It's always bothered me in mediations when – in what I suspect is largely an attempt to bolster their own confidence – someone claims they've got a "75% chance of winning at trial". 

The trained statistician in me yearns to understand how they have come to such a conclusion – and what do they mean by it.  Presumably they're saying that if the same case was separately heard by 100 different judges then they would prevail 75 times.  Obviously that's never going to happen in practice.  In any event how does such analysis square with the other side's perspective – in a perfect market of equal knowledge and equal expertise then does this imply that the other side are running their case with only a 25% expectation of success?  

Of course, the law reports bear weighty evidence that for every winner there is also a loser - a long-run success rate for litigants as a whole of only 50%.  Mediators are well used to discovering that both sides are telling themselves they have a 75% chance of success; and part of our job is to address such anomalous joint perspectives by facilitating information exchange and encouraging fuller analysis and evaluation – in effect helping parties move towards a position of shared information and mutual understanding. 

And yet, even if mediators do achieve the nirvana of reducing parties' aggregated evaluations of their likelihood of success down to only 100%, we still have to consider the matters of how individuals respond to risk in conflict.

And to confuse the situation even more, some of us will be quite prepared to take on a 60% chance of winning, but would be scared off by a 40% chance of failure.  This isn't just a question of the diminishing marginal utility of money (the richer we get the less value we place on the next £1,000) or of what Richard Thaler1 called the endowment effect (whereby we place a higher value on things we already own) but also, more worryingly, on simply how the question is asked – as a general rule, we are so risk averse that we would rather, say, chose a medicine which saves 40% of patients rather than one which we are told will still see 60% die. 

Helping parties conduct an appropriate risk analysis is an essential part of the toolkit of the commercial mediator, and mediators are generally very good at what is called "reality testing" – helping them identify and think about the relative strengths and weaknesses of their and their opponent's situation. 

The size of the problem

Less formal attention is paid to the question of how parties respond to risk in these conflict situations.  As my former statistics professor, Dennis Lindley, one of the foremost experts in decision-making under uncertainty, wrote: "My hope is that today, somewhere in the world, there is a young person with sufficient skill and enthusiasm to be given at least five years to spend their time thinking about decision-making under conflict...Conflict is the most important mathematical and social problem of the present time2".

As to the scale of the problem, CEDR has previously reported3 that we estimate that conflict costs British business £33 billion a year once the usually unmeasured losses through wasted management time, distraction and damaged reputations are taken into account.  However, no doubt largely because of the failure to measure this cost – which never appears on any manager's financial report – conflict goes unaddressed, with a majority of managers reporting4 inaction as their organisation's main method of conflict resolution, and "avoidance and pretending it is not there" as a regular course of action. 

Inadequate training coupled with our natural conflict aversion clearly go some way to explaining this phenomenon of wasted cost and opportunity, but in this article I would like also to explore other aspects of our psyche which seem to apply even where the full extent of conflict risks are brought to our attention.  Key questions include how do we evaluate risks; what factors influence our appetite to take on such risks; and are we always the rational actors that we like to think we are?

Personalisation of risk

As a starting point, we should recognise that helping people to identify risk is not sufficient.  For all too often, even when we accept intellectually the theoretical possibility that something will go wrong, we comfort ourselves with the assurance that "surely it won't happen to me", that somehow we're immune from the fortunes that affect ordinary folk.  Reflecting what social scientists now term the Lake Wobegone Effect5, numerous studies have shown that across a wide range of characteristics and traits, a statistically improbable number of people tend to regard themselves, or their luck or skill, as being above average. 

Of course, you and I are too sophisticated to fall for such conceit.  But ask yourself this: "Am I a good driver?  Above average?"  Ask your colleagues for their rating of their driving ability too, and see if your experience bears out the results of repeated studies which reveal that around 80% of us claim to be above average drivers – a figure that remains remarkably constant even when the survey is taken amongst participants in driver re-education programmes or traffic school.  Thus, even people who have convictions, of the criminal variety for poor driving, nevertheless retain faith in their ability to be above average!

The same thing happens in conflict situations.  Particularly at the point of entry into conflict, when emotion or grievance is at its height, we can see ourselves as immune from risk, dazzled as we are by our own self-confident righteousness, and failing to see any daylight from other perspectives, often with the result that we get a nasty surprise later on.

