Engineering the Service Mindset

Hi, my name is Charles Hollingsworth and I am the Engineering Director for Xaviant. I have had the opportunity to transition our team from building traditional data-driven client/server applications to console/PC games. We have accomplished this by leveraging the powerful CryEngine3 platform, training our developers and then augmenting our team with experienced game engineers. We have adopted a dual mission on our engineering team: developing great software for gameplay AND providing a rich, stable infrastructure that maximizes the ability for programmers to “program,” designers to “design,” and artists to “art.”

20121107-142941.jpg

Focusing on core competencies from our enterprise-scale, commercial software pedigree has allowed us to construct our (game development adapted) Lean-thinking house to accomplish our mission.

Although buzzwords and fluffy diagrams are always cool and certainly required for any self-respecting blogpost, I would like to specifically elaborate on the mindset and tangible systems that facilitate our commitment to this model. Developing engine functionality for spell casting, loot management, AI, etc. is the obvious responsibility of any engineering team, but the delivery, accessibility, and accuracy of this functionality is where we strive to differentiate ourselves. Specifically, in this post, I will explain our Respect for People pillar and give examples of how we remain committed to this principle.

Respect for People, from an engineering perspective, means that we are firmly committed to supporting those who create the essence of player experience. This requires a bit of humility which is difficult to swallow since it is obvious to everyone (in the universe) that engineering is the most important component of game development. Seriously though, it is easy to say we are “Design Driven,” but this is not easily quantified on a day-to-day basis. So, we have set ourselves up as a support group. We verbally and organizationally commit to “a spirit of service” toward our internal customers which manifests itself in things like:

  • Extending the classic software “continuous integration” to include art and design
    • Hourly builds
    • Each hourly build is stability-tested by automated gameplay scripts and level walkthroughs
    • Real-time visual feedback on displays throughout our work space (image above)
  • Prioritization of development tasks based on impact to artist and designers tasks
  • Avoiding 100% utilization so that we can immediately respond to technical roadblocks
  • Updating task tracking system so that internal “customers” know the status of their technical roadblock

Adopting this “support” mindset has allowed us to avoid the tension that often exists between engineering and design/art. Credibility is earned here by eliminating wasteful technical overhead so artists and designers can spend the maximum time “building awesome.”

  • Hourly builds of all code changes, including bug fixes, available to all artists and designers with a 5 minute installation
  • Instant feedback from our ErrorTrax game log monitoring
    • Detailed error information
    • Meta-information (location, camera angle, file information) for quick defect reproduction

Maintaining this level of service raises expectations among the entire team and provides a healthy feedback loop for requesting tools and optimizations to further enhance workflow.

Additionally, “gameplay developers” extend our engineering team by actually sitting in teams with designers. This proximity encourages rapid iteration and allows the developer to precisely experience the designer’s workflow to create best practices gameplay systems.

Our flat organizational structure requires that every team member is personally responsible for eliminating waste and increasing the speed and accuracy of the aforementioned processes. In future posts, I will discuss how we promote this “learning” environment and implement the other pillars of our team’s philosophy: agile development and continuous improvement.

Level Process: Game-play Collaboration (…and stomping the ego)

My name is Tim Lindsey and I’m the Design Director here at Xaviant. The articles that members of the Xaviant team are posting on this blog each two weeks are intended to give you a glimpse into our studio, our methods, or challenges and every so often, our products.  This week we decided to deliver a small video preview featuring a section of one of our levels.  The focus is on how the team members collaborate without killing one another. :-)

Commentary by:

  • Sam Elder on Level Design
  • John Grello on Concept
  • Dan Brown on Environment Art
  • Scott Warren on Lighting

As always, please click through and view the HD feed to appreciate our work. Thank you.

