Psychology and IGRORAZRABOTKA:

Obtaining data that can improve GAME *

Bill Fulton

End. Start cm. # 6’02GAME.EXE

EDITORIAL: The first part of the publication was devoted to a brief description of the problem "intuitive approach" to game design, it has been shown the benefits of the use of professional psychological methods to collect data on the players and set out four criteria for an adequate feedback from users. Now you will learn how on the basis of these studies, Bill Fulton built his work group at Microsoft.


There are many feedback systems used by game designers around di (or, more precisely, that they have to use it). Most developers, like all authors are aware that the responses to their projects are useful because they allow to improve their quality. Moreover, TVO-rhenium most authors reach a quality worthy of publication only after processing, based on these same responses. I’ll list the systems that I know, and each will try to determine on-how good it is.

There are two main categories of feedback: responses received from game industry professionals, amateurs and evaluation (ie ordinary players). Although these sources will certainly affect each other, it is convenient to speak of them separately.

The feedback from the professionals

There are two main sources for-feedback:

1. The team itself developers NE-one to the main source of feedback: the people working on the game, continually give out comments such as «this character — no,» «a weapon too powerful,» etc. This is, without doubt, useful information as perfectly meets the criteria of the second to fourth (feedback there on time, it is quite detailed, and get it quite easy), but the criterion number one it is, alas, does not correspond to: who knows how many u -rokov agree that the weapons too powerful?

2. Feedback from industry experts. Consultants Game Design, me-managers in the company, the publisher, log-sheets — they may be useful to express in opinion. This feedback is able to satisfy the criterion of the number three (ie be moderately detailed), but with the second criterion-Receiving (occurrence time) is usually a problem: often the period before the response is long, so that the recommendations will appear after you they can use. And, of course, an exact match responses with the opinion of the players once again called into question, although in this case it is possible predpo-pose that professional judgment will be more accurate than the developers, because they deal with a lot of games.

Although feedback from the profession-channels — a typical tool for most of today’s teams to work fine on the criteria of two, three and four, it largely Ba-inrush on faith and hope, when it comes to the number one criteria

— How accurately it reflects the opinion of the opinion of the pro players of users. The reason for this is easiest to show a simple thought experiment. What do you think, how many games a player sees a typical year? And what is the figure for the professional? They differ in at least one order of magnitude. Pros make up one percent of the same players, the most knowledgeable about the games, whose tastes may just be too much developed (and exotic, note), rather than the typical player. Yes, of course, some professionals may phenomenally well predict the preferences of the masses, but where to find such? ..

All of this leads to the fact that the responses from the professionals (although they be-zuslovno needed) is not sufficient, it is not the best method for defined-division of whether players like project or parts thereof. In the end, all that they say,

— It is their personal opinion on it?

Responses from non-professionals

However, the developers themselves know about the problems with their own opinions, do not necessarily reflect the tastes of the players. To improve D results, they often try to get feedback from those who are likely to give them a more accurate and obvious candidates for this role — the same users. Traditional ways in which to receive information from them are listed below, and simultaneously given and an analysis of these four methods of our criteria:

1. newsgroups, beta, fan mail. In the simplest case it is sufficient to carefully read Internet forums, watching the reaction of the players in the game. The main problem with the second criterion (timeliness). That game came to the players, it should be at least in the beta, if at all finished. Therefore, this kind of feedback comes too late to change something is almost impossible. In addition, information is often, to put it mildly, is not sufficiently detailed («This character — a complete city!»). Perhaps only one thing good here: responses are rather «cheap» — and on time, and cost.

2. Testing for, relatively speaking, a relative. Typically, this method involves ispolzova-set far from the industry people (born-tion, neighbors, etc.). They sit down, ur-rayut — and tell you what’s wrong. Such information is usually detailed enough, can be quite accurate, but often do not always act-Hoc (because of problems with the search-BPE to time on its organization) and can be quite costly because of the large amount of time.

3. Focus groups. Typically, this type of feedback has been from the giver. He gathers a small group (four to eight players), and once-used to say to them about the game. They can give play to the current version, but this is not mandatory. A typical problem here — late as usual focus groups connected in the final stages of the project, and the results are usually not too detailed. And, unfortunately, the cost of the work is very large.

Thus, communication with non-professionals potentially very useful, because it involves the study of people just do not have anything to do with the industry, which will play the game. But with this method, and a lot of related problems. For example, it is not clear how feedback reflects the views of the typical players (because, say, forums wrote a very definite kind of people, and during a lively conversation players feel compelled to say something good), and people who are testing often do not have proper education in psychology, so that the results can be accidentally «shifted» in an arbitrary way. For information on how to minimize these problems and create a feedback system, which works well on all four criteria, I’ll tell you.



We at Microsoft calls the process of game designers responses «user testing», and people who are engaged in it, experts in the field user testing.

Experimental psychology studies how Therefore gg ~ 7

J | Mech Commander 2.

obtain meaningful data from people over 70 years, and our group is trying to use these works. Of course, not every psychological study — good research, just as not every code — good code. Psychologists — the same people, and their results vary in quality. However, Su-recognized methodologies exist-cal doctrines that have proved their usefulness and our process is time-ops primarily on them. Lest you fall asleep from boredom, I will not try to put 70 years of research in the report, but just try to describe the daily work, which deals with our group telskogotestirovaniya user.

Methods …

Group test provides three main services: usability testing, testing gameplay and reviews. Below we look at them in detail.

1. Usability testing the use-tion is usually associated with Neboli-Shimi groups of subjects, of Coto rymi monitored. Within two or three days we visited six to nine people — they come to the individual two-hour session. In a typical study, each of them spends some time studying the game is free, and then performs a series of very clearly defined tasks.

Typical results include user comments, they are keeping, the time spent on tasks you-complements, and error rate. Usability testing is-use — an excellent way to identify problems that originators time simply did not know and to understand the thoughts and feelings to play and how they reflect the on-their interaction with the game. This type of testing is-uses in the software industry has a lot of years. It usually refers to the area of ​​human-computer interaction, not psychology, but the methods that are used in it, it is easy to trace them

* Once again we remind you that the proposed to your attention the publication of the essence of the report of the founder of the group user testing in the gaming division of Microsoft, Bill Fulton (Bill Fulton), read it on the March of this year, the Conference of the developers of computer games (Computer Games Developers Conference) in the session, dedicated to game design.

Like this post? Please share to your friends: