This week BarrelEye made steady progress on Torch, building up our momentum for the next three weeks, which we expect to be the busiest part of our semester.

Our week began with our second guest playtest. After the EA employees we invited were unable to attend, we sought out fellow ETC students who had not tried our game yet. We arranged a focused test with a group of three players, giving us the freedom to hold an extended interview with the group afterward. Two of our testers self-identified as non-gamers and one identified as a gamer, giving us a good group dynamic that may be reminiscent of the show floor audience we are targeting. One of the most interesting results from this playtest was how each player interfaced with the game differently – it’s becoming clear that our controls will require lots of iteration.

During the week we had multiple opportunities to demo the game to our client as well. In each case they confirmed that our progress is visible and they think we are headed in the right direction. However, we still feel that we are just beginning to find the fun in our game. Feedback from all sides has suggested that players only need to feel lost for a few seconds at a time to create a satisfying contrast when they enter the light, and that we can push them toward success and more interesting interactions sooner. It’s all about timing! To help us rapidly iterate on the level design and explore that timing we put together a simple map editor for the game.

In addition to adding features, our programmers did lots of planning work this week. They jumped into the details of our network architecture and pinned down their strategy for integrating the many platforms and technologies we are using. We also pulled in technologists from the OCCO to review our networking plan and give us feedback on our approach. These partners have committed to provide us with ongoing feedback and assistance as we uncover new engineering challenges with our platform. Another bit of good news is that we’ve received a few more details on the smart TV hardware we will eventually deploy to, which has allowed us to review our plans for feasibility. So far we think we will be just within the capabilities of the TV, with a few “bonus” features set aside to explore if we have time.

Thursday marked the beginning of our third work sprint, and the team carefully reviewed our remaining tasks before Halves and distributed the work to make sure everything will get done. Our biggest concern by far is the unforeseen implementation challenges the new smart TV hardware will bring, so we may have a little crunch time next week to solve those problems as early as possible. Otherwise we are on schedule and having a lot of fun as we watch all the elements of Torch fall into place. There will be a lot to share in the next newsletter! Until then, have a great week!

Photos © Jaewan Park.

 NEWSLETTER #7  PDF FORMAT – DOWNLOAD

After a week of planning, the team started in on our next phase of development with enthusiasm. We began by making as many improvements to our Unity prototype as we could, so we would have something more impressive to show later in the week. In addition to small changes to our camera and control scheme we made big strides by implementing a player objective, adding 3D models for characters and some of our props, and moving several of the game attributes into a config file that our designers can modify directly.

At the same time, we launched into a deeper exploration of the TV and its capabilities. Our engineering group had huge scope concerns about building a game engine on top of Marmalade, so we started looking for tools that would get us started. We discovered IwGame, a free game engine built on top of Marmalade that has just the basic features we need to get started. Our team tested an IwGame sample on the TV and started to get familiar with the system so we can benchmark its performance next week.

Even with IwGame, we discovered that our workload was too great to create the alpha we wanted to by October 24th. So we approached our client now and asked for an additional week – a new alpha deadline on October 31, two days before our mid-semester presentation. They happily agreed, glad that we were planning ahead and wanting us to produce our best-quality work rather than cut features and rush the game. The schedule is still going to be tight, but it’s now within our reach.

With that in mind, we’ve been more proactive this week about reaching out to the OCCO for help. We’ve enlisted the assistance of Roger Powell to help us with our game audio. He asked us to provide him with a small sample program that would work on the TV, then started exploring that audio capabilities of the platform on his own time. We’ll be working with Roger sometime next week to start putting audio into our test program. We’ve also spoken with Joe Sutton about our networking strategy for the game; he advised us to show that Unity can talk to Marmalade before we get too far into our network plan.

In keeping with the plan we outlined at Quarters, we arranged a small playtest on Tuesday in the ETC space. This was an internal test without specific guest testers. Our own team tested the game at first, but it didn’t take long for our advisors and fellow students to jump in. It was encouraging to discover that our small changes to the game had drawn people’s attention to the TV more, which was one of our goals. Players also expressed a desire for the game to move faster, and for us to explore alternate control systems. We left the playtest with a small set of action items. Though quick and casual, we hope these informal playtests will keep our team focused on the player experience.

On Wednesday we had another informal playtest, this time with Rich Hilleman (Chief Creative Officer) and Jeff Skalski (Producer, BioWare Mythic). They were both pleased with our progress so far. Rich reminded us that we must create a dense experience out of necessity, and challenged us to reconsider a few of our design decisions from that point of view. He also emphasized the value of making the player feel silly, both in the context of the show floor and for players standing side-by-side. Jeff echoed this, reminding us to play with the social dynamics our game creates.

