GDC-3.1-Kanban - How to Make Your Production Scream

by Clinton Keith ( President of CK Consulting )

Kanban is :

• a tool for managing the flow of assets or information in a process.
• translated from Scrum, where it is visualizing the flow of valuable stuff being built.
• something that does not rely on a plan of 'should', since Reality > Planning.

3 Basic Rules

• Visualize workflow so there are clear goals for delivery.
• Limit work-in-progress to reduce multitasking/batching assets.
• Measure and improve flow, seeking ways to introduce measurable improvements.

Scrum vs. Kanban

• Scrum is based on a To Do-Ongoing-Done approach, making it better for slower or more definitive projects.
• Kanban works with a To Do-Ongoing1(#)-Buffer(#)-Ongoing2(#)-Done setup, where # is how many tasks are in that process at max capacity, making it better for faster hand-off projects.

Setting-Up Kanban

• Visualize the process, map it, add metrics, inspect it, and adapt everything for your situation.
• The process is the 'as-is' of the situation.
• Consider sticking a Process column to a Kanban board.
• There is only 1 asset from beginning to end.
• Set WIP limits as identified by numbers (#) next to each column title.
• Pay attention to limits - if they are overshot or never reached, they may show that there needs to be more/less of something.
• Create an Emergency Lane, where if any item is in the lane and in a specific column, all other things in the column are put on hold until the item leaves.

Edits/Mods to Kanban

• Split columns and assets between specialized teams.
• Add a marking for assets requiring special attention.

Sprint Planning from Scrum is not a feature of Kanban; it is replaced by 2 week cadences that reflect on what has happened when they end.  Planning lies mainly in Burn-Down charts, where the y-axis is how many tasks are being taken-on during a cadence, the x-axis being major time landmarks during the 2 weeks.

Summary:

• Kanban can be started quickly, it improves workflow with its features, and it regularly leads to a +50% improvement.

// Up next from GDC: The Walking Dead: Crafting a Stylized World for the Mature Franchise.

GDC-3.0-The Design of New Enemies for Halo 4

by Scott Warner ( Lead Game Designer - 343 Industries )

There was a 'fatigue' from fighting the Covenant in every Halo game so far.
Where was the '30 seconds of Fun'?

• Find this from everything from Halo to Halo Reach.
• Apply it to the Prometheans in Halo 4!

Foundations:

• Every alien race is unique in Gameplay, Visual Silhouette, and Personality.
• Enemies support experimentation (Halo is an open sandbox game).
• There's a Leader/Minion hierarchy.
• Some are easy to dispatch, others are real conflicts when encountered.
• There are 'minibosses' with a squad/squad leader.
• The Halo Encounter must exist for Prometheans.
• Classically, it's Plan, Take by Surprise, Beat Back, Break Them, Mop-up, and Repeat.
• Enemy characters (Leaders) raise or lower Minion moral and action.
• The Halo approach to difficulty is Smarter-Faster-Stronger.
• Higher difficulty only speeds-up gameplay; it doesn't change it fundamentally.
• PROMETHEANS MUST USE EXISTING CONTENT CREATION TOOLS!

The Gold Triangle: Shoot - Hit - Throw

Promethean Inspirations:

• Voltron
• Forerunners are involved beyond anything else in the galaxy.
• Adaption of form to fit events and environments.
• T1000
• They can take any form, but stay relatively humanoid.
• The Thing
• Cranium Rats
• Units are smarter the more there are.

Early Enemy Goals:

• There should be a combination of 'familiar' and 'alien'.
• They cooperate with each other to defeat threats.
• They adapt their form and the environment.

The Knight is something easily understood and relate-able.

The Crawler (Pawn) started as a biped.

The Watcher (Bishop) was wanted to be something like a Necromancer from Diablo; a game changer.

Other enemies were Rooks, Queens, and Kings, all of which are likely to be included in future releases.

Watcher Prototyping:

• Had the abilities of a shield, resurrection, deflection, spawning, focus fire, physics impulse, a 'junk attack', etc.
• From focus fire on, players just didn't get what the abilities did.

Knight Prototyping:

• They used squad tactics in a combination of alien and human strategy.
• Use cover like a human, but use a bound move like a Metroid ball.
• Enter and exit battle through the environment.
• Dismemberment is shown when damaged, answering 'am I damaging this thing or not?'
• Have a deadly melee attack!

Problem: Pipeline.

• Pipeling rewriting for Maya was a pain.

Solutions:

• Bring in more engineering staff.
• Reduce the scope, cutting the Rook, Queen, and King enemy types.

Problem: Who are they, the Prometheans?

• They were not fleshed-out with gameplay features seeming to be random.

Solution:

• Center everything around the Knight.
• Finally say 'this is it' and drop some features.

Problem: Pawn

• The team didn't like the biped design because it felt like the Covenant.

Solution:

• Walk on walls as a quadruped!

Post Mortem Wrongs:

• Early team communication
• Long absence of high-level vision
• Watcher encourages one-dimensional solutions
• Prometheans are devoid of emotion
• Forerunners are like bullet-sponges

Post Mortem Rights:

• Early gameplay contraints and a foundation
• Enemies players enjoy
• Creating different enemies from either the Covenant or Flood
• Team orientation

// Next: Kanban-How To Make Your Production Scream.

GDC-2.1-The Future of Storytelling: How Medium Shapes Story

 by Jesse Schell ( CEO - Schell Games )

Weaknesses of current games:

                Verbs

                                Right-now verbs of running, shooting, yelling, etc.

                                ‘Below the neck’

                No use of tragedy

                There is a lot of lost unification

                                All great stories tell the ending by foretelling it at the beginning.

 

Elemental Tetrad of aesthetic, mechanics, story, and technology.

                Either mechanics or technology changes will likely lead to better stories.

Prophecy of Chris Swain:

                Film was taken seriously when film learned to talk.

                Games are waiting to learn to ‘listen’.

                                This will allow for a conversation between the game and the player.

Games must: sense emotion, understand spoken word, and remember what a specific player has done.

Who will we talk to?  Our avatars (virtual companion!) is the answer.

Don’t think ‘platform’, think ‘venue’.

                The Hearth – OK for story games.

                Anywhere – Great for mobile games, but terrible for story games.

                Reading Nook – A good venue for story games played on things such as tablets.

                Workbench – The home of PC gaming and a good home for story games (future tech would include the Oculus).

 

We need: Above the neck verbs, irreversible tragedy, and everything coming together in connection and reference.

What happens to avatars when the player dies?

                Inheritance: Virtual companions can be sought to teach about the player that was alive.

// Visit jesseschell.com to find the slides!

// Next: The Design of New Enemies for Halo 4.