Black Rock retrospective

Black Rock retrospective

By Develop

January 27th 2009 at 12:01AM

In this frank retrospective on the production of the studioâ??s first new IP for Disney, Disney Black Rock production director Alice Guy examines the teamâ??s approach to embracing â??focused developmentâ?? processesâ?¦

As Black Rock Studio celebrates its second birthday under Disney, it’s a great opportunity to reflect back on some of the changes.

Our culture’s remained as strong as ever, but we’ve had the opportunity to radically rethink our approach to production. And looking at our first Disney-funded title, Pure, we’re on the right track. In our previous life as Climax Racing we produced twenty games for a number of publishers. Our two standout franchises were MotoGP (THQ) and the Fury franchise (SCEA). We always had a focus on racing games – which served us well. But we knew there were areas where we could take it further. We yearned to focus on quality above all else, have more creative control; and a better work-life balance for our staff.

As an independent we were all things to all publishers. We were working on any formats – from console to handheld to PC. We had individual teams working on an individual platform which was a nightmare for technology sharing and game design. Now we not only focus on just two platforms (PS3 and 360), outsourcing other SKUs; we jointly develop PS3 and 360 on one team which has brought great rewards.

We were tied to immovable delivery schedules and timelines. On top of this we weren’t able to say no to a publisher’s requests, and too often they used the number of bullet points on the back of a box as the measurement of product quality.

CRITICAL SITUATION
One of our last games under Climax Racing was ATV Offroad Fury 4. GameSpy wrote this in a review: “the developers not only threw in the kitchen sink, but they went to the neighbour’s house and threw in his kitchen sink as well”. There were layers of complexity and the team was spread precariously thin. Few areas got the focus and attention they really deserved. Those that did really stood out and got great feedback.

Fury 4’s submission coincided with the acquisition and it influenced a lot of the changes we would make to how we approached production. Although many in the studio were nervous about Disney’s arrival, the acquisition was a huge momentum for change. It provided a clean slate, a chance to reshape the studio. Crucially we’ve been entrusted with creative control, and we’re able to have open and honest communication with our publisher.  Disney also buys into having focused products which is great.

Indeed, Gordon Ramsey is our new patron saint. Take Kitchen Nightmares – restaurants are seen drowning with lengthy menus and chefs struggling to get dishes out on time. Overly complex recipes make a loss and the ingredients jar and undermine each other. So what’s Gordon solution? Simple, focused menus; dishes that suit the restaurant’s theme; and fewer ingredients in each dish so the chef’s skills can shine through. Pure has been our first product where the team – led by ‘Head Chef’, game director Jason Avent – has had the focus and freedom to truly achieve a clear vision. The key drivers have been a more focused design; a more flexible and iterative production process; and a broader management structure on the team.

THE BIG IDEA
On the design front the ‘Big Ideas Sheet’ has been a massive win. The bullseye design lists out four to five key ideas – the premise being that there’s a ‘bullseye’ for each title with a limited number of supporting features. Everything in the game feeds into these. If a feature doesn’t then it’s not worth doing.

The game directors use the Big Ideas Sheet to direct and focus the team by constantly reminding everyone that “If it's not on the sheet, it's not in the game”. The publisher also buys into these objectives, and marketing knows how to pitch the title without getting tied down in specifics which may change.

It’s important to note that these Big Ideas aren’t set in stone right from the beginning – they can change as the game’s core concepts are developed. For Pure the Big Ideas Sheet was iterated several times. Its top ranking feature changed two months or so into mass production. ‘Vertigo Rush Moments’ – the gut reaction experienced when the crest you thought was a bump in the road turns out to be a cliff face - became the emotional response the team focused on. It feeds into the jump mechanics, the trick cameras, the vertigo rush effects and is absolutely key to why Pure is fun. It’s proven to be a simple tool to maintain everyone’s focus and ensure we’re not making the game unnecessarily complex.

BEING AGILE
There have been other much deeper changes though, such as our move towards Agile Development.

As anyone who’s worked in games knows, it’s hugely challenging trying to schedule game development. Things change on a daily basis and schedules can look very different at the end of the week than they did at the start. We needed a production methodology in place that would allow us to be flexible; allow us to change direction – and crucially minimise the impact on the team. MotoGP 06 had adopted ‘agile’ behaviours without really knowing anything about Agile but without total control it couldn’t be as focused as it needed to be. The acquisition provided the impetus and security needed to really address our production methodologies and answer some of the issues we’d been facing.

Agile is fast becoming the standard for the games industry but it’s actually been around in the manufacturing business for decades – with Lean production (its forefather) originating in Tokyo in the 1960s. Agile, as I’m sure many Develop readers will know, focuses on people and partnerships over processes and tools; working software over reams of documentation; listening to the customer; and responding to change over doggedly following plans. For a developer, and the production department in particular, it’s important to react quickly and easily to changes. Agile provides tools and mechanisms to manage complexity and unpredictability.