-tim

 

The Siege Report: Our Post Conference Post

My name is Tim Lindsey and I’m the Design Director here at Xaviant. The articles that members of the Xaviant team are posting on this blog each two weeks are intended to give you a glimpse into our studio, our methods, or challenges and every so often, our products. In this post I wrap up Xaviant’s trip to the Atlanta SIEGE Conference and Lichdom’s first public Pre-Alpha public testing.

siegetestroom01_v1

My initial attempt at writing this article was a summary geared towards a public audience and, frankly, was boring.  After some review I realized that anyone reading this far should really only care about one thing.  How is Xaviant using their SIEGE play-testers and the data the testers delivered to better their product?  Below you will find an internal email, quoted in its entirety, that I sent my staff Monday morning after the testing.  The message is honest, highlights where we must improve to impress our fans, and reinforces our core values as committed game developers.  Enjoy this peek into the studio culture.

From: Tim Lindsey

To: All Lichdom Staff

This weekend over a hundred gamers visited the Xaviant Play-test suite at the Atlanta Siege Conference.  This is a brief and, most importantly, honest update for all Lichdom developers on the experience.  There is no “blowing smoke” in this message.  After reading this please feel free to ask for more detail from any of the designers or myself.

Our Siege mission was to “build community.”  Instead of delivering a demo where the conversation can be very one sided we decided to host play-tests to encourage a cooperative experience.  To facilitate this experience volunteers received 20 – 30 minutes of game time and 20 – 30 minutes of interview time with Craig, Jim, Ben, Sam, Michael or myself.  There were always 3 designers in the interviews collecting notes and encouraging feedback.  To reinforce the community building goals we reiterated that testers were playing a “Pre-Alpha” so that Xaviant could respond more easily to the players concerns and interests.

The feedback was very consistent over the course of our 16 interview sessions and even continued to be consistent with impromptu interviews with other local developers from CCP Games, Hi-Rez Studios, and Tripwire Interactive.  Below are snippets of the feedback;
    • “This is like Skyrim.”
      • We are being associated with a AAA product.
    • “I want to dodge.”
      • Design for this in the works.
    • “I wanted to jump when I saw the scroll.”
      • Jump was a non-topic until players saw a location that felt like it necessitated a jump (the scroll on the ruins piece).  More work needed on cleaning up these heights.
    • “I hit Q and E by accident.”
      • We are fighting the taught norm on this one a bit.  More testing to come with a quick toggle between 1,2,3 and Q,E,R.
    • “I spammed AOE.” and “I didn’t get Cone.” and “I used Cone constantly.”
      • Believe it or not, this means we have different emergent play styles that players are disagreeing over.  I won’t say that we have our spell attacks perfect but this feedback is a sign that we are heading in a positive direction.  
    • “I liked the targeting.” and “I wish I could choose when to aim myself.”
      • Simply put, targeting was a topic when it failed to predict the target correctly.  It was appreciated for the majority of the test and allowed players to focus on the casting and the effects.  Speaking of effects…
    • “The Fire effects are beautiful.” and “I didn’t like the ice effects…I thought it would be more of a blizzard”
      • Fire is the bar for visual and auditory quality and was lauded all day long.  Ice lacks the impact of Fire in terms of game-play impact and vfx thus the comments.  We have some solutions for this on paper to discuss.
    • “I didn’t realize there was loot.”
      • Lots of work to be done on showing people what they have been awarded.
    • “More defensive abilities.”
      • There was a strong desire for spells that manipulated space and bought the player more time.  This feedback will very directly influence how our additional spells will be designed and balanced.
    • “The environments are awesome.”
      • ’nuff said
    • “I’ll buy it when your done.”
      • I heard this as people exited the interview and as I saw them in the halls.  This statement means two things to us right now.  First, this is someone saying that we’ve earned some trust.  Second, this is someone saying that our efforts have value – this game belongs in their collection.
    • In conclusion.
      • “It was fun.”
This is just the beginning, folks, we now have gamers (hopefully even Fans) interested in our ideas and looking for us to step up and deliver.
-tim

Play-Testing “Lichdom” at SiegeCon: October 5th – 7th

Xaviant is assembling a Play-Test Lab at the Siege Conference in Atlanta this year and anyone is welcome to schedule a 40 minute play-test session with us.  The purpose of this travelling testing lab is simply to get feedback on some of our core game-play mechanics and pre-Alpha content. Folks, we are early in our production cycle so what you will play is not a finished game but the test build is stable and enjoyable.  Since this is a testing session their will be a short Question & Answer session and Brainstorming with a Xaviant Designer that concludes the session.  Your ideas are welcome here.