As a mediator, I often have to spend a lot of time with parties re-visiting their risk analysis, encouraging them to reflect fully on the strengths of their opponent's position, on the weaknesses of their own, and on the vagaries of what former Defense Secretary Rumsfeld6 once termed "the known unknowns" and "the unknown unknowns".

Perception of risk

Of course, it is only if we can get all of the risks of conflict laid out on the table that we can have any chance of undertaking a proper evaluation.  However, unfortunately, we are not very good at assessing the level of risks even when they are brought to our attention.  I'm a very relaxed traveller, but as I write this article at some 36,000 feet above the Middle East en route to London even I can feel a heightened a sense of nervousness as I see my companion reading the front page news of yesterday's air disaster in the South Atlantic.  Obviously I'm not at any greater risk than the last time I flew, but the possibility is now at the forefront of my mind, and makes me worry just a little more.  And I also worry more than I should about dying in a fire, or of cancer, or a snakebite.  Why?   Because like most people I am not very good at calibrating different levels of risk.  Research7 on risk has shown that we all tend to over-estimate the risks of suffering from one of the higher profile or more spectacular causes of death, and we under-estimate our exposure to the more mundane, or at least less luridly publicised, causes (such as diabetes, stroke or asthma).

The old joke about doctors reveals how this plays out in conflict situations – we all celebrate our successes, and bury our failures. And thus it is our successes which are at the forefront of our mind – not only do we rejoice in past victories, but we pay more attention to formulating the winning arguments to support our own position than we do to listening and responding to the merits of the other side's position. 

Quantification of risk

And our difficulties in the perception of risk are only heightened once mathematical figures or patterns become involved.  There may be a stereotype that certain groups, such as lawyers, are not very good with numbers, but in fact our failure to compute probabilities, to assess risks or patterns of behaviour, is a far more universal failing.

Try yourself on the following questions:

1. If I toss a coin seven times, which of the following sequences of Heads and Tails is most likely to be observed?

                   (a). HHHHHHH         (b). HHHTTTT         (c). THHTHTT

2. A Claimant knows that, in order to succeed at trial, it will have to persuade the judge on 10 separate issues.  Counsel advises that they have an 85% chance of succeeding on each individual point.  Should the Claimant pursue the case to trial?

3. How many people must there be in a room before it becomes more likely than not that two of them share the same birthday?

                   (a). 23                             (b). 103                  (c). 183

Each of these questions challenges our ability to compute probabilities associated with multiple events.  We all know that on a single toss of a coin there is a 50:50 chance that it will come down heads.  But when looking at multiple events, our inclination is to look for patterns.   Some of us fall prey to the gambler's fallacy: "I've flipped heads with this coin five times consecutively, so the chance of tails coming out on the sixth flip is much greater than heads"; whilst others fall for the clustering illusion (also known as the patternicity effect) – a tendency to find meaningful patterns in meaningless noise.  When he said that: "Once is happenstance; twice is coincidence; three times is enemy action", Auric Goldfinger8, James Bond's wealthy nemesis, may have had particular concerns about people interfering with his plans, but he would have been wrong to read anything into the apparent patterns in answers (a) and (b) to my first question.  In fact, assuming a fair coin toss, each of the three sequences is equally likely.

My other two questions highlight the difficulties many of us have with composite probability assessments – of computing risks which are dependent on a series of events.  Intuitively, most of us would consider that the claimant in question 2 has a fairly good chance of success, but in fact when the 85% probabilities are multiplied together9, the combined result drops very quickly, down to below 20% in fact.

The birthday question is an application of the same principle, with the maths10 leading to the even more counter-intuitive result that only 23 randomly chosen individuals need be in a room before it be comes more likely than not that two will share the same birthday. 

So what does all of this say about our risk assessments in conflict situations?  Simply that we're not very good at gauging the odds.  We see patterns which don't exist, leading to unfounded predictions, and we under-estimate the implications of compound probabilities, leading to over-confidence about our chances of success in complex situations. 

