For games that will be used for analysis, a complete and accurate record of the chronology of the gaming is critical.
Several means are proposed below for maintaining a record during game play; generally a combination of these means will work best. Generally a scribe should maintain a written gist of discussions. This is a "failsafe" measure: very little can go wrong. This can be augmented by audio or video recording. However, these do come with some disadvantages that should be carefully considered on a case-by-case basis.
Records maintained by members of the Study Team should always be supplemented by various means to gain feedback directly from the participants.
As earlier as possible in the study, a workbook for the Study Team should be started. As the study unfolds this should be maintained so it contains a record of meetings, decisions, and workshops.
The Project Workbook should also maintain a running list of constraints, limitations, and assumptions (CLA). At the conclusion of the Project Workbook should contain the full set of CLA, however at some point these can be reduced to only those that, in retrospect, affected the application of the study results. If uncertain about whether some CLA may have affected results, include it in the final report.
An investment in maintaining an accurate and complete workbook will be recovered in the timely generation of reports following the war gaming. Substantial parts of reports can often be little more than a cut-and-paste from the project workbook, with some light editing.
The main problem is that the task of maintaining the project workbook will seem menial, and a task that can be easily deferred. It can be seductive to leave recording the current issues until time is more available. However, it is best to record events as they happen. One member of the Study Team should be the custodian for the Project Workbook, although any member should be able to insert material.
During the play of a game (Step 12), a scribe should be tasked from the Study Team to gist the discussion as it unfolds. The scribe should be sufficiently familiar with the material that the gist can include flags when particularly relevant points emerge. Upon conclusion of the game, those flags can be used to "rewind" the chronology to certain events to see how play unfolded.
It is usually beneficial to maintain the gist on a computer so text can be edited directly into reports come the time. However, a hand-written manuscript may be preferable: (1) it means no computer screen impedes the line of sight from scribe to other participants or to maps and charts that may be around the room, and (2) the scribe may be able to work faster by hand than over a computer. If the gaming activity goes on for long periods without break (not a recommended practice), procedures should be in place to rotate personnel through the scribe position.
While other methods may be used as well (see below for audio and video recording), a good gist provides the study team with a means to readily jump to key portions of the record. Means like audio and video recordings still have their shortcomings and a gist taken by a motivated and knowledgeable scribe will provide immediate context when the participants or study team have to review some part of the proceedings.
There are four main problems:
(1) The scribe may not be sufficiently familiar with some of the more arcane parts of the discourse and so may miss the significance of some of the debate. Some of this may be in jargon terms if the facilitator does not enforce rules to explain terms that may not be common knowledge. So a poorly trained or inexperienced scribe can miss critical points or, worse, may even record them in a way that will subsequently mislead the participants and the rest of the Study Team.
(2) Maintaining the gist may become tedious after a time and the quality may diminish accordingly. A scribe needs to be motivated throughout, or there needs to be a pool of scribes who can rotate the duties.
(3) For efficiency the scribe may use codes for certain items (e.g., initials for the various speakers). If impromptu codes are used, either these codes need to be familiar to all, or the scribe who used them must be available to translate when needed.
(4) There may be side discussions that the scribe cannot follow due to the physical layout of the room. Or participants may direct remarks to each other and not speak up so the scribe can follow (the facilitator may have to intervene to have participants repeat points for the scribe, or may have to give the scribe a synopsis of any discussions that might be missing).
The onus will be on the facilitator to ensure that the scribe can maintain a good gist. If participants use unexplained jargon or speak so the scribe cannot follow, it will be up to the facilitator to have the participants assist the scribe to maintain an accurate and complete record.
Audio recording can collect a participant's debating points exactly how they were made -- for good or ill. Digital audio recording has a particular advantage in that voice-to-text software may be used to quickly generate a written record (e.g., using software like Dragon Naturally Speaking). However this is not yet technically perfect, and subject to more errors when a debate gets particularly energized and several participants are talking at once.
Microphones need to be positioned at several well chosen locations around the room. Recording equipment needs to be accessible to technical staff to do sound checks and to inspect batteries and wiring. The microphones should be designed for "conference room" debates, otherwise they may not be able to pick up all the conversations the Study Team will want.
Audio recordings provide the participants views in their own words -- often useful in resolving ambiguities over what has been said. With audio recordings, there can be no question of the words a participant may have used; although there may still be a matter of interpretation. This may be aggravated when there may be incomplete ideas as a result of the intensity of debate, where key phrases have been obscured by poor quality recording, or when several participants were talking at once.
The microphones cannot record what they cannot hear -- e.g., conversations at breaks or out of range of the equipment. So microphones need to be in good locations. Also technicians may have to check equipment regularly to replace batteries or recording media. If there is no time-coding embedded in the recordings, it may be difficult to scan quickly to a portion of particular interest. There may be potential legal issues of recording conversations (make sure all are aware that recording is in progress, and that they acquiesce) -- provide release forms if necessary.
Video cameras and microphones can be set up to cover the room or rooms within which the game is conducted. If there are multiple cameras, an editing suite that can handle multiple video tracks will be required to extract the relevant parts of the discourse by switching cameras to follow the discussion.
The record can be augmented with video excerpts of the proceedings. If there are ambiguities in the record, referring to videos may clarify the situation.
Many potential problems with video recording are similar to audio recording: