Field Anomaly Relaxation (FAR) is a technique that maps a interactions of change drivers on a systemic scale, structuring them as 'sectors' and 'factors'. From the interactions between these complex interconnected tree of branching scenarios can be created. See also morphological analysis.

[Scenarios] provide a relatively unbounded forum for creative thoughts and ideas regarding organisational renewal such that novel concepts and imperatives for change could be expressed within a broadened perspective of the future. Many emergency response organisations: law enforcement, military, fire fighters, paramedics and disaster relief, to name just a few, use scenarios on almost a daily basis to sharpen their capacity to deal with the unexpected. Their common characteristic of being necessarily focussed on the short-term does not reduce the valuable use they make of scenarios and scenario thinking. Through a comprehensive set of 'what if' statements and structured events they are able to rapidly cycle themselves through a bewildering array of problem situations. The result of this focussed training and the associated flexibility of mind it demands is a highly agile team that can rapidly adapt structures, techniques and procedures to overcome situations in which indecision, hesitancy and lengthy planning are not options. Pilots undergo multiple repetitions of emergency situations in simulators to create this necessary rapidity of response and action in order to reduce problem-solving to an instinctive sequence of behaviours, in effect allowing the mind to remain clear and unencumbered to deal with those few unique contextual aspects that are outside the domain of previous simulations. These scenarios, and more importantly the use of them for long-term planning, are certainly not predictors of events, nor are they attempting to forecast the likelihood of particular situations they simply aid in the provocation of new ideas. Russell Rhyne, the original architect of the Field Anomaly Relaxation technique, advocates adherence to a four-step process to projecting futures (Rhyne, 1974, 1981, 1995 & 1998):

• Step 1: Form an initial view of the alternative futures that could unfold within the area of interest.
• Step 2: Construct a language using Sectors that will become the dimensions of describing the area of interest; and Factors, which become the alternative states within each Sector and array these on a matrix to form Whole Field (full sector range) descriptors of all possible configurations.
• Step 3: Eliminate those factor pairs that are illogical or cannot co-exist, forming a reduced set of whole field configurations.
• Step 4: Position the surviving whole field configurations on a 'tree' whose branches represent possible future states and transitions from one configuration to the next.

The process adopted involved: generating a large list of drivers, with the criteria being that they must be beyond the organisation's ability to influence; compressing this list of drivers into broadened 'themes' that embodied the main thought being presented as the driver; and then entering step 2. Rhyne (1981, p. 349) has identified the importance of doing something useful with the results (purposeful action is the neat term that Checkland has provided us with) of futures thinking that engages a wider body of thought about change “…the payoff from projections of this kind [FAR] usually dribbles away unless solid uses for them are visualised from the start and pursued The first step was to identify the drivers that would be most influential in shaping events. As stated earlier these drivers need to be beyond the planner's ability to influence, the only acceptable response being adaptation. (…) These (…) drivers were then collapsed using affinity diagrams to reveal the major forces at work. To make the future believable it is necessary to string the transitions into a plausible history that marks out how a future world could evolve from the present. Having constructed the Faustian tree and, thereby, creating a series of pathways through which the single future trajectory may travel, a set of rich narratives are called for. These narratives link the present to the future in a manner that adds substance, depth and most importantly breadth, effectively putting 'flesh on the bones' of the earlier, more structured thinking stages. * page 9

From A Practitioner's Experience of Using Field Anomaly Relaxation (FAR) to Craft Futures -Guy A. Duczynski

The starting point of Field Anomaly Relaxation (FAR) is Lewin’s social field theory to the effect that we all live within ‘fields’ of interactions with other people and events. (…) This notion is used in morphological forecasting (see morphological analysis), morphology meaning ‘the form and structure of anything’. The method was invented and named by Zwicky before World War II (Zwicky, 1969). It uses a formal description, called a Zwicky Box (renamed a Sector/Factor array in FAR) to explore, for example, all conceivable forms of aircraft propulsion. (…) FAR exploits that idea to explore the imaginable patterns within social fields, eliminating any which do not satisfy a gestalt, whole-pattern, assessment of internal coherence. The remaining, internally consistent, patterns are then used as stepping stones to create paths into the future. The steps across the stones enable FAR to generate story scenarios, not just end- state pictures. (…) Each of the components of the field must have several conceivable conditions; economic growth may be high, low, stagnant and so on, differing in kind as well as in degree. The idea is to create ‘filing space’ for all plausible possibilities. FAR is a four-stage process:

