Co-Development: Open & Agile Game Development

Starr LongExecutive Producer

Greetings SXSW Attendees You are here to hopefully see me talk about what I call Co-Development: Open & Agile Game Development. If you were expecting another talk you are probably in the wrong room (unless I’m in the wrong room!) This is going to be a 40 minute talk so there will be plenty of time for Q&A at the end. Additionally my last slide will have my contact info for any follow ups you might have.

Starr Long● 25 Year Game Industry Vet ● Ultima Online ● Tabula Rasa ● Disney Connected Learning ● Shroud of the Avatar

My name is Starr Long and I’ve been making games for 25 years. I started in QA and after a few years was able to convince the bosses to let me direct my first project: Ultima Online which was one of the very first MMOs and is currently the longest continuously running MMO ever.

Since that time I worked at NCsoft and Disney.

For the last 2ish years I’ve been working on Shroud of the Avatar which is the 2nd highest crowdfunded video game to date at $7.5M. Today I’m going to talk not about that game but rather how we are making it.

Co-Development: Core Concepts● Transparency ● Cadence ● Feedback Loop

Co-Development for us consists of 3 core concepts: 1. Transparency: For us this means being incredibly open and honest with the community. No NDAs! 2. Cadence: Establish and maintain a rhythm for everything from emails to game patches. Be predictable on this axis 3. Feedback Loop: Listen and respond in very visible ways

Transparency● Daily, Weekly, Monthly,

& Quarterly Updates ● Forums & IRC ● Scheduling, Funding,

Staffing Details ● Unfiltered Responses ● No NDAs

Why? Transparency is the foundation on which this entire thing is built. Without transparency your users don’t have the information they need to give actionable feedback. Without transparency they do not have trust because they do not have visibility into what is happening. In our case since we are crowd funded this is critically important because our users are also the source of our funds so trust is critical.

How? We accomplish this transparency in a number of ways both formal and informal but the main one is a steady stream of updates on regular intervals. • Quarterly: Every quarter we provide backers a schedule for the next quarters 3 monthly releases that includes a detailed list of items we will deliver with those monthly releases. • Monthly: On the last Wednesday of each month we provide instructions for the monthly release that goes live the next day. In those instructions we list out all the deliverables promised versus delivered. We cut and paste the original information from the Quarterly update

and strike through anything we failed to deliver so players do not feel like we are hiding anything. • Weekly: Each Friday we put out a newsletter that is a summary of that week’s work. It includes previews of items going into the next release (more on that later), items being added to the store, community events, news articles, etc. It takes almost at least a full day to

pull together and is anywhere between 8 and 15 pages long. • Daily:

1. Standup Notes (formal): Every single day our team has a series of 3 standup meetings (ala Agile methodology) where each team member says what they worked on yesterday, if they are blocked, and what they will be working on that day. After the standups we post the notes we take to our forums so every single day our users know what every dev is working on. This lets them know EXACTLY how their dollars are being spent. In some cases (which I will go into more detail when I talk about Feedback) they also get to see how we are responding to their feedback

2. Forums & IRC (informal): Every single day someone from our team (often me) is on our forums interacting with users. This is often informal: liking a post, responding to a thread, sharing in a joke.

With all of our communications we do not filter our responses through PR nor marketing. We do sanity check with ourselves sometimes but mostly just unfiltered direct responses. Additionally we don’t make our users sign any NDAs and we allow them to repost content freely. We only ask that they make sure to note / mention that we are still in Pre-Alpha

Cadence● Daily, Weekly, Monthly, &

Quarterly Updates ● Monthly updates to the game ● Same day and time, no

matter what ● New Content, bug fixes,

polish, etc. ● When > What

Establishing a series of regular intervals for the project is good for both the team and the players. As I mentioned earlier these intervals can be about communication (standup notes, newsletters, etc.). However the most important interval is game updates. For us those are monthly on a Thursday at 10:30 AM. In general we aim for the last Thursday of each month but that varies slightly around holidays and such. Each quarter we lock down those dates and we NEVER waiver from them. We feel it is important to stick to that time and day more so than stick to a certain set of content. For us the “WHEN” is greater than the “WHAT”. Why? Because we think the users getting this constant sense of progress is one of the keys to Open Development. Without it they begin to lose that feeling of trust brought about by always knowing what is going on and where we are. It is also good for the team because it forces them to wrap things up to a deliverable state to a point that I never see when the content is only internal facing. After each release we do a video hangout post mortem where we weigh in with our observations, answer user questions, and provide a preview of any changes coming in the next release based on the current release.

Feedback Loop● Iteration ● Subjective + Objective Data = Iteration ● ~30% of schedule for response ● Recognize contributors

The final component is creating a Feedback Loop between the users and the team. To best facilitate feedback we try to get content / features into the game as quickly as possible with only enough content to demonstrate functionality. That way we can get feedback before a system is too far along to make any changes. Once we have content in the game we begin collecting feedback: • Subjective: This is written feedback from the forums, bugs reported, requests sent into support, etc. • Objective: Data about game usages (spells cast, enemies killed, time on tasks, etc.) We then take those two sets of data and correlate them. Items that have high correlation between objective and subjective data we schedule for upcoming releases and prioritize against other tasks. We schedule approximately 30% of each monthly release to feedback response.

Additionally we recognize individual and group contributors. Sometimes it is as simple as linking to a user’s forum post in a weekly update and mentioning that “we liked this idea that jelloface92 had so much that its going in next release!” Sometimes for us we make it much more formal. We actually created two awards modeled after Great Britain: • ONBE: Order of New Britannian Empire. This is awarded to individuals who contribute to the community as in very positive and visible ways. They get a title in game plus an actual physical medal that we mail them or better yet award in person • Royal Warrant: We award this to groups who contribute and/or lead in significant ways. The best example is the Bards of Poets’ Circle who formed to make music for the game and are now responsible for 90% of all the game music. The warrant we gave them requires

anyone wanting to contribute music to the game to go through them (vs. going directly to us). They then vet all that music based on guidelines and bounty lists that we provided. This public and specific recognition tightens the bonds and blurs the boundaries between the users and the developers to the point that ideally we all feel like one giant dev team.

For us the result has been one of the most consistently supportive and positive communities I have ever seen in 25 years.

Challenges● Iteration w/Feedback takes longer ● Vision vs. Feedback: Staying on Target ● Educating Users on Early Development ● Getting Discouraged by Negativity

Working this way is incredibly challenging.

First it takes much longer to do it this way then traditional methods. I spend at least 50% of my time just interacting with community and processing their feedback, gauging mood, etc. Then we spend time to react to that feedback (as described in last slide). The net result is a slowing down of about 20-30% however we hope we will gain that back on the backend with a shorter beta cycle but that isn’t yet proven.

Second it is hard to stay on target when your vision is being constantly questioned and critiqued. You can easily be led astray if your vision isn’t well articulated or you don’t believe in it enough to stay with it.

Most users have no idea what an early game looks like and the feedback they give is often focused on what they expect from a finished game. For us it usually takes 2 monthly release cycles to get users focused correctly and get them to understand.

Lastly it is really easy to get discouraged and lose momentum when users are constantly barraging you with critique. It is very hard to mitigate this and mostly you have to just stay strong in the face of the adversity. Belief in the vision as mentioned above is a strong defense here.

Q & A

Contact: Starr Long

Portalarium & Stellar Effect

slong@portalarium.com

stellareffect.com

shroudoftheavatar.com

Thank you for your time. Please take the survey!