Monday, May 1, 2017

GDC 2017 - Art Leadership Roundtable - Day 3


Despite being the final day of the conference, the room was packed once more with attendees.  Since Day 2 had been successful with audience-sourced topics, we did the same thing on Day 3.

The first question touched on aligning vision.  Specifically, how do you align the vision between different people who have different interpretations of a vision.  The attendees offered the following suggestions:

  • The first and most obvious suggestion was documentation.  The creation of clear style guides and/or reference material is the quickest way to communicate intent.  While some flexibility may yet remain, the style guide should establish the boundaries in which to operate.
  • Beyond the style guide, it was also suggested to create and document benchmark assets for visual style.  In addition to showing successful examples, there should also be benchmark examples which show why particular examples do NOT work.
    • It was also suggested that benchmark assets not be presented as solely screenshots but rather also the asset files (models, textures and animation) should be archived and openly available.
    • Furthermore, it was commented that visual style examples do evolve over a longer production cycle.  As such, it is worthwhile to update the benchmark assets and documentation as needed.
  • From a technical standpoint, there should also be reasonably clear specifications and limitations so that all content fits within the performance threshold.
  • Training videos was also another strong recommendation.  These should focus on how content is made and, equally important, how crafting choices are made.
  • Outside of documentation, there are several strategies that can be employed to address divergent interpretations or conflicting visions.
    • Organize group critique sessions.  These should be moderated by a Lead, but foster an environment of open discussion.
    • As an early-stage exercise, establish the goals and then have the individuals assemble reference.  After a time, have the group reform and compare reference.  What choices were similar?  What choices were different and why?  Use this as a springboard for dialog, but don't forget to narrow and align the vision.
  • There were also a few production recommendations offered.
    • During pre-production, create a small subset of atlas materials that adhere to the color script.  This allows the team to focus on modeling and world building without getting lost too soon on material definitions.  A cohesive material/color library will help to identify modeling inconsistencies.
    • As a director, don't get stuck micromanaging individual content.  Where specific assets run astray, delegate to the Lead or Senior artists to identify and resolve.  Instead, as the director, puts lots of images on the wall (concept, production assets, etc.) to build a library of visual style that everyone can see while walking through the office.
    • Take the time to ensure that the pipelines are capable of achieving the desired visual style.

The second question focused on team organization.  How do you assemble your teams?  Do you localize groups based on similar skills (ex. concept artists all in one room) or do you organize around goals?  Or something completely different?  The following comments and criteria were expressed.
  • Generally speaking, artists do want to work together.  They want the opportunity to inspire, motivate and learn from one another.  Related to the previous topic, this is also the best environment for creating unity around a vision/style.
  • It depends largely on the attitude of the group and how communicative they are.  A high-performing group which makes a conscientious effort to interact with peers in other departments would work fine in a more isolated work space.  However, an insular group who possesses a more withdrawn attitude are better integrated directly with their client group(s).  Seniority and maturity are likely to be a factor here as well.
  • Another criteria is the number of artists as well as the number of projects.  In some cases, scale of the organization may limit your options.
  • Regarding layout, it was recommended that you consider the spatial impact.  Organizing groups by job type can sometimes have the unintended consequence of creating silos and expectations that run contrary to collaboration.
  • In addition, there was a warning against the development of a "ticketing system" approach to development.  When groups are isolated, there's a risk that lose an understanding of the context that drives their work.  Development may take on an assembly line mentality where one is not desired.
  • Speaking in favor of organizing by craft, the quality of the individuals of the group are likely to grow at an accelerated rate.  Being surrounded by similar developers, the individual is far more likely to ask questions and pick up new skills in this environment.
  • One suggestion was to organize based on which communication is more valuable to the organization.  Do you desire/expect to see more communication within groups or across groups?  Choose your spatial organization based on the goals.
  • It is also important to consider the stage of the project.  For example, organizing by craft may make the most sense during pre-production when there's a lot of prototyping and testing.  However, production stage may require a reorganization into cross-discipline work groups.
  • It was also suggested that you not approach this with a binary mindset.  Different structures may be more or less effective based on different groups.  Take the time to identify the goals and develop a strategy; don't get in the habit of changing and over-correcting based on tactical needs.  Moreover, consider getting the groups impacted involved in the discussions around these decisions.
  • Is there a middle ground here?  A few people suggested that groups be organized around feature or product, but that other strategies be employed to preserve the type of culture you would desire from a group that was organized by craft.
    • Preserve and promote regular critique sessions within a project or across projects.
    • Invest in art-centric mixers to promote visibility and the opportunity to ask questions/explore.  Drawing clubs or art challenges.
    • Look for cross-discipline culture activities - game jams, rapid prototyping, experimentation
  • Ultimately, the leaders of the project/organization are the ones responsible for building the community and bridging communication.  Regardless of spatial organization, the way the leaders interact and behave sets the tone for communication.  In short, lead by example.
The final question of the day was asking how do leaders/directors want to grow?  What can managers do to assist you?