• Step 1 requires one to develop some kind of imaginative view of the future into which the decision must unfold (this can be done by several people in the team, guided by one or more questions)
• Step 2 requires one to identify the critical uncertainties and their ranges of possibility, expressed in a sector/factor matrix. That is superficially attractive but the construction of a good Sector/Factor array, in the sense of one that will fully illuminate the strategic question, requires contemplative thought and may need more than one attempt. The analyst may have to question and criticise, very much as the Devil’s advocate, to ensure that the study gets the right degree of simplification into its model of the future.
• Step 3 eliminates the anomalies (‘could I imagine a world like that?’ eliminate if the answer is no, then cluster combinations to get to a smaller amount of 'worlds'). The anomalies must not be eliminated by the team members sitting round a computer. That tends, for some reason, not to lead to thorough discussion of gestalt appreciations. The ideal way is discussion round a table, with the debate taking as long as necessary. The debate is vital as it sometimes happens that an anomaly which has been eliminated should not have been and it may need to be reinstated. However, someone has to keep a record of what is decided and software is needed to perform the eliminations, to calculate the remaining configurations and to print them out for review, and that is the analyst’s job.
• Step 4 strings the surviving configurations (cluster) together to form time lines (‘Can I see this world leading to that one?’). This process may take a week or so, not of full-time effort but of sessions looking at the board and intervals of other work while the ideas filter through the minds of the study team.

As a rough guide, one might say that a one-cycle FAR ought to be able to be done with about 20 person-weeks of effort, with a second cycle, if it is necessary, requiring somewhat less. (…) We have to live in the real world and accept that, manifestly unsatisfactory as it might be in theory, it is often necessary to do the best one can in much less time. We therefore turn to a simplified version of FAR which has been found to work better than one might have expected and is certainly better than doing nothing at all to think about the future. One of the attributes of a good analyst is the ability to produce at least some kind of results by the time at which they are needed and not to pursue rigorous perfection which will be too late to be of practical use. A complementary skill is, of course, to explain the caveats attached to a ‘quick and dirty’ study in such a way as to give the user a sensible understanding of the trustworthiness of those results. In terms of FAR, this may mean that a rough study has to be carried out within a few days as opposed to the ideal of some months. No one suggests that such an approach is fully satisfactory; the point is that it is somewhat more satisfactory than doing nothing to think about the future. Simplified FAR (~2 days) Simplified FAR depends on building outline time lines from a few consistent futures, as opposed to eliminating all the inconsistencies and forming scenarios from the remaining consistent configurations. Simplified FAR uses the following steps:

• [0. maja’ addition: make a mind map of a normative future and select several 'sectors' - critical uncertainties]
• 1. Form a Sector/Factor array in the usual fashion with as much documentation as time permits but, at the very least, explanations of what the Sectors (columns) mean.
• 2. Find a consistent configuration representing the current situation. (In any FAR, simplified or full-scale, if none of the configurations represents the present, there is a fundamental flaw in the Sector/Factor array.) Write a short description, in five or ten words, of what that condition represents, record it on a yellow sticker and place it the foot of a white board or flipchart. Do not describe it as status quo.
• 3. Find at least two consistent configurations that are believable, in the gestalt sense, as conditions for the end of the time horizon being used. In practice three are often easily found. Again, write short descriptions on a yellow sticker and put them on the board. Space them apart across the board from ‘worst’ to ‘best’; the yellow sticker should state clearly why they are bad, good or in between. In between does not necessarily mean half way between good and bad. It might, for instance, mean ‘pretty good’ or ‘rather worrying’.
• 4. If time allows, find a few more consistent, believable, configurations as end points on the time scale, and perhaps for some intermediate states, and record them as before.
• 5. It should now be easy to connect these six to eight stickers into credible time lines. If it is not, revisit steps 3 or 4.</blockquote>

From MORPHOLOGICAL FORECASTING– FIELD ANOMALY RELAXATION (FAR) By Geoffrey Coyle