The Business and Design of Free-To-Play Games


Finding Fatal Flaws: Lessons from Kongregate – CC09

Here are raw notes from a discussion of lessons learned from Flash games on Kongregate, presented at Casual Connect 2009 by Jim Greer (CEO) and Greg McClanahan of Kongregate.

Kongregate hosts 16,000 Flash games, 1,000 are added monthly. It's oriented towards core gamers and includes achievements, friends list, chat, comments and ratings.

  • 85% of Kongregate gamers are male, majority are teenagers & college kids.
  • 8 million unique users per month

Game plays are not as long-tail as expected:

  • 1st game - 12m plays a month
  • 2nd game - 10m
  • 20th game - 2m
  • 60th game - 1m
  • Top 10% of games - 90% of playtime
  • Top 1% - 50% of playtime

Regular players vs addicts

  • Epic War 3 - deep single player game - 33% converts to regular players (5x plays), 10% converts to addicts (20x plays). Conversion is low because players already finished the game by that time.
  • Dolphic Olympics 2 - Tony Hawk style game - 24% to reg players, 13% converts to addicts - Did well for a single player game. It's not IMMEDIATELY obvious how great/fun the game it is, players need to get good at it first. When put badges into this game, ratings skyrocketed since people were pushed by achievements to try out other areas of the game.
  • Platform  Racing 2 - 38% to regular players, 27% to addicts. Simple platformer mario-style that won user's choice award. Has UGC and long legs.
  • Starfighter: Disputed Galaxy - 32%, 22%
  • Free Rider 2 - 17%, 18% - Once you get over hump of creating levels, it is more likely to convert you.
  • Desktop Tower Defense - 7%, 15% - Have to build lots of towers before getting the value.

When putting in achivements, more often ratings go down. When people really care about your game because they like the achievements, problems (such as with controls) become more obvious. Or people play games they don't like just to get the achievements.

Newgrounds vs Kongregate ratings
In many cases similar, in some cases not. Suggested rationale for difference included.

  • Desktop TD Pro - 3.9 vs 4.4 (NG vs KO) - KO: gameplay is awesome, so it's ok we already saw the game before
  • Protector 3 - 4 vs 4.3 - NG: we've already seen this
  • Post it Draw it - 3.7 vs 4.3 - NG: wtf
  • Epic Battle Fantasy - Parodies of Final Fantasy fights - KO: ripping off FF, and is not so fun
  • Nuclear Eagle - 4.1 vs 3.7 - KO got frustrated by not being able to finish (too hard)
  • Xenotactic - 4.4 vs 3.9 - NG, someone reskinned DTD
  • Thing Thing 4 - 4.4 vs 3.7 - KO about 5,000 people who uploaded stuff to the site (8 m)
  • NG has higher % of people who consider themselves developers. So when they see something technically impressive, they react well.

Flaws in Flash Games

  • Need a mute button
  • Have save files, and pause buttons (users really want all of that)

Build fun first
Lots of new devs think it's like doing a homework; that putting in enough 'good enough' work in each area gets your game to be  fun.

Button Hunt 2 - 3.83 rating, 760k gameplays
Shopping Cart Hero - 4.07 rating, 2.2m plays
Don't Shit Your Pants - 4.01 rating, 835k plays

New devs - The product counts. You are NOT paid by the hour!

Players don't read instructions

  • It is developers responsibility to teach, not the players
  • This is something that friends/family won't critique you on
  • Have a dialogue that pops up when you learn a new mechanic, don't assume that people are going to remember instructions from the title screen
  • Allow flexibility with controls - say use WASD but also use the arrows

Super Energy Apocalypse Recycled - 4.11, 650k gameplays

  • Two thirds of the game is a tutorial. A great way of doing it as opposed to 'We hope you read the instructions on the title screen'

Challenge does not equal frustration

  • It's a virtue to have a good difficulty curve
  • Bad: POD, hidden timer
  • Good: N+, high skill component, reward players on regular basis, then game can be brutally hard

Art games work without much gameplay - Sprout, Anika's Odyssey - Beautiful, bizarre, text-free, remain intuitive and enjoyable

Good strategies need to be the fun ones
Players will play the game to complete the objectives, not to have the most fun

Ether Cannon 3.27, 184k
Ether War: 4.11, 838k


  • Style, aesthetics, relatability are more important than impressive tech
  • 3D graphics do not impress, since players already play console games
  • The final 10% is the most important - Tweaking, and getting reliable feedback from non-friends


Club Penguin Founder Lane Merrifield – Keynote and Q&A

Here we are at Austin GDC! First order of business was attending the Lane Merrifield Keynote, entitled "At Their Service: Making a Difference By Putting Players First," where I took copious notes. I have 40 mins while our first video interview (Mike Zummo of Acclaim) downloads from our camera, so here are my notes, cleaned up a bit:

Also, here is our photofeed from today, updated when we can.