  • One response was that the leader often gets bogged down in non-critical meetings.  Having a manager stand in would free them up for more time creating or reviewing art.
    • Another attendee countered that they wished their art director would send leads or senior artists to meetings more frequently.  They would appreciate both the perspective as well as the higher-level visibility into development.
  • One lead expressed a desire to periodically step out of the lead role.  They would appreciate having time set aside where they can focus on content creation so that they don't lose their perspective or craft skills.
  • Lastly, one attendee mentioned that they value the mentorship they've received, but that almost all of it had been about leadership.  They would value having the opportunity to receive mentorship centered around different artistic craft skills.

That concludes the notes from Day 3 of the Art Leadership Roundtable.  Thank you again to everyone who attended.  I hope you found the sessions valuable and that these notes are useful.  Hope to see you all again next year!

Speaker Evaluation

Art Leadership Roundtable: Day 3

Thursday, March 3 at 10:00 AM

Room 120, North Hall
------------------------------
--------------------------------------------------------------------------------------------------
Total Headcount: 90

Roundtable Session Ranking within Visual Arts Track: your session is ranked 4 of 12

Roundtable Session Ranking within GDC 2017: your session ranked 43 of 421

Session Totals (This Session)
Response
Count
Percentage of Responses
Excellent
20
83%
Good
4
17%
Poor
0
0.00%
Terrible
0
0.00%


Comments

Loved these sessions!
Great host! So valuable!!
I wish these were longer! Ideally we timebox topics and, at times, contributors.
Great facilitation, esp. when picking people who haven't spoken as much. Also, I recommend to never pick more than the first 3 answers and then re-eval who wants to add something new to the table.
Please, more art related roundtables. Lots of valuable information on these talks. Would be great if the elaboration time was cut in half for each question though, because it sometimes felt like we were going in loops and would have been nice to address more issues with the time we had.
Like a AA meeting for different types of leaders. Love it!
I love Keith. He creates a safe space to speak and encourages lively conversations. He's great at summarizing points when the main thread is getting lost.
More facilitation is needed. We had some one sort of hijack the last question and change it to a question that severed mostly themselves.
We spent too much time on the first question about aligning artists when it felt answered in the first 4-5 responses.
Excellent conversation I think I will go to more round tables in the future
Love this session- best part of GDC for me. Super helpful to see how other studios/art leaders approach problems.
I got a lot of information and connections I needed from GDC
Preferred the "free question" style of this roundtable to the theme based discussion of the first roundtable of the week. Though far too much time was spent on the multiple projects in a studio question, this was only even slightly applicable to (at max) half the room and went on for the majority of the time. So feel that could have been moved along quicker to leave room for more questions.
Keith Roundtable is my favorite talk almost every year. I make it a point to come every time. I'd even suggest adding an Art Leadership Mixer for people who come every year as there are a lot of familiar faces, and we don't get to mingle much.

Sunday, April 30, 2017

GDC 2017 - Art Leadership Roundtable - Day 2


For the next day, I really wanted to just open the roundtable up to any attendee who had questions or topics that they were wanted to discuss.  For those who haven't attended my roundtables in the past, you should know that it's pretty much first-come-first serve.