We’ve adopted many Scrum methodologies. Planning is lightweight and user-friendly leads aren’t spending hours on it, and teams can largely be left alone to focus on the tasks

in hand.  Teams work on a more short-term basis – three-monthly milestones are split into short ‘Sprints’ – three weeks long. For the sprints it’s a good balance of accomplishment versus planning. For the milestones you get major visible progress.

TEAM WORK
The team structure’s completely changed. We’d played with ‘strike team’ concepts in the past but this was the first time we completely split out the team into feature areas (known as ‘sub teams’) and mixed up programmers, designers and artists. Sub teams manage their own workload. Membership can change from sprint to sprint/milestone to milestone depending on priorities and capacity levels.

Teams communicate more effectively and focus on joint goals. They’re accountable to each other, and have an understanding of common issues. Also, features aren’t worked in isolation by one person alone – so if someone leaves or is off sick we’ve not lost that knowledge.

There’s also a much broader management structure. The group of project directors focus on setting the vision, goals and direction. Then each sub team has a dedicated producer, and Leads for each discipline. Producers act as the ‘Scrum Masters’ (we’ve not adopted the term itself internally, but have replicated the role they play) and help teams plan and progress the priorities for that sprint and track dependencies.

The Leads provide clear points of contact and drivers for the different areas. The number of reports per Lead is kept to a minimum, and more producers means their time is freed up to focus on mentoring and setting the quality levels for their teams.

[img:463]

PRODUCERS ACTUALLY MATTER
From a production point of view the investment in a production department – before we simply had one producer per project compared to a team that ramps up to five per project now – has been fantastic. Producers no longer maintain detailed schedules with granular tasks assigned to individuals. Instead we have a ‘Product Backlog’ which outlines high level features, and the sub teams that will need to work on them. Initial estimates are assigned to teams (not individuals) so you can review capacity and move people accordingly but don’t get bogged down in the detail.

Before every sprint the project’s senior management reprioritises the backlog so the highest priority features are at the top, ready for the team to remove and assign for that sprint. Crucially teams work on features until they’re complete – aside from the end phase fine-tuning to avoid having a game full of 70 per cent implemented features developed in isolation.

Features are fully designed, pre-produced and estimated in detail as they’re worked on with an appropriate ‘development partner’. You don’t hang your hat on estimates set at the start of a project; or rely on design schemes from the start of development.

We are now developing this approach further by ensuring features that span several sub teams are planned as a whole, rather than (as we did previously) splitting out the backlog and prioritisation sessions into sub teams. This will aid dependency-tracking and ensure the whole feature is testable faster.

This approach, along with a small QA team supporting development throughout production, means the game is always kept working, enabling continuous testing and review. User tests can take place sooner; areas for improvement can be spotted quicker, then developed further or, in some instances started from scratch.

ITERATE, ITERATE, ITERATE
Iteration is something we’ve embraced, and we really challenge teams to look critically at the game and improve and iterate wherever necessary. It’s not easy for people to throw work away but it’s important to understand this isn’t wasteful – it’s the only way to develop a triple-A title. Yes, it makes production and scheduling harder, and sometimes the scope has to be narrowed in order to make the numbers work, but that’s a compromise we’re willing to make. What’s the point of getting that extra feature in if other areas are below par?

The key is working code; a flexible team and publisher; and taking time out to change and fix as you go – not hope you’ll have time to sort it at the end. For us, Agile Development equals iterative development which means you’re continuously striving to stay ahead of the curve. Pure certainly benefited from this approach. It’s fascinating to look back at how the game progressed, particularly areas such as the HUD, track design, and environments. Supporting this approach is the studio’s investment in ‘previs’ (previsualisation) and rapid prototyping. You make changes when they’re cheap – i.e before feature implementation. It might be videos, real world models, Flash, right through to trying things as cheaply as possible in software and the game.

PURE THINKING
All of this has helped define what being part of the team here is all about. It’s about being passionate, innovative, flexible and creative; and a self-driven team player. It’s about being willing to throw work away; being open to critique and peer review; and looking for ways to rapidly prototype work.

In return we offer staff a better work/life balance. We have the control now to better protect our staff from the pressures of crunch, and we focus on working smart. It’s not perfect yet, there’s always areas for improvement, but it’s leaps and bounds beyond where we were as an independent.

In addition, empowerment and engagement is a huge area of focus for the management team and we’re anxious not to play lip service to it. We want to empower the team to help direct and shape both our products and processes – only then do you get the best ideas, participation, motivation and quality.  Everyone knows about the production ‘triangle’ of time, cost and quality. You could say there’s a dimension missing though – productivity – and we’ve found that when individuals and teams are empowered, self-managing and motivated the productivity improves in a very measurable way.

To conclude, it’s been amazing to see the changes over the past two years. Sure there have been challenges, and there’ll be more to come, but the journey’s been hugely rewarding so far. We’re on the way to reaching our goal – to be the best racing studio in the world.

www.blackrockstudio.com

Alice Guy is production director of Disney’s Brighton, UK-based Black Rock Studio, responsible for managing and leading the strategic development of the production, QA and audio departments and spearheading production process improvements across the studio