Blog Archive

Saturday, October 29, 2005

Live blogging fr AGC: Fun Meters, Lazzaro and Mellon

28 October 2005, Texas, Austin, Austin Game Conference

Mellon & Lazzaro

"Friday, 4:30pm - 5:30pm
Fun Meters for Games: The Ins and Outs of Measuring the Player Experience
As development costs and risks go up, innovation in game design falls. Yet because the power of Next-Gen hardware creates possibilities for new forms of game play, we need innovation more than ever. How can we increase innovation in the design process without risking millions on a potential flop? A proven technique is to apply iterative innovate, play test, analyze, and design cycles early in development. If cheap, accurate measures of fun were made easily available to designers, what impact might that have on their ability to adjust the challenge ramp, change tuning parameters, develop more popular features, increase emotions, or experiment with a new game mechanic? This talk provides an overview of known techniques for measuring fun, such as direct observation of player actions and measurement of their emotive and verbal responses. We also discuss what types of questions such measures answer, their strengths and weaknesses, and how automation increases the use of metrics early in the design cycle. Come participate in this discussion of what kinds of fun can be measured, and what might be done with such data early in the design cycle when changes are easier to make.
- Nicole Lazzaro, President, XEODesign, Inc.
- Larry Mellon, Chief Architect, VP (Engineering), Emergent Game Technologies"

Nicole Lazzaro

How make a fun meeter.
Need to pull in data from various sources.
Need to be repeatable.
What is fun differs between different people and different games

Previous study: "why we play games"

Where is the innovative game play?

How will you get is funded.

How would a fun meeter for pac-man be?
Next generation: packing in more features
Pac spore

Mobile off-Computer Play

Next gen gambling...
Blocking issue: big games hard to change at last minute.
Build incrementally.

Fun meeters today.
Sims online. Checking what behaviors were used the most.
Geographically: hot spots and dead spots?

Only way a funmeeter is going to work is if it is not too late, not too close to shipping.

Nicole: looking at facial expressons.

4 flavors of fun; from the white paper.
Hard fun - frustreation and fiero
Easy fun - surprise and curiosity
Altered states: change internal sensations (like therapy) excite and relax
The people factor: social mechanisms, shadenfreude + naches

Slide from  Fun Meters session at ACG 2005 by Lazzaro and Mellon

Sucessful games usually have 3 of these factors.

A nice variety of the flow diagram.

Testing leapfrogs fly pen top.
Created a mockup of it.
Sat down with useres. Roleplayed with users having a script.
Didnt wait for implementations, could do it early with the paper mockup.

Now a kind of game thing,

Writing things to measure for fun on big pieces of paper.

how measure fun?

Noice in the signal when testing.
Especially when players are interpeting the experience.
Professional testers: not the target group.

Risk of designing for a vocal minority.

Facial expression: one can see a player looking digusted, but he might not remember it afterwards.

Can also listen to reviewrs. Are they repersentative for your players? Also its way too late.

Online serveys. Look great in a speradsheet, but players are self-selected.

Slide from Fun Meters session at ACG 2005 by Lazzaro and Mellon

So what to do then.

Observe player experience.
Takes time to analyse the data.
Context. Imortant to look at those data too, like realtions and actions of others.

A real smile has wrinkles by the eye.

Paper fun meeters.

Tokens, yellow stickers, use simple methods to get to not have to analyse so much video.

Sims online, seeing that ppl spent in-game money on windows. No idea why, but next expansion will have more windows.

Best results when using several meeters in the same testsession.

Measuring factors in the game that has an impact.

Need to take advantage of that we can do.

Knowing when:
Player testing and reviews
Data probes
Observarion and prototying play.
= early!

We need more than better visuals in next-gen games.
Visual impact eventually tapers off

Breakout games; establishing new genres.
Pac man
The sims

Measured innovation, early in the design.
Fun meeters of the future: unite.

Metrics addictive to designers.

all photos from this session: