We are used to, and think we understand, percentage probability, says Jonathan Haydn-Williams – but remember that percentage estimates of success do not make it certain and are less reliable the earlier they are made.
We live in a world measured by statistics and expressed, in particular, by percentages.
Sports pundits talk of ‘playing the percentages’. Politics gorges on percentages. In economics, percentages enable us to get some grasp of figures so large as to be otherwise incomprehensible.
Such is our predilection for percentages that we like to use them, not just as a common currency for expressing statistical data of past events, but also as a means of expressing the probability of future events. Hence, the probability of a team winning a football game or of rain disrupting a cricket match is expressed in percentage terms. It is hardly surprising, therefore, that lawyers are wont to, and clients expect them to, express the probability of winning a piece of litigation as a percentage chance.
Parties to disputes and/or their insurers or funders need to have some means of making decisions about whether to pursue litigation, whether to settle a claim, how much to settle for etc. A percentage chance of success, even hedged with the usual – and necessary – caveats, may feel more precise than mere words (eg ‘reasonable case’, ‘good case’) and gives them a familiar handle on the uncertainty they are facing. However, litigation is inherently uncertain and parties tend to attach more significance to a percentage probability than is perhaps prudent.
Maybe for psychological reasons beyond the scope of this article, a client who is told that there is a 70 per cent chance of success may not pay enough attention to the ‘flipside’ of a 30 per cent chance of losing. The ‘70 per cent case’ label can take hold. What the lawyer’s opinion means, of course, is that, based on the knowledge at the time and on the lawyer’s experience, and very roughly speaking, if the same case were run 10 times before 10 different judges, one might expect the client to win in seven of them and to lose in three of them.
In those three lost cases, the probability of a future event happening would have turned into the certainty of an event that has occurred. A not uncommon reaction from the losing party may be surprise or shock: ‘How could this happen when my lawyer told me I had a 70% chance of success?’
The factors which can affect the outcome of litigation are many: the availability of documentation and what it states, the memory of witnesses and how they perform when giving evidence in court, the fallibility of expert evidence. The earlier the stage at which an opinion is given, the greater are the unknowns, such as what documents the opponent may produce and what its witnesses will say. I do not recall ever giving a percentage estimate of winning a case of more than 75 per cent, because there is no ‘crystal ball’ in which future unknowns are revealed.
In England and Wales, civil claims are tried by a single judge, who decides on both law and facts, usually without a jury. The judge’s decision is difficult to appeal, in that there is no absolute right of appeal: instead a losing party has to seek leave to do so, which is often refused. Judges are only human and two judges’ analysis of the same case may differ markedly. After all, if the outcome of a case were obvious, it is likely that it would have been resolved by agreement before it reached the stage of trial. In some countries, a panel of judges decides a case, but in England and Wales that seldom happens below appeal level.
A number of years ago, I was involved in a case in which a barrister gave my then firm’s client an opinion that its defence had an 80 per cent chance of success. To our surprise, the opponent issued an application for summary judgment. To our even greater surprise, the junior judge hearing the application stated that, in spite of superficial complexity, he thought it a simple case to which our client did not have a realistic defence: judgment for the claimant. Fortunately, in that instance there was a right of appeal to a more senior judge, who disagreed with the decision and overturned the judgment. Rather than subject themselves to the further uncertainties of litigation, the parties reached a compromise settlement soon afterwards.
The laws or principles of probability have their own inherent difficulties of application, as explored in Tom Stoppard’s play Rosencrantz and Guildenstern are Dead . Rosencrantz begins tossing a coin. It comes up ‘heads’ time after time. Guildenstern takes an interest, starts supplying coins from his purse and calling ‘tails’. But ‘heads’ keep coming up and Rosencrantz keeps winning the coins. After 78 ‘heads’ in a row, Guildenstein muses that: ‘A weaker man might be moved to re-examine his faith, if nothing else at least in the law of probability.’
There is a paradox. Every time the coin is tossed, the chance of ‘heads’ is 50 per cent and that of ‘tails’ also 50 per cet. The outcome of one flip of a coin is not affected by the results of the previous flips. But the chance of 78 consecutive ‘heads’ is very small indeed. Guildenstern sticks with ‘tails’, playing the percentage game, but losing. He may think he has an ever-improving chance of winning, but it always remains 50 per cent on each toss.
If one suggested to parties to litigation that they should save all future legal costs and toss a coin instead, they would be appalled. Yet the same parties might well be willing to pursue expensive litigation if told they had a 20 per cent more chance of succeeding in their claim or defence than if they just tossed a coin (ie 70 per cent). Some might litigate on a 60 per cent – and some even on a 50 per cent – chance of success. But whilst the 50 per cent odds on the toss of a coin are fixed under the laws of probability (in spite of Guildenstern’s losing streak), the odds on litigation success have no such basis and are dependent on the individual lawyer’s judgment.
Thus far, the spotlight has only fallen on probability, but decisions are (or should be) made on the basis of risk assessment, which has two components: the probability of an event occurring and the severity of the event should it happen. A higher probability of a non-serious event occurring may carry less risk than a lower probability of a very damaging event.
For a claimant, the effect of losing and not receiving the remedy sought will depend on that claimant’s needs. Losing out on a damages award that would be nice to have, but not vital for the ‘bottom line’, is very different from failing to obtain an injunction to prevent a competitor acting in a way that is destroying your business. A defendant has a ‘mirror’ position, except that the defendant did not have a choice of whether to be part of the proceedings (though is not powerless as regards trying to bring them to an end, eg by making an offer of settlement).
Losing a case in England and Wales usually results in having to pay a significant proportion of the opponent’s legal costs (how much has become more uncertain in recent years), as well as one’s own. So, the likely size of costs bills is a significant factor is determining litigation risk in England and Wales. If a potential loss is growing, an unchanged probability gives rise to an increased risk factor: a point not to be ignored by litigants as the legal costs get larger.
This guide is for general information and interest only and should not be relied upon as providing specific legal advice. If you require any further information about the issues raised in this article please contact the author.