Previous Entry Share Next Entry
GDC 2010: Creating Successful Social Games: Understanding Player Behavior
Bar Harbor
Creating Successful Social Games: Understanding Player Behavior
Speaker/s: Mark Skaggs (Zynga)
Day / Time / Location: Thursday 10:30-11:30 Room 134, North Hall
Track / Format: Production / Lecture
Description: You can build the fastest growing social game the world has ever seen, but if you don't know how to analyze what your players are doing each day, your chances for real success are doomed before you start. We'll cover the key player metrics you'll want to start tracking the day you release your game as well as which become important over time. Why reinvent the wheel? If you're just getting started making social games or have been doing it for a while, save yourself time and effort on your way to success by attending this session.

Metric Mindset
test, review, modify, rinse repeat

Some things to measure:
How many people get through your install process?
How many players last through the first five minutes of the game?
How many people are playing today?
How many people tell their friends about this game?
How often do they come back to continue playing?
When do they stop playing?
How much money do they spend in the game?
Do they enjoy playing

Real data in real time is much better than old-fashioned guessing
Measure what you want to know about
If something is important, measure and track it

Identify when/where you re losing people; try and plug holes

Tutorial change comparison anecdote
Slashed tutorial by 40% without much care: player retention to end of tutorial increased +25%, albeit worse player understanding of game.
Refined tutorial significantly, significant chunk of work: Improved player experience -- but only +0.5% retention increase.

Metric = "let's test it" "What do numbers say?"
What color should the text be? anecdote
Text for cross-game ad line at top of screen
"Red on white is clearly correct"
"Let's test some alternatives"
Numbers clearly show Red is *lowest* scoring!
Pinky-purple is much harder to read -- yet gets about three times as many clicks!

Measuring gets better results than guessing
also removes ego issues, can get you past entrenched preconceived notions
[Alexx observation: CoDWaW Zombie mode has one map per DLC, despite popularity. Did Treyarch ignore numbers on who was playing what? I don't have the numbers myself, so I'm just guessing here...]

Hard to define 'fun'; easy to observe behavior
If player repeats behavior, it's fun
[Alexx observation: This is a great short-term measurement, but what about long-term consequences? I have had experiences where I keep clicking compulsively for an hour, but at the end of it I feel manipulated and wasteful, and quit the game forever. Can you measure that sort of longer-term effect? Harder, but should be possible.]

graph player actions
find patterns
identify which patterns match 'fun', and which match 'boring'
create opportunities for 'fun' behavior patterns to arise

  • 1
I have had experiences where I keep clicking compulsively for an hour, but at the end of it I feel manipulated and wasteful, and quit the game forever

That might be a success story for some of the games Zynga produces. Especially if you paid first. "Well, that game was $1, and it was fun for an hour or two. I wonder what else is in the App Store..."

  • 1

Log in

No account? Create an account