The first person to raise their hand asked the following:  How do you deal with a (senior) developer who openly challenges you?  The following suggestions were offered in response.

  • Speak to your own boss.  As much as this may feel like "telling on someone," the purpose should be for you to develop a strategy to deal with the issue.  If it becomes too problematic, then the issue can be escalated.  However, working through the problem with a colleague may provide different insight.
  • Avoid "stooping to their level" in the sense of disrespect should not be met with equal disrespect.  This will fail to resolve the issue and further compounds the problem by causing you to act in a way which is not professional.
  • Make sure the difference in roles is clear.  Sometimes individuals are insubordinate towards authority, but more often individuals may not be aware of responsibility and accountability.
  • One attendee suggested that you try to "reframe to be the supporter."  This is a delicate strategy to employ.  This doesn't mean that you reinforce their challenges, but rather offer them the chance to express why their idea is better.  In short, you take the stance of finding the best solution rather than whose it comes from.  The onus then follows on the other person to advocate, while you hold the responsibility of final say.  At a minimum, this can help reinforce an atmosphere of collaboration with the larger group.
  • You should also have a direct conversation with this individual in private.  Act quickly and try to get to the core of the matter.  Listen to their story.  As a lead, you do want others to feel like they are part of the process.  However, that doesn't mean that you have to tolerate open hostility.
  • Another attendee recommended the book "Crucial Conversations" as a guide on how to approach critical conversations.
The next topic that was put before the roundtable was centered upon a small studio who is currently growing.  While they have no formal managers, they have identified a need to create a more clear structure.  As such, they were looking for advice on how to manage through this type of change.

  • The first recommendation was to avoid putting people into positions you just need to fill, regardless of skills.  Rather, match the skills of your staff to the positions that are available.  If you have positions with no skill match, then you're better off opening the position to applicants.
  • For the comfort of the team going through this restructure, make sure you clarify what the goals are and what you want out of the structure.  This gives the team to best possibility of supporting the change and potentially helping achieving the goal.
  • Assuming that your new structure also defines new career paths, make it possible for developers to have career paths that don't require becoming a lead or manager.  Develop a career path for craft experts as well.
  • One attendee suggested avoiding the term "manager" all together and instead finding "owners" to avoid frustration during the change.
  • Clarify that you are changing the responsibility structure, not the communication structure.  Developers sometimes confuse organizational hierarchy with how they interact and this should not be the case.
  • Lastly, do not rush the process.  Organizational change is difficult.  Work through the problems and resistance rather than avoiding difficult conversations.
The last major topic of the day focused on crunch.  How do people handle crunch, or any situation where there is simply more work than time.


How do people handle crunch -- situation where there is simply more work than time?
  • In this particular instance, the art director and the lead were taking on the majority of the extra work to "save" everyone else.
    • While this is a common and somewhat noble goal, a number of people voiced concern that this isn't practical or effective in the long run.  The AD/Lead is still required to maintain the vision.  They are expected to guide others, not to take on the responsibility alone.  You have to pull back to the broad view to preserve perspective.
    • Statistically speaking, there is reduced return on investment if the organization relies on the same small group of individuals to handle crunch in every instance.
    • Delegate anyway.  Delegation is how you challenge others and is also how people advance in their own careers.
  • Have you looked at reducing scope?  Although this is an obvious question, it should unveil some potential problems in your development process.  Sometimes scope is crafted with little insight into development bandwidth.  Sometimes there are late additions -- without the requisite subtractions.  It is often too late to resolve crunch late in the process, but crunch should be seen as an indicator of where something went "wrong" (not bad).  Use this as a litmus test for how to improve your planning.  What are the early warning signs?  If those are observed, rescope quickly.
  • Look for alternative solutions to scope cuts.  Are there other resources that haven't been explored/utilized?  For example, are there individuals in other departments who aren't crunching, but who are interested in helping out or learning more in a different area.
  • Although "crunch" has a strong negative connotation, I asked the roundtable (by show of hands) who had experienced positive crunch.  A number of hands went into the air.  Crunch can be a positive experience.  For example, crunching because you're passionate and want to make something even better is good.  Crunching because your pipeline is terrible or because the culture mandates it is bad.  Two key things were proffered as critical to crunch effectiveness and morale.
    • Community building.  It's awful to be forced to crunch because someone else is.  However, it can be culture-affirming if the crunch is felt across the group.  This is most effective when it is managed bottom-up.  "We want to work together to get this done.  This is important to us."  This is decidedly different from top-down managed crunch.  "We need you to come in this weekend.  You aren't working hard enough."
    • Timing is everything.  I'm sure most people have seen a project enthusiasm or tranisition curve.  Most studios have learned to make crunch cyclical rather than persistent.  However, there haven't been many who take into account the excitement curve.  Unsurprisingly, crunch at the bottom of a curve is far more problematic than crunching during ascension, when enthusiasm is growing.