Lane's Talk

  • Lane's plane landed at 4:30 this morning due to Ike travel complications. He jokes that if he's slow today, it's not because he's Canadian. I take minor offense, being Canadian myself. :)
  • By putting the player first, better customer support is natural follow through. But we also build better games and stronger businesses.
  • Lane is a huge fan of Walt Disney and was with the company long before they acquired Club Penguin. Lane first job was manning a float in the Lion King parade at Disneyland (shows his employee pass from Disneyland, including Flock of Seagulls hair). He rode the float every day after school and also controlled a 300-pound animatronic crocodile.
  • While at Disneyland, Lane observed that trash got picked up fast and drinks got replaced instantly when spilled. All the staff had a great attitude of service and took care of each other behind scenes. The environment was all about taking care of each other and service.
  • The passion for serving at Disney came from Walt Disney himself. He built it for his kids. Without realizing the parallels, Lance and Lane built Club Penguin for their kids to play in. Lane strove every day to find news ways to make it safe and fun for his kids.
  • Lane makes the point that "Serving must be genuine" - providing his dad's "shirt off your back" gauge of someone's service integrity as proof. In a snowy Canadian winter, someone who would give you their shirt is gold.
  • (Big story about how he's called Jason at his local starbucks. Point was likely about Customer Service Reps (CSRs) not taking time to know their customers.)
  • Lane then reads from a note he sent to his development team after Club Penguin initially launched amid the usual dev team chaos and late nights. The letter was a story about when Lane saw parents and teachers shepherding kids across a snowy road. His analogy to the net was that some companies thought crosswalks and stoplights were the solution to keeping kids safe on net, but kids still felt isolated and unsafe and confused. Danger still existed to those trying to cross. Club Penguin believes kids need advocates - holding their hand, walking with them - and Lane hopes that as a company they always go one step further to pick up that hand and cross the street together.
  • "Crush the Joy" - Lance, co-founder of Club Penguin, coined this phrase. Lance is convinced that before anyone joins the team, they need to have their ego turned down a bit. Hence, Crush the Joy. Dev team members need to work together to serve audience and let go of ego that gets in the way. Devs often wind up front and centre of what we build - we end up building for ourselves and not the player.
  • Club Penguin has worked tirelessly to find people who had a passion for serving kids. Without it, they wouldn't be hired.
  • Pixar has a saying: we put people through a gauntlet to get in here, but once they are here, we treat them so well they don't want to leave.
  • Club Penguin was once fired by a staffing firm as CP consistently turned down highly qualified candidates who weren't right for the position. CP has hired a lot of CS people from Starbucks because they know how to talk to people face to face.CP also hires people out of summer camps, boys and girls clubs, teacher's assistants, etc to take care of player base.
  • Most of the CSRs have rarely done more than email and web browsing with PC before coming in - deep tech knowledge is not as important as people skills. CP does not use automated responses... CSR team personally responds to 5-7000 emails a day.
  • Programmers stand over CSRs shoulders to measure things like mouse travel to make minor changes to CSR interface to optimize work flow.
  • Because real people answer player inquiries, players are more likely to relay qualitative info back to CSRs about issues like game difficulty, suggestions for "white fur" found randomly in game, etc.  CP approached this feedback in a revolutionary fashion: "we didn't do what the kids said would be lame and did do what the kids said would be cool" - applause
  • "If it doesn't matter to a kid, it doesn't matter" - this is the mantra CP works toward. Easy to get distracted and focused on ourselves - easy to get caught up making what we think is cool or what our friends think is cool, but need to remember what the kids think is cool. Should always be striving to serve our audience. That will create better games, work environments and a better business as a result.

Q&A Period

How do you differentiate between vocal minority and whole audience when acting on feedback?

  • CP deals with a certain demographic that can barely type and read. Full time staff reads blogs and forums, play testing is done in office, etc.

Do you use any metric systems for pooling user data?

  • CP does not have a lot of systems for that. We rely heavily on people... emails come in by the thousands focused around certain topics, CSRs gather that info and compile top 10 issues and send out to dev team.

Does CP have any plans to do microtransactions?

  • This is a tough thing for CP to do, due to age of audience. Subscriptions are easier for parents to moderate spending. Subscription model makes it easier to focus on parental needs, versus an ad-supported model. For instance, CP just launched a timer that kicks kids off after a parent-determined aount of time.The cool thing is that there are no advertisers to get mad at CP for doing that - parents wanted it, parents got it.

Any other recommendations for fellow kids world devs, besides serving?

  • Know your audience, keep them in the forefront. CP is fortunate we did not take on any VC as we didn't think it was going to go as far as it did. We used our own credit cards to start it up and didn't need to be accountable to anyone but parents and kids. There were no VCs in our ear talking about the latest research and focus groups, etc.

Would you like to open up new avenues for users to give feedback?

  • There is never a bad way to get feedback from customers. CP also had phone support to talk directly with users. They are open to any new method for feedback.

You slated a percentage of profits to go to charity in the beginning. Was that maintained when you were acquired by Disney?

