purporting a “get it done” strategy. The reality is that 80% of 2016 was spent on the UI overhaul. That remaining 20% has been frantically spent changing our course to get the game finished.

What has changed is that we’ve fully dumped the PVP mode in the name of focusing on what the game is doing best right now (which is pretty much combat). From there we’ve vastly simplified the map to create a more traditional RPG experience. For all the years spent trying to make the map work, the reality is that it had too much going on which made it vie with the also complex battles. One had to die for the other to live, and we chose the one that’s already good.

What’s next, then, now that we have a working prototype with Update 9 is to change the map presentation to fit its new systems. A change I loathe doing when we need to get the game done, but necessary. It fits alongside the other major architecture change of making the map support simultaneous turns. From there, it’s a matter of finalizing the mechanics, planning the complete content of the game, and then implementing it. Once the foundation is rock solid we can finally start full production on the assets, add missing features (saves, settings, etc), and generally just GET IT DONE.

We’re not in a good state, but it’s better than it was.