Speaker Evaluation

Once again, here is the unedited report from the day's roundtable

Art Leadership Roundtable: Day 2

Thursday, March 2 at 3:30 pm

Room 120, North Hall
------------------------------
--------------------------------------------------------------------------------------------------
Total Headcount: 81

Roundtable Session Ranking within Visual Arts Track: your session is ranked 3 of 12

Roundtable Session Ranking within GDC 2017: your session ranked 33 of 421

Session Totals (This Session)
Response
Count
Percentage of Responses
Excellent
25
86%
Good
4
14%
Poor
0
0.00%
Terrible
0
0.00%


Comments

It was my first roundtable, and I am definitely going to take part in the one that is planned for tomorrow.
If I could I would take part in art related roundtables three times a day for the whole conference.
Do more art related roundtables.
Loved the roundtable, will definitely come back for all of these!
Lots of valuable information on these talks. Would be great if the elaboration time was cut in half for each question though, because it sometimes felt like we were going in loops and would have been nice to address more issues with the time we had.
Good second session, too bad a question was repeated from day 1 which I didn't find interesting at all.
Focus was less than on day 1, but all in all an ok session.
I was expecting more of an art related discussion in this roundtable, but it was all about management. Thats not to say it was bad, it was a good discussion about managing people and dealing with crunch, etc., but I think this should have been billed as a management and production talk. I know I'm not really the intended audience since Im a programmer, but I had hoped to get some insight into visual development. From my perspective, you could have had a room full of lead engineers and had the exact same discussion.
I couldn't get into the first day because it was full, and I missed the third installment, because I didnt get there by 10, but I really enjoyed this roundtable. I would love to see more of these.
This was one of my favorite things in the conference.
Invaluable! Not only because the moderator was excellent, but also for all the attendees experience which was both enlightening and practical. Again, super great.
Highlight of my GDC for talks. Loved the topics and sharing.
My favourite of the art leadership sessions.
Would have been good to have more time to meet or introduce people. Last year there was a quick introduction for each person in the room.
Keith Roundtable is my favorite talk almost every year. I make it a point to come every time. Id even suggest adding an Art Leadership Mixer for people who come every year as there are a lot of familiar faces, and we dont get to mingle much.

GDC 2017 - Art Leadership Roundtable - Day 1



Once again, I was shocked and humbled to see a room full of people interested in this topic of conversation.  I also think this was the first time the room had reached maximum capacity TEN MINUTES before it was scheduled to start.  Humbly, I think we must be doing good things.  More importantly, I think we're adding value and giving back to the community.

Anyway, I started the first day with an opening topic.  In past years, we've discussed what a successful leader does.  This year, I wanted to discuss traits.  Rather than what the person DOES, but instead who the person IS and how they operate.  What attributes does the leader embody?

The opening brainstorm generated a lot of "familiar topics."
  • Take initiative
  • Maturity
  • Problem Solving / Decisiveness
  • Time and Experience
  • Resourcefulness
  • Effective Critique
  • React to Change
  • Adapt Communication Style
  • Reliability / Consistency
  • Helping Others
  • Passion for Work
  • Inspirational / Motivating
This was a good starting list.  I pointed out that we talk about these topics a lot.  However, when prompted by show of hands, most of the attendees identified that they have (personally or through others) experienced situations where artists did not know how to translate these values into actions.  They simply didn't know how to manifest these attributes.