Over reliance on limited data

We also often base decisions on very limited data.  Consider an informal experiment reported by Stuart Sutherland11 in which he compared his own perceptions of Australians following a visit to Earls Court in London with his findings from a subsequent visit to Sydney.  Based on his Earls Court experience, he described Australian men as being "loud and hearty" and "slightly uncouth", but his later visit to Sydney found its residents to be "extremely courteous and gentle".

Motivated scepticism

This tendency to place excessive reliance upon limited data is exacerbated when that data has been gathered through our personal experience, or where it confirms our existing prejudices or beliefs, possibly leading to the conclusion that the way individuals respond in conflict risk situations depends, at least to some extent, on their degree of familiarity with other similar situations. 

Taber and Lodge12 have proposed a model of "motivated scepticism", whereby when we are presented with a balanced set of pro and con arguments, we tend to place greater reliance upon those which support our prior views; and we place less reliance on – and even seek to disprove – opposing arguments and evidence.  Even more worryingly, this biasing also affects our future researches – when looking for fresh material to consider, we tend to seek out confirmatory evidence.  And the consequence of these biases is that, as time goes on, and more and more evidence is considered, our attitudes become even more polarised. 

This behaviour presents a challenge in dispute resolution, explaining as it does why so many disputants become more and more embedded in their position the longer the matter drags on.  

Group think

The role of the group also has an impact upon parties' evaluation of risk in conflict situations.  In a phenomenon known as "risky shift", people in groups make decisions about risk differently from when they alone, with the group generally likely to take riskier decisions. 

A number of reasons have been advanced to explain this behaviour – possibly the shared responsibility of a group decision eases the burden on the individual (i.e. a diffusion of responsibility); perhaps higher-risk takers have higher social status and/or are more persuasive in influencing the group decision; or perhaps it is simply that, as people pay more attention to a possible action through group discussion, they become more familiar and comfortable with it and hence perceive less risk.

Whatever the reason, the implications for conflict management are clear.  With group decision-making involved in most corporate disputes, there is a clear need for the questioning, or at least sceptical, voice.  But not every team is sufficiently robust that it can cope with the dissident, and it is not uncommon for a further level of internal conflict to break out if one individual is regarded as not being a team player, particularly in stressful situations where the team is seeking to bond together against a common concern. 

Clearly this is one area in which a mediator, as a trusted outsider, can play a key role. I would suggest that a more effective solution would be to have sufficient conflict literacy and self-awareness within a team such that it can tolerate someone – a designated cynic if you will – who can take a long, hard and dispassionate look at the merits of a case, to ensure not only that the chances of success are above average but also to determine the best strategy for persuading the other side of the results of such analysis. 

 

We continue to learn that the human animal responds to risk in ways that are even more complex than hitherto thought.  As a result, negotiation is now emerging as a multi-disciplinary field involving many specialisations, from lawyer to psychologist to coach to game theorist, all of which add to the mix now being employed in successful conflict management.  And in the future, maybe having a mathematician or behavioural economist to hand will also become common practice? 

Footnotes

1. R Thaler, Toward a positive theory of consumer choice. Journal of Economic Behavior and Organization, 1980

2. DV Lindley, Understanding Uncertainty, 2006, Wiley-Interscience

3. CEDR, 2006

4. Roffey Park, Management Agenda, 2004

5. "Lake Wobegone...where all the women are strong, all the men are good-looking, and all the children are above average", Garrison Keillor, A Prairie Home Companion

6. "There are known knowns. These are things we know that we know. There are known unknowns. That is to say, there are things that we know we don't know. But there are also unknown unknowns. There are things we don't know we don't know", US Defence Secretary Donald Rumsfeld, February 2002

7. Solvic, Fischhoff & Lichtenstein, 1979

8. Ian Fleming, Goldfinger, 1959

9. The formula to compute overall chance of success is: (0.85)10 or (85% * 85% *.... * 85% * 85%) =  19.7%

10. The formula for the chances of no duplicated birthdays is: 365/365 * 364/365 * 363/365 * 362/365 * 361/365 * 360/365 * etc.  This produces a result below 50% with the 23rd multiplication (i.e. ...* 343/365)

11. Stuart Sutherland, Irrationality, 1992

12. Charles S Taber & Milton Lodge, Motivated skepticism in the evaluation of political beliefs, 2002.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.