Boss For a Day. . .

Chip Lambert won our "Boss For a Day" contest award. He was selected randomly from a pool of users who sent a Tweet out announcing our Torque 3D 1.2 release. Chip is an Applications Developer with Bluefield College and an independent game developer. Chip has been a member of the GarageGames/Torque community since 2001.

We had a number of goals in mind for running this contest. First we obviously wanted some free organic marketing from users tweeting about our release (thank you to all who participated). Second we wanted to give some real value to the winner (Chip will determine whether we succeed there). Finally we felt that from much of our interaction with community members in the forums, via email, and in our IRC room that one of the largest obstacles to people getting a finished product is not a lack of talent or resources but rather a lack of clear Project Management. We know that today Chip is going to get a lot of real work done for him. But we believe that moving forward with the help of a detailed plan beyond today that Chip will have everything he needs to take his project (it’s really cool by the way) to a deliverable state.

As the Operations and Development Director here at GarageGames it is my job to run Project Management. At its core I am a traffic cop and air traffic controller who coordinates the efforts of our programmers, developers, artists, QA team, tech writers, marketing and community support team across all of our products. The short version is it is my job to help schedule WHAT is getting done by WHO and WHEN and in what ORDER.

In Project Management there are some common terms that get thrown around for the process of determining the project plan: goal, milestone, sprint, priority, resource, estimates, dependencies and risks. These are the critical pieces that we are focusing on for Chip’s project and all projects at GarageGames. Next are the steps or phases of the actual sprint. For today we’ve had to shorten what would normally cover a week or weeks into a single work day.

The first step is to determine a Goal and to prioritize the details of the Goal. The Goal leads to the actual Milestone deliverable for the project. This deliverable defines the WHAT of the project. Chip provided us with a list of Goals which Eric then defined tasks from. We have divided these tasks into 3 priorities – High, Medium, and Low. Once the priority has been set for the tasks, we can hold the Sprint Kickoff meeting.

We started Chip’s day with a Sprint Kickoff meeting where we determined WHO would be assigned each of the tasks for the sprint. Once the tasks are assigned to resources estimates were provided for the tasks. With these estimates, I built the project plan which gave us the WHEN of the Sprint.

The day was then broken up into a number of cycles: Development, Review, Reaction, QA, QA Reaction and finally a Release. Now, given that we are doing all of this in one day, the Development cycle covered the majority of the day at 5 and half hours, with the other cycles all being one half hour each.

When everything was wrapped up into a project plan and schedule, this is how the resource allotment worked out: 6.5 man hours of engineering time, 6.5 man hours of design/production/scripting time and 16 man hours of art time. All of this before we spent any QA time on the project. All said, we spent 29.5 hours on Chip’s project before we took time to run a quick QA pass on things.

Now take a step back and think about how this looks when a sprint is not 1 day with a team of 5 or 6 but rather 7 to 14 days long with a team of 10-12 engineers, 5 artists, 3 technical writers, 3 production staff, a product manager and a web developer…all who have tasks across all of the Torque products new and old. That’s where the air traffic controller part of the Project Manager comes in. And I haven’t even mentioned dependencies or blockers yet. A good Project Manager has to stack all of the milestones and the tasks in each project, all in the right order to make sure nothing crashes. This involves lots of communication via face to face conversations, instant messages, phone calls, meetings and task updates.

What are the critical tools at my disposal? MSProject (the dark and arcane tool), JIRA, MSExcel, Visio, a good shared document system (Google documents or our Confluence WIKI) and communication. Each of these serves a different role in project management and together they help me build the entire Project Roadmap to present to management and to the team.

The final tool that I have used to great success here at GarageGames and elsewhere is the dreaded Project Post Mortem…but that’s a subject for a later blog.

We can’t and won’t share everything about what we have done today for Chip Lambert to protect his intellectual property. But look forward to more Project Management blogs covering the WHO, WHAT, WHEN, and HOW of video game development.