I then asked the attendees to share a bit more about themselves.  Tell the group about what you did (specifically) that caught the attention of others.  How did you demonstrate leadership
  • One animator shared their experience.  They started by approaching their Lead and asking for additional work.  This person helped to develop a pool of assignments/improvements that would make the game better.  They showed initiative in driving this forward and, through direct action, yielded visible improvements tot he game.
  • One concept artist shared that they worked in an environment with no official Lead.  As a result, they worked directly side-by-side with the Art Director.  This visibility, enabled the concept artist to produce solutions for a variety of different departments.
  • A principal artist said that their career moved forward as they started mentoring others.  Mentorship is likely a whole other topic for a future roundtable.  However, through mentorship, this artist was able to improve not just themselves but the quality of the group.
  • An art director expressed their history of picking up work from a variety of different disciplines as needed.  This experience translated into the ability to create "high level solutions" that resolved problems across multiple departments.
  • A lead artist shared their experience in becoming an active problem solver and assembling diverse groups that could propose alternate solutions.  In addition, they talked about being fun and approachable and trying to embrace the "opposite of the 'no' attitude."  In short, this person worked to build a community to tackle problems as they arose.
In order to have a complete discussion about success, we also needed to talk about failure.  Unfortunately, sometimes people misunderstand how to apply the attributes or commit missteps on their path to leadership.  I asked the attendees to share experiences (no names, please) to help foster greater understanding of what not to do.

  • A lead artist shared a story of another artist within his group.  This individual was frustrated and felt that, given their experience, they were ready to be a lead.  As such, the artist began a campaign to "prove himself" ready for the lead role on a new project.  However, the campaign became more important than the work itself.  As the artist didn't get the lead role (in a time frame that he wanted) he became embittered and performance dropped dramatically.  The lesson here being that the work should always take priority over the campaign (or self-promotion) and that impatience is a clear indicator that someone is not yet ready for a leadership role.
  • Another story that was shared involved the departure of a lead character artist.  The studio had two senior character artists, both of whom wanted to vacant role.  These two artists were close friends, but the pursuit of the open position generated animosity and hostility that was visible to the other members of the team.  Ultimately, the studio didn't select either artist for the lead role and ended up hiring from outside the organization.   The lesson here being that if you're willing to sacrifice relationships for roles, you are also not ready for leadership.  Leadership is relationships.
  • One attendee shared a story about a talented technical artist.  They were being groomed by a manager for the lead role.  As such, this individual was tasked with leading the development of a demo and was scheduled to present the demo for the client.  When the time came to demo, the artist had left on vacation and the demo was incomplete.  The lesson has to be inferred, but it is clear that the artist knew the demo was in a poor state and failed to communicate this reality to superiors or to work to resolve the problems that caused the demo to be incomplete.
  • Lastly, the story was shared of a tech artist who wanted to transition to engineering/programming.  Rather than working directly with his lead or manager, the tech artist approached the CTO directly.  While the company has an "open door" policy, the CTO's first question was if he had spoken to his leads.  The CTO became concerned when they learned the artist hadn't discussed it with others.  While this wasn't a gross failure, the tech artists had "tarnished their brand" by giving the impression of being overly political.
Given all of this, we have a gap.  We have attributes that are valued in leadership.  We have examples of success, but also individuals who don't know how to translate those values into actions.  Finally, we have examples of individuals failing in their attempt to demonstrate these values.

