Saturday, April 29, 2006

How to Find a Good Game Designer

Hiring good, experienced designers is hard. It is so because design is seen as such a nebulous stitching-everything-else-together task that it's not as easy to evaluate via the interview process. An animator can bring his showreel, a coder can bring his samples, and an artist can bring his album. Hell, even a producer can bring his sample schedules and track record of games produced on time and on budget etc.

What does a designer bring? Levels that they've done (assuming that they're level designers), perhaps articles or books that they've written, videos of their former games in action, a meta-critic score, a list of linkedin endorsements, a list of titles? Do any of these actually tell you anything?

Levels actually tell you very little. A level video is played at optimum speed, often edited and the prettiness of the art etc can be quite distracting (in both good and bad ways). The level design of many of the corridor sections in Halo is basically corridor-enemy-corridor-enemy, for instance. How do you show that it's "good", especially in an interview setting, and distinguish its goodness from, say, the enemy AI scripting or the balance of the game's weapons which you may have had nothing to do with?

Articles and books can be misleading. They show a mind at work, but not necessarily the right sort of mind. Quite a few journalists and academics have forged into design, only to find that they aren't good as designers. Speaking as an author of a few airy articles in my day, does my output really tell you anything about whether I am actually able to sit down and design a game? Not really.

A list of titles and achievements can mean anything and it's very very easy to big up one's involvement in a title to sound more than it is. For instance, judge

"I designed half of the levels for platform game X".
"I designed over 50% of the crucial levels that proved critical to the testing and implementation of the gameplay as well as developed the narrative backbone of the game."

Academic qualifications can similarly be over-emphasised, software skill lists are nice but rarely relevant as most design-related software can be learned quickly enough, and references invariably say nice things. Previous experience in other disciplines is also nice to have, but again no real indicator of anything in terms of design skills.

While all this material is good to bring along and everyone expects an interview to have a bit of razzmatazz and spin associated, they can all be faked to a great extent. It's an easy lie to say that a bit of design work in a game was yours. Who can really gainsay that you were the one that sorted out the balance issues with the plasma rifle, or that you were the one that set out the length of the levels, or drew the 2D maps and workflows. Who can really say exactly how much of the sample GDD that you've brought along is actually your work, or the copy of the design wiki?

With the best of intentions in the world, what these different standards of screening design candidates do is reward presentation-oriented candidates. Presentation is itself an important skill for a designer, but the interview process doesn't address the other important skill. (I.e. can they design a game?) And there is much of a consensus across the industry that there are a lot more people who think they're good designers and can craft a spiel about how good they are rather than actually being good designers.

So what to do?

A lot of game designers have good secondary skills, and those are the ones that appear most in the interview. Very few have good primary skills because they're simply not tested in the interview in the same way. Here's the skills breakdown:

Primary: Abstraction, Rules, Mechanics, Constraints, User Interface

Secondary: Scripting, Feedback, Databases, Playtesting, AI, Balancing, Gameplay analysis, Progression, Writing, Enemies, Characters, Maps, 3D environments, Diagrams, Cinematics, Documenting, Pitching, Talking, Scheduling, Idea Management, Market Research, Focus groups, PR, Direction

In my view the interview process lets a candidate demonstrate their secondary skills, but we have no means of examining primary skills. So, we often end up hiring designers who are good on paper, but actually bad designers. The solution is to devise a test. Since the primary skills are all really about fundamentals, a test should also be about fundamentals too:

The particleblog design test
Place the candidate in a room and give them the following items.

1 deck of cards
4 six-sided dice
A pad of paper and three pens of different colours
1 pack of index cards (blank)
1 whiteboard and eraseable marker pen
1 bag of 50 black tokens
1 bag of 50 white tokens
2 table-tennis bats
1 table-tennis ball

And you tell them that they have 4 hours to create a game.

Here are the rules of the test:

1. The game must only use the components presented.
2. The game must be in a playable condition at the end of 4 hours.
3. The game must be playable by anyone (i.e. no obscure knowledge of trivia etc)
4. They do not have to use all the components. If they just want to create a puzzle game using only the whiteboard and the tokens, that's fine.
5. They must not replicate a game which already exists
6. They must write out the rules of the game, because...
7. They don't get to present the game

Rule 7 is absolutely vital, because it removes the personal sales pitch from the process. Ideally what the designer should do is hand the rules to the interviewers and go home, leaving the interviewers to learn and play the game for themselves. The objective of this exercise is to test raw talent.

If the game that he invents is fun (or potentially fun), then he has talent. If it is not, then he does not, and all the pitching and documentation and PowerPoint presentations in the world are useless in the design context. That candidate would probably be better suited to a PR, production or creative writer role.

The game that the candidate invents would also be highly indicative of the type of designer that they are. A candidate who invents an interesting sport, for instance, is likely to be a good Type 1 designer. Whereas, a candidate who creates dungeon-crawling roleplaying game is probably going to have good Type 2 skills. A candidate who creates a game about building little societies would likely be a great Type 3 designer, and one who creates co-operative game for several players would do well in a Type 4 environment. It's like a Rorschach test for game designers.

At least that's my guess anyway.

So is anybody up for the challenge?
If you are, assemble the components, set yourself the time limit and make the game. Afterwards, post the rules here in the comments section and we'll evaluate them.

Particleblog's comments have moved to The Play Room.