The rest of the week rushed by. We adjusted the Unity prototype in preparation for a playtest on Monday, and had a chance to show Drew Davidson our progress. We also kept exploring IwGame, putting together a load-test app to find out how many animated sprites the TV can render before we run into performance issues. Finally, several members of our team were out on Friday to attend IndieCade, the International Festival of Independent Games. We’ll start next week with a guest playtest and by the end we’ll be into our third sprint. Now it’s back to work at BarrelEye…

Photo © Eren Song.

 NEWSLETTER #6  PDF FORMAT – DOWNLOAD

It’s a great feeling when a project has a week of planning. It’s an opportunity to find a stable spot, get oriented, and take a collective deep breath before continuing on the journey. BarrelEye’s fifth week, though busy, was full of reflection and preparation.

After a successful playtest last Friday we took some time to review the feedback we had collected. Our guests had provided us with a wealth of suggestions and we walked through the list together, discussing the changes that made the most sense for Torch. We generated a yes/maybe/no list with reasons for many of the design decisions we made. It’s nice to have the whole team invested in the design.

We also spent a few days preparing for our quarter-semester presentation. As part of this process we examined the work we have done, so that we could explain it to our audience. We also looked forward to the next twelve weeks, to describe our roadmap to success. In the process we discovered that while the last four weeks of planning and prototyping have been productive, it is time to lay the cornerstone of Torch and build.

In addition to a successful Quarters presentation, Wednesday marked the end of our first work sprint. After the presentation the team convened upstairs for a sprint retrospective. We spent about 45 minutes discussing the state of the project, what we felt was working and what we needed to change. This was followed by discipline-specific meetings to discuss and estimate the features we want to include in the game, and thus need to build during the next two sprints.

The next morning we met to formally plan sprint #2. With our tasks divided into stories (or features) we selected a subset to focus on during the next two weeks. We hope that set of stories will be within our reach, and that it will best represent our vision for the game at our next progress update on October 10th. We spent the afternoon getting organized and doing individual task estimates for our next burndown chart. The time spent planning is worth having a clear sense of the work ahead of us.

Today we turn the corner into the next stage of our project. It was a busy week and while we’re not exactly refreshed, it’s nice to have a plan and move forward with confidence about our direction. We’re a little intimidated by the amount of work ahead of us, so we’ll be reaching out to the OCCO for support as much as possible. Next week is all about building the foundation of Torch and investigating the least familiar parts of our plan, especially the networking and graphics technologies we are using. See you then!

 NEWSLETTER #5  PDF FORMAT – DOWNLOAD

“He dares to be a fool,
and that is the first step
in the direction of wisdom.”
– James Huneker

The team shared a nervous energy this week as we prepared for our playtest. We had made plans, prepared our pitch, promised a revolution and now – a meager seventeen days later – it was time to reveal our hand. Our plan was to demo our first digital prototype, and we had boldly sent out a public invite both to our fellow students and to the EA team. “We’re using Unity,” we thought, “how hard can it be?”

Of course, it’s never that easy.

On Monday and Tuesday our engineers spent all their time working through Unity networking to set up our prototype. Every few hours they would have a breakthrough and would celebrate, until they realized they had uncovered three more problems to be addressed. Networking can be tricky business.

By Wednesday we were starting to worry. The prototype was working, but it was all grey floors and grey walls and cube people. We understood the game, but our guests would not. Art started dropping temporary assets in anywhere they could, and by that night the game looked quite impressive. We expected to leave our guests amazed.

We knew that Thursday would be our final push, because Friday morning would not afford enough time for fixes. After attending morning classes, the team gathered to test out the game one more time. That’s when we realized it wasn’t working on the phones. Our engineers spent hours trying to track down the problem, then rolled back to the last working build, then spent more hours reconstructing crucial elements for our playtest. Art could only continue working on research and presentation materials, and wait patiently for an answer. At ten o’clock we were back to cube people and grey floors, and we’d thrown out so many other features we wanted to show. But we agreed to stop – the build was stable, and we weren’t about to tempt fate any further.

The morning of the playtest we prepared with mounting anxiety. Somehow the prototype felt sub-par, having watched a more polished version slip through our fingers the day before. Why had we ever invited our client to this, our most vulnerable moment? How would they respond to boxes sliding around on a screen, when just a week ago we were showing beautiful concept art?

So, somehow, it was a surprise when our guests showed up and incredible feedback started pouring in. People understood the prototype nature of our test. They picked up on what the shapes represented and got the hang of the experience faster than we expected. Best of all, some of them started to have fun! We should have known – we did invite game developers to our test, after all.

We came away from our playtest with a mountain of feedback that we will sort through next week in preparation for Quarters, but also with a new appreciation for the value of early testing. It wasn’t easy to show our game knowing that it was so incomplete, but we would have missed out on so much insight if we had not done so. If our next several playtests are half as productive as this one, then Torch is going to be a great game.

 NEWSLETTER #4  PDF FORMAT – DOWNLOAD