  • Yes it was maintained and continues to this day. CP does not talk about it a lot as we don't want to turn it into a marketing pitch. "Coins for Change" was our first public charity event - it allowed kids to give their coins to one of three charities. Then, CP took $1M and distributed it proportionally according to how kids put their coins between the three charities. Two billion coins were given, which is remarkable due how difficult it is to acquire coins in the game. Disney never questioned keeping the charity program.

How do you give kids what they want when, depending on the age group, kids can be bloodthirsty llittle monsters? How do you distinguish about what they say they want and what they really want?

  • Kids always ask for beds for their igloos, but it leads to inappropriate stuff so CP will never do it.

What are some of the key personalities or traits you look for when hiring?

  • A lot of it comes from gut If the people doing the hiring are not the type of person CP is looking for themselves, it gets a lot harder to hire the right people. Up in Canada, it's a very relaxed environment well suited to producing those sort of people.

Does CP appeal to an older audience?

  • CP actually has a wider demographic than people realize: 6-14 years old. It takes on a geeky cool thing at the older ages, like Napoleon Dynamite. Older kids spend more time on Xbox but will still hop on for snowball fight wtih friends in CP.

What percentage of your budget goes to Customer Service?

  • Two thirds of our staff, 150 ppl, are dedicated to CS. We are opening offices in Brighton, Australia, Brazil, etc.

What kind of predator protection tools do you use?

  • Predators are not much of a problem for CP due to our filtering tech and massive human presence on CSR. We remove several hundred words a day from system. i.e. "lollipop" takes on undertones due to pop song, so we remove it. Then we review it a few months later to see if it can be reintroduced. A lot less gets through than people realize. Reportable incidences are how these issues are measured and to date CP has not had one. Nothing is ever perfect, but in a sea full of unlocked cars, CP is low-jacked and locked up. Plenty of easier targets for predators.

With two thirds of staff being CSRs, are costs a concern? Would you ever outsource?

  • CP worked hard to build a scalable model and work on efficiencies in CS. We support English alone in Australia, UK, North America, so it's really important to have people in each country that know that region's slang, etc. CP is so passionate about CSR system that we have never brought up outsourcing it.

How will your product evolve to accommodate a more mature userbase?

  • CP has a substantially large writing staff working hard to make sure there is relevance across the board, but realistically, CP will never be able to compete with a 17 year old's console gaming habits, etc. So at some point we need to let them go. Now with Disney we can look at developing other products for those older demographics.

I heard Penguin Times is the most widely read newspaper in the world - is that true?

  • An article recently compared it to a lot of leading newspapers and it is substantially large, but CP does not keep a lot of hard metrics on it. CP did track it for a while and it was impressive - especially intimidating to the 2-3 writers who work on it full time.

How did you come up with the backstory you have?

  • Story is everything in entertainment. From day 1 we had a writer who started as a support person. She built a team, but a lot of the ideas come from the kids. Lot of logical - and illogical - ideas. CP introduces characters very slowly - we don't introduce 30 chars at once and call it a day. 90% of our internal conversations revolve around what CP can't do re: story, but ultimately the 10% that makes it in, makes sense to audience.

How often do you get feedback concerning the flash platform?

  • Lance is great at working backwards. CP was originally developed in Flash 6 and most of it would run today in that environment. We've always looked for the highest install base and built for that. One of the biggest issues is ports not being open on routers, so we spent a ton of time making sure it just works for everyone. New issues are constantly coming up - i.e. new browsers coming out - but our audience is quite patient as we work through issues.

What other non-English markets do you plan to launch in?

  • CP has 7 or 8 translations in the works now. None are public yet, but our goal is to roll them out quickly. That is one of the reasons CP joined Disney. In the next 6 months there will be several announcements around that.

At what point did you realize you needed more than out of pocket money to move forward?

  • We started CP with aggressive budget, low salaries. We worked really hard to build a scalable model, so as audience grew, CP could cover costs. There were lots of VCs out there, but CP needed infrastructure help not just money. John Lassiter and the crew at Disney were always consistent with what they said would happen and are still consistent.

How can you handle the support tools and requests from other regions and languages?

  • That is one of the reasons we can't just push a button to launch in other territories. We need to ensure the tools work for those regions. CP brings people from that country to their office for 6 weeks, then sends people from CP to the regional office for 6 weeks.

What was the biggest challenge working with Disney?

  • The largest challenge CP has was the size of Disney. Every Disney division has a 50:1 ratio compared to CP's team, so it was easy to be inundated with emails due to difference in size. Bob Eiger called Lane after acquisition and said "If you ever feel that Disney's size is stifling you, I want to be the first to know"

What level of importance does mobile play have for your audience?

  • We are trying to track demographics of our audience and how fast they are adopting mobile. We have a few things in the works for mobile now, but we want to make sure it is done purposefully and not just to check a box. Need to make sure it's as fun or more fun than PC.

How do you deal with banning players?

  • One thing we do is to purposefully keep CSRs who ban from seeing whether it is a member account or non-member account, to prevent bias from creeping in. We issue 24 hour, 72 hour and permanent bans. We're able to see not just chat logs but also what a user was trying to say - i.e. what didn't make it through filters.