I hosted a roundtable discussion on designing online puzzle games at the 1997 Computer Game Developers Conference in Santa Clara, California. The attendees included a few who are already making online puzzle games, more people who are thinking of building or producing online puzzle games, and many people more generally interested in puzzles, especially for education. Here are highlights of the discussion.


Although puzzles enjoy wide popularity in newspapers and magazines, they occupy a small niche in the computer game world. The internet is about to change all that.
   Web developers are finding puzzles a good way to draw repeat visitors. The broad appeal of puzzles reaches beyond hardcore gamers to include women, kids, adults, and people who don't like violence. Puzzles are quick to build, fast to download, and don't require fast networks.

S U N A P R 2 7

75% of the people who do crossword puzzles and other word games are women. In general, the audience for puzzle games is largely female.

A few puzzle games have become major hits. Besides Tetris, The Incredible Machine continues to be a steady seller many years after its release. In educational software, The Castle of Dr. Brain and The Oregon Trail have sold large numbers (hundreds of thousands).

In the future retail software may go toward lower-priced packages in jewel boxes, which will mean lower price points and more room for more titles. That's probably good for puzzles software, which can afford lower development costs.

What is the fundamental appeal of puzzle games> people like to optimize, to work out systems. We do it every day. Look for constraints in your experience and build puzzles around them. One of the best examples is programming, which is certainly a form of problem solving. An example of a puzzle game built around the experience of programming is Rocky's Boots (a logic game from The Learning Company, now out of print).

Other motivations: finding a structure or pattern. Visual appeal, as in Tetris or MetaSquares (on AOL). Time waster, as in Solitaire.

People have many different motivations for playing puzzles: the novelty of seeing what's new on the next level, competiion, mental challenge. Ideally you want all these in one game, but that's hard to achieve. So instead variety packs are popular.

Players want to know that they can do well, achieve some degree success. Keeping the rules simple is important. This can be done partially by dividing a game into smaller portions. Good if you can learn by just watching.

Three kinds of competition: with self, one on one, team against team. Different people prefer different types of competition.

M O N A P R 2 8

From the perspective of an experienced puzzle designer who is new to computers, what the computer offers is movement and sound, which can give rise to whole new types of card games, even though cards have been around for a long time.

How do you determine if a puzzle is too hard? Focus groups. Online testers (players who have volunteered to take on more responsibility).

What makes a good puzzle fun? Should make you feel clever. A hierarchy of progressively less cryptic hints can help. Aesthetic, cuteness appeal. Gradually increasing difficulty.

Are prizes legal? Not equally so in all states. For a contest to be considered a lottery, which is illegal in some states, it must satisfy three conditions: pay to enter, winners chosen by pure luck, and substantial prizes offered to winners. If a contest does not meet one or more of these conditions, then it is not considered a lottery.

Is anyone using hybrid media (Internet + CD-ROM) puzzle combinations? What about Director 6?

Dave Phillips of Playnet has several puzzle games including MEGAMAZE, has a game in alpha he would like to market.

T U E A P R 2 9

Castle Infinity (online environment for kids from Starwave) tends to be self-policing. If someone tries to be a bully, everyone else just walks away. Also, there are adults online with special heads on their avatars that identify them as authorities.

How do you prototype a puzzle to know if it's good or not before you program it? In the case of MetaSquares on AOL, the first time we played it was online. Perhaps that wasn't the smartest decision, but I had a strong intuition that it would be a good game. Furthermore I knew that we could tweak the gameplay after we built the first working prototype. In particular the scoring system was developed after we had played a number of games and started to get a feeling for the gameplay.

What makes a good puzzle game? Alexy Pajitnov of Microsoft commented that one of the most important things to focus on is the rhythm of gameplay. Get that right and you're halfway there.

Copyright 2000 Scott Kim.
All rights reserved.