Title: CRITICAL STAGE ANALYSIS
1CRITICAL STAGE ANALYSIS
- Integrating CSA into the game development process
2What is Critical Stage Analysis
- Simple mid-project feedback
- Designed to get feedback from all team members
- Requires team to take responsibility for cited
problems ASAP - Intended to improve game quality and reduce
crunch time by fixing problems early
3The Origin of the Process
- Wolfgang Hamann of Radical Entertainment
- Article published in the July 17 2003 issue of
Game Developers Magazine (available at
gamasutra.com) - Tutorial session at 2003 GDC (s are archived
at the GDC website)
4Wolfgangs Theory
- We plan for our game production process to be
very linear but it never really is. - Were unprepared for the loops were thrown for
and we have no tools to help us see whats coming
in advance. - CSA adds a feedback loop that allows us to see
our true production path more clearly and be
better prepared.
5Who am I to talk to you about Critical Stage
Analysis
- Im just a producer using the process
- Amaze Entertainment small Austin studio
- Handheld projects on Nintendo DS
- Licensed properties 3rd party developer
- Small teams about 15 people or less
- About a year of development time (or less!)
6CSA A couple of caveats
- Hamann thinks postmortems should be scrapped I
disagree. - The literature presents this as a very simple
process and while the doing a survey is simple
executing solutions for problems identified
mid-project is not.
7Surveying the team
- The big three questions with answers ranked in
order of importance - Name five things that went well during the stage
- Name five things that went wrong during the stage
- Name five things that could be improved for the
future. - The rating process is considered critical.
8Analyze data and deal with problems
- Data should be gathered and distributed to all
team leads. - Team leads should meet to analyze the data
determine solutions assign ownership. - Data and solutions should be presented to the
entire team for discussion. - Everything should happen within one week of the
milestone or stage that has just past.
9The Ideal Timeline
- Survey made available no more than three days
after a milestone. - Team members will spend about 10-15 minutes
answering survey. - Information should be compiled and distributed to
leads no more than two days after it is gathered
and should be discussed immediately after
receipt. - Discuss the results in the next team meeting
within the week.
10Tips and tricks
- Give team members multiple ways to fill out
surveys. - In team meetings bring up remaining issues from
the last CSA first. - Lead with the positives.
- Always deal openly and honestly about all issues
dont hide things from the team. - Unsolvable problems should be broached.
- Unedited CSAs should remain accessible.
- Follow up is essential.
11Measures of success
- Fewer issues are raised repeatedly
- Fewer things ranked with high importance over the
course of the project
12What I wanted out of CSA
- Feedback before the game was over
- Feedback from the quietest team members
- Feedback on both big and little issues
- A new communication tool
- Morale measurement
- A chance for team members to bestow accolades
upon one another.
13Things I changed
- I added a couple more questions essentially to
help get a measure of morale - How satisified were you with the look and feel of
this milestone and why - How satisfied were you with the deliverable we
sent overall and why
14Things I changed
- Everybody was required to fill out the survey in
the same format. - Surveymonkey.com
- Zoomerang.com
- I didnt follow the suggested timeline.
15My CSA Timeline
- Monday after milestone send out CSA survey to
team. Deadline to fill out noon on Thursday. - Wednesday afternoon send out reminder to do
survey. - Thursday afternoon producer compiles data and
sends to all leads and directors. - Friday morning meeting with producer leads and
directors to determine what kind of solutions can
be offered for all problems identified. - Friday afternoon producer creates a nifty CSA
presentation - The following weeks team meeting team
participates in a CSA presentation
16Things Ive learned
- The CSA process has been a great thing most of
the time. - People did indeed laud their coworkers.
- We did indeed identify problems we were able to
solve earlier in the process. - I started getting a whole new level of feedback
from team members about the project both macro
and micro issues. - Some project teams have become more emotionally
invested in the process than others. - Its very hard to keep things anonymous.
17Things Ive learned
- The producer has a huge influence on gets
discussed in the CSA meeting. - Its difficult to tell what the rankings mean.
- The two questions about what went wrong and what
could be improved basically always ended up being
10 things that went wrong. - People who dont communicate well in meetings
dont always communicate well on the surveys
either. - Its hard to give people unlimited ways to
participate. - It takes a lot of time.
18Things Ive learned
- People may wait until a CSA survey to present a
problem and thats a problem. - CSA results really give you a leg up when youre
getting ready for a postmortem! - Opposing viewpoints come up all the time thus
you cant make everybody happy. - Some team members may be triggered in unexpected
ways by the process.
19But the most important thing is
- Doing a bunch of surveys doesnt do a thing
unless you as a team commit to fixing the issues
that are raised by them. - Team members need to feel that team leads and
directors have got their back and are willing to
make changes to the production schedule game
design and perhaps even the staffing and budget
plan if an important problem sited needs to be
solved.
20Future Plans
- Start an online repository of CSAs to be used
company-wide. - Ask even fewer questions.
- Ask contract vendors to participate.
- Get the client involved.
21Thats it!
- Questions Comments
- Do you have a CSA-type process youve used How
does it work for you