riftwoodrays
  • Interested individuals will need to register at the Xaviant hall table to receive a time slot and review/sign an NDA.
  • Testers are expected to arrive at our table in the hall at Siege 5 minutes before their scheduled session time.
  • Xaviant will provide a few questions regarding game-play to keep in mind as you play.
  • After playing the group will be interviewed by a Designer and the group is encouraged to speak candidly.
  • After leaving the play-test lab testers are welcome to discuss game-play with other testers but we ask that you respect the NDA and not discuss or post game-play testing information, details or opinions online.

If your interested in helping us out as a play-tester then stop by the Xaviant table in the hall outside of the Conference sessions.  Sign up with someone at the table and bring some constructive criticism.  This opportunity will also give anyone testing an opportunity to talk to Xaviant Developers. We have limited slots per hour so please make sure you visit us early to secure your spot!

Thanks for considering the opportunity and we hope you join our Lichdom community.

-tim

Keeper of The Xaviant Philosophy

My name is Craig Lipinski and I’m the Production Director here at Xaviant. The articles that members of the Xaviant team are posting on this blog each two weeks are intended to give you a glimpse into our studio, our methods, or challenges and every so often, our products. This article focuses on our production methods and gives some insight into our company philosophy. I hope you enjoy.

I believe that videogames are a form of art. The art lies not just in creating something that is visually and aurally pleasing, but finding that elusive secret sauce known as “fun” along the way. Leonardo Da Vinci once said, “art is never finished, only abandoned.” Planning and scheduling something that is never finished can be quite the daunting task; and when you try, you begin to see why our industry has gained such a stigma for missed release dates and excessive employee overtime. Our core philosophy is that we will build a fun game, in a reasonable amount of time, while not taking over our employees’ personal lives in the process. Let me give you some insights into how we plan to accomplish this.

productionmapmq

First and foremost, we believe firmly in both individual employee and team ownership of work. We built our company by finding great people and then hiring the people that they recommended. We formed these folks into cross-discipline teams, each of which owns a specific aspect of the product (world building, combat, player reward, etc). In our previous postings, we discussed the creation of the pipelines that allow these teams to go off and create amazing things. They can do this because we allow them to chart their own path while we help guide them to the finish. We will always have deadlines, but this trust/ownership relationship creates an environment where employees figure out how to get things done and work hard to do so; and in return we ensure that the goals given are aggressive but reasonable.

Second, we are agile. Two of the primary principles of agile development are “inspect and adapt” and “respond to change over following a plan.” We set weekly goals, review the results as a team, and then quickly plan the next week’s goals. To help accomplish this, we place a major focus on playtesting. We hold daily internal playtesting sessions in which each team member is assigned one day a week to spend an hour testing the game. We collect feedback from everyone that plays, and then we inspect that feedback and adapt our designs accordingly to ensure that the gameplay remains fun. We still keep a high level plan with a targeted completion date, but the low level details remain fluid as we continuously respond to change along the way.

Third, as cliché as it may sound, we place a great emphasis on working smarter, not harder. Our first internal milestone after the creation of our visual targets was constructed to be a “team metrics” milestone. We pushed the team to see how much playable game they could deliver in a three month time frame. We used the data that we collected to create an estimated schedule for reaching the final gold date. We used not just the raw data, but also applied smart analytics to help refine the estimate. For example, we know that the amount of future content the team is capable of producing will continually increase due to factors such as content reuse, and the simple fact that the team is constantly getting better at what they do. In addition to smart scheduling, we also focus on smart engineering. This allows the team to work as unhindered as possible by technical roadblocks. Be on the lookout for a future blog post from our Engineering Director to learn more.

We are confident that we can deliver a high-quality, fun, and engaging experience in a reasonable amount of time without depending on an unreasonable amount of employee overtime. I hope that you’ve enjoyed reading this short insight into our methods and philosophy, and be sure to keep checking back here in the future to learn more!