One piece of advise I always give, is never ask if something is possible, ask how long it will take.

To a good developer, everything is possible. Programmers know that (almost?) everything is doable with significant time to develop, their is almost always some new way of doing something that hasn’t yet been invented. So if you ask is this possible, the correct answer is (in 99.999999%) of the cases, yes. Exactly the same for design, art and all the other fields.
For example:
Is it possible to paint the roof of a massive church with an amazing stunning visage that will last for centuries and be considered a masterpiece, well yes (see the Sistine Chapel), you just don’t want to know how long or how much it will cost.

The more interesting question, “is how long will it take?”, if the estimate involves the words, “I know how to do or X seems to do that quite well”, then with luck the production team can sleep soundly. However if it’s considered “not sure I’ve seen anybody do that before”, then ignore any time estimate, its research and while given infinite time and resources likely solvable, infinite tends to upset spreadsheets.

Unfortunately many producers can’t handle that honest unknown assessment, Its ingrained into them that, “we need estimates, we need milestones” which make no sense at all for research. Research can take wrong paths and simply not go forward at any track-able pace, so the concept of milestones fail (a milestone is meant to show a linear extrapolation towards the result, but in research situations it’s not necessary linear progression, so extrapolating based on milestones is dangerous at best).

This leads to a fairly reasonable answer that then no business can do research. After all how can you cost something that’s unknown?

Well if you step back and reject some traditional thoughts about production, its is indeed doable.

1. Trust – you can only let those you trust, not to waste time, to have the lateral thinking and insight needed and the mindset of ‘playing’ with the field. Hiring a researcher is hard, as those qualities are often only noticeable after you’ve seen someone through the fire of a product release, the only other guide, is have they done research elsewhere. That trust is vital, as you have to know they are working on the problem and have to ability to evaluate if something can be done (and honest enough to tell you). A good researcher may spend 6 months on a promising idea, to find it falls down at the last hurdle and that problem is fundamental and will take a long time to work around and so suggest its not worth continuing.

2. Cost – you need to realize that the cost of research (in games anyway) is mostly salary which you do know, the only unknown is how long you’re paying for them to find something. Once you’ve decided that someone is in a research role, how long it takes becomes to a certain degree unimportant. The cost is fixed, the outcome is variable. The hard part comes from where you place that cost in project based costing, my advise is usually to treat it the same what you cost HR staff and other non project related people, its simply a cost of doing business, making sure your studio is pushing frontiers in areas the company feels is right for them. To make it more palatable (at least in UK and Europe) there are tax breaks and even grants to encourage research, having researcher roles on the books makes that much easier to claim back.

3. Not dependent – you have to decouple it from any important dependencies, if you’ve sold your project to a publisher on a research topic, your risk level just went sky-high. Research should feed back into products when its ready not when its needed. Its always tempting to promise something amazing that’s halfway through its research, but try to avoid it, wait till everyone has decided “this works in all cases and is awesome”, before selling it to clients.

4. Inquiring mind vibe – without milestones, how do you know they are doing anything? Well its usually easy, a researcher will talk and ask questions to other people, when you ask them hows it going, they will get manic or depressive and go into a long explanation of whats currently good/bad. As a producer you don’t even have to understand, all you have to know that quiet is a bad sign. Research is inherently about the thrill of discovery and telling someone is a hard thing to keep in, if they are quiet (whether happy or depressed) its generally a bad sign (though not just a one-off quiet period, give it a little while, everyone has bad days/weeks!).

5. Communication – Even failed research is often very enlightening, so a great way of understanding where things are, is to encourage communicating with the rest of the company/team etc. However try not to do regular director or management presentations, let them speak to there peers who will understand it. Often things just aren’t presentable in a nice form except to people who understand the field well. Mini tech presentation in lunch hours, or wiki with work in progress notes and feedback comments, allow you to get those not in research also stretching their mind and hopefully speeding the research up as more brains is always better (ask any zombie ;) ). Its good to show everyone (especially those paying the bills!) when something is demonstrable, but forcing it to fit some non research schedule, doesn’t help the research. Nothing wrong with asking “got anything laymen can see”, but just allow the answer to be, “erm no not yet”.

6. Feel useful – After a piece of research is done, find someway quite quickly of using it or at least acknowledge its usefulness. A research piece on a new control system might not be applicable for the current games, but how about a mini game using it or a conference/blog post on it? A piece of graphics research might be too demanding now, but books and conferences allow it to be useful in the wider sense. Often the idea of giving something away after paying the research is hard to swallow, its hard enough to convince that paying for a researcher can be useful, let alone then giving those results to your competitors for free! But usually the secondary effects, of improved company PR and status, having others improve and adapt it and just that it will improve team morale are worth it. The other fact is that often its the researching itself not the results itself which are most valuable. When you do come to use the results, having the person who knows it from top to bottom and all the caveats is the ideal situation. The worst situation is to bury the research after its done, nothing says “why bother?” more than spending some time on something to see it put away like the Ark in the first Indiana Jones.

Having people doing research is a really good way of getting experienced staff brains working on stuff that really stands out. Some people might like it just as a break after a long hard project, others it’s a career move that doesn’t mean management or a the burdens of continual shipping.
Most good games developers are creatives, letting them explore that away from the often soul-destroying reality of shipping products with its harsh rules of cuts, marketing decisions and other fun elements that we all love about the “business” is a great way of keeping talent and keeping fresh ideas flowing through a company.