How do organizations "close the gap."
  • Unfortunately, sometimes people can't advance because they are already in a critical role.  In those cases, individuals need to be taught how to train their replacement.  Likewise, the studio should be investing in new talent -- if someone can't advance their career because they are in a critical role, the studio is failing to pass the "bus test."
  • Don't automatically put great artists in lead roles.  Great leads need to be in lead roles.  Poor leads, regardless of art quality, can motivate others to leave.
  • Celebrate individuals progressing their career.  When someone gets promoted, do more than simply give a name and a new title.  Call attention to the specifics in their work so that others can see clear examples of what they did.  This provides two benefits:  a clear example of career progression and a bulwark against peers who may not know what was done and, as such, feel the promotion was unwarranted.
  • If this fits your organizational culture, develop a lead interview process.  Make it clear to the organization how the lead is going to be selected and solicit "applicants."  Though this will likely require careful messaging, the transparency may prove beneficial once a lead is selected.
  • The idea of a flat structure was also expressed -- removing the need to pursue official leadership roles.  This was a contested topic, but a few key ideas were expressed:
    • Even in a flat structure, someone still needs to own the product vision.
    • Flattening the structure should be about flattening the communication layers.  Invest in tools (or processes) which encourage transparency and facilitate regular communication.
    • Organizations need a lot of time to acclimate to new structures - so be careful implementing this in a pre-existing culture.
    • Responsibility and accountability are still required.
In the final moment, I asked attending leaders to share with us their strategies for balancing the leadership role:
  • Understand your strengths.  Focus on the things at which you excel.  Mentor and delegate the rest.
  • Schedule "lead time."  Hold office hours or dedicate specific time on your calendar for yourself alone.
  • Preserve elements of complexity or which will require frequent iteration.
  • Find time to create content.  Don't lose perspective.  Understand the pipeline and the limitations.
  • Hire for your needs and hire those who can follow the type of direction you provide (strict or flexible)
  • Provide clarity of the role (too frequently leads can get sucked into doing work outside of their areas of direct responsibility) and evolve the role over time if necessary.
  • Invest in an outsourcing manager.  Don't insist the lead focus on both the internal team and external developers.

Speaker Evaluation

Once again, here is the unedited report from the day's roundtable

Art Leadership Roundtable: Day 1

Wednesday, March 1 at 3:30 pm

Room 120, North Hall
------------------------------
--------------------------------------------------------------------------------------------------
Total Headcount: 81

Roundtable Session Ranking within Visual Arts Track: your session is ranked 9 of 12

Roundtable Session Ranking within GDC 2017: your session ranked 123 of 421

Session Totals (This Session)
Response
Count
Percentage of Responses
Excellent
17
68%
Good
8
32%
Poor
0
0.00%
Terrible
0
0.00%


Comments
(Note: The following comments are aggregated across all 3 days)
'This was the second year in a row that I have attended his art lead discussions and I am always blown away with all of the useful information. It has significantly helped me and how I do my leadership role.
The moderator is a very talented communicator who was able to distill information and keep the talk on track.
Awesome stuff.
It was great to hear others' POV's and experiences. Lots of valuable information on these talks.
Would be great if the elaboration time was cut in half for each question though, because it sometimes felt like we were going in loops and would have been nice to address more issues with the time we had.
I was really motivated to get to this talk, and got a lot of useful feedback out of it. However, I noticed a shift in focus for the attendees: experience or personal background divided a focus, which meant a lot of discussing our focus as a group in itself.
I loved hearing about the different experiences (nosedives) or how others organize their teams or work.
Personal stories/time work better than problemsolving for me!
Great experience to speak with so many other varied artists from studios all over the world.
This session could have been longer and we all would still be talking.
One of the most useful talks I attended.
The round table discussions are proving to be one of my favorite, most useful and most importantly of all, applicable of all GDC sessions. Thank you!
Loved this roundtable, a lot of topics where brought up and I realized I was having these feelings before without realizing it. Thank you!
It was ok but there weren't as many solutions offered to some of the issues the leads were facing.
Always good fun! Thank you Keith!




Sunday, February 26, 2017

Game Developers Conference 2017



It all begins tomorrow.

Monday we have the Art Direction Bootcamp, co-organized by me and Andrew Maximov.
We have an incredibly group of talents joining us on the stage again this year, including Denis Rogic, Greg Foertsch, Liam Wong, Jason Connell, Brian Horton, Eric Pangilinan and Christian Nakata.

What a group, huh?

Also, you can find me Wednesday, Thursday, and Friday moderating the Art Leadership Roundtable.

See you there.