Previous Entry Share Next Entry
GDC 2010: Motivating Casual Players: Non-Traditional Character Progression and Player Retention
Bar Harbor
alexx_kay
Motivating Casual Players: Non-Traditional Character Progression and Player Retention
Speaker/s: Laralyn McWilliams (Sony Online Entertainment)
Day / Time / Location: Saturday 10:30-11:30 Room 133, North Hall
Track / Format: Business and Management / Lecture
Description: RPGs and MMOs rely on leveling up, stat increases, and item unlocking for the carrots that retain players. Multiplayer shooters and fighting games rely on competition and a skill ramp that encourages players to fight for positioning on ladders and leader boards. Traditional single-player games rely on new areas, new abilities, and storylines to encourage continued play. Yet all of these things are forms of progression.
This presentation presents examples of alternative player progression, using games like Free Realms, Pogo and Halo for development examples. It compares and contrasts progression at both ends of the spectrum to develop a methodology that steps you through the usual and some unusual choices in progression. It also takes a critical look at how casual games, social games and large brands are innovating in progression.

Alexx: Lots of great info on how casual players differ from 'core' gamers.

Free Realms
At launch, had following character progression:
XP, levels, stats, most XP from quests, loot rolls
EPIC FAIL

Traditional progression didn't work; nor did character progression
Made significant changes over 10 months
huge improvement in retention and other factors

measurement again -- their initial intuitions turned out to be false; they tried new stuff, measured results, improved the game tons.

casual players defined:
do lots of other stuff, not just games
play in small chunks of time
haven't played lots of different things
not familiar with basic (to 'core' gamers) notions like XP.
poor at controls
never used WASD
may not have used mouse and keyboard simultaneously before
stop playing at first barrier, don't come back

Looking at data, clear and obvious separation between core and casual players
This difference completely swamps any differences in age and gender!
Casuals are a big enough percentage that Sony is concentrating on them

Progression != XP
reward all play styles fairly
respect time, effort, money equally

Players want Rewards for winning
They also want at least a minor reward for LOSING!
Reminding them about recent wins when they log on is powerful

Progression is revenue
Gives player a goal to make them want to keep playing
spend (money or time) to continue growth cycle

measure growth and change
Give big rewards at regular thresholds
Ensure a clear connection between behavior and reward
Operant conditioning

small % of players got higher than level 10 in any job
smaller % higher than level 10 in multiple jobs
even smaller % spending 'stars' earned by leveling up
made progress clearer, smoother, ditched 'star' mechanic

Game as released: 75% XP from quests, 25% from minigames
measured small % of players actually doing quests, many more playing minigames
Swapped XP balance to 75% minigames, 25% quests

Problem: Casual players want to start playing within 30 seconds, will quit if it takes too long
Introduced Game Guide mechanic: lets players join a minigame quickly, without having to navigate 3D world

"don't make me do math" turns out to not go far enough
casual players didn't understand stats even w/o math
removed most stats
combat items now have 1 stat, and 2 abilities

Casual players don't look stuff up on web sites; they don't know what loot drop possibilities are
Even if a boss drops the most awesome weapon evar, and drops it 100% guaranteed -- most players won't know he exists!
Minigame (or boss fight) start window now shows all possible rewards
When you win, you see all the possible 'drops' in a big wheel, that you spin to see what you get.
"This was a huge win for us and everyone should copy it!" (Alexx: I agree)


Social progression
guilds
level up relationships
FB friends

They held an in-game concert; many people attended
Hours after concert ended -- people were still hanging out!
Looked at data, surprising result:
people were spending more time socializing than on ALL minigames COMBINED

They created party zones, private party mechanics ("rent a room for a party")
Public parties get listed in game guide, very popular

"I'll spend money to show off"
Original item creation prioritized Weapons and useful tools, then silly costumes, with toys and tricks as afterthought
This turned out to be exactly backwards; Casual players value toys highly, then silly costumes, with weapons and such being mostly valued according to their aesthetic properties.

Now they make things like banana suits, or a boombox that makes people dance -- and then turns into a rickroll!

Casual players have no clue about formal 'group' mechanics and NEVER use them.
Casual player assumptions:
friends list = group
proximity = group
changed mechanics to deemphasize groups, support lobbies
Casual players don't understand lobbies *either*
Then tried silent matchmaking -- still not very used
Currently working on mechanics to automatically group and ungroup players when appropriate

Gaia Online is a good social success story to look at
social mechanics as character progression; XP for joining guilds
gold rewards for making forum posts
YoVille

Player progression
trophies/achievements
high scores
Feature to make a TCG card of your character, based on actual level
people choose options that make cards more individual, NOT more powerful
Player progression is even stickier than character progression!
understood by people who don't game -- grandad can congratulate you for success on FB game w/o needing to understand it
for casual players, player progression IS character progression
People make avatars with the same gender, skin, hair, even name

Player progression success stories to look at:
Waypoint
tying multiple games together
Pogo avatar customizer
Big Sea (big fish?)
everyone in the room gets a reward when ANYone progresses -- hugely powerful social effects!

Progression is a 3-step process
1. need
2. interaction
3. reward
Needs
rewards only work when they fill needs
Core wants skill, information
Casual want coins, social status, individuality
Both want relationships, ownership, entertainment, sense of progress
Interactions
No market segmentation of interactions; core and casual play the same minigames
Rewards
Core: leaderboards, knowledge, level up, useful item
Casual: money, profile popularity (just looking at page), appearance, socialization
Both: Friendship, fun

Classic 3-step Progression:
need Wealth - kill npc - get money

Use new combinations to think outside the box, create new gameplay:
need relationship - trade with npc - appearance
need info - kill NPC - get popularity (need to find mobile boss, first ten people to kill today get name on poster in center of town)
need skill - explore - appearance

casual players
don't always equip new items, even if reminded
don't value stats or levels
compare dress, but not levels
don't want to grind
like appearance
like status *when reflected in appearance*
like uniqueness
like frequent smaller rewards
only understand large changes in effectiveness - don't even TELL them about 1% stuff, they will only feel cheated

Q&A
casual players just don't like the same sort of game mechanics as core players -- training won't help, they just don't like it
No expectation that these players will ever 'graduate' to core MMOs
Casual has no patience for long tutorial -- must be playing within 30 seconds
If they are having fun, they will puzzle out advanced features later on

Players will play short periods just for fun or random rewards -- known rewards drive long term play -- need both

Introduced color tinting of clothes, to help make people unique
Removed up front avatar customization, to get people into gameplay quicker
can change character appearance once in game

selling tools to improve minigame performance tanked, except for combat
people just didn't 'get it'
powerups in the field are understood, but equipping parts beforehand is mystifying
selling *silly* tools for 3d world worked

wanted to protect kids from in-game bullying
snowballs have no hit reaction, just a particle effect
GMs are ever-present
BUT -- kids love to grief each other and will pay for it
Compromise: effects caused by other players are cancelable
Someone else places a tombstone, which turns everyone nearby into a skeleton -- but if you want to show off your brand new outfit, you can cancel the skeleton effect on yourself.

  • 1
Lots of interesting stuff here. I'm trying to synthesize.

It sounds like casual players don't play to power-game. They don't play to get 'better' at it or to learn it or to achieve. They play to have fun, and possibly unlock new shinies, and to express themselves in front of their friends.

This maps well to social behavior in the real world, with many people happy just to know what to do in life, and have a way to say, "This is mine, you can tell by the yellow ball on the antenna", rather than striving to become better at their job or build the biggest house or whatnot.

Makes me wonder if that's a real correlation with personality profile, or just happenstance, or if it's more complex (e.g. people I know play casual games when in a casual mood, and grind/powergame when they want that instead).

Currently working on mechanics to automatically group and ungroup players when appropriate

Reminds me of one of the ideas I was playing with at Trenza (never implemented, since we didn't survive long enough): adding extra layers to a "friend" list.

Buddy lists were critical to the design I was building, since it was an unsharded 3-space -- the people you "found" to interact with when you entered a space was strongly influenced by your f-list. But one thing I quickly realized was that an explicit buddy list wouldn't cut it, so I wound up adding "acquaintances" as a formal concept.

The result was a somewhat analog scale of who would be visible to you in the space. Explicit bans would score very low; explicit friends very high; and people you had previously interacted with in conversation somewhere in-between, scored based on how much you had interacted. The theory was that most people would fall into the latter, and the result would be mostly-implicit buddy lists.

One of these days, I may pull the idea out and play with it more -- the nice thing about Trenza going under is that the patent on this stuff never did get submitted...

  • 1
?

Log in

No account? Create an account