What Is Scientific Case Consulting?

Scientific case consulting (SCC) uses the scientific method and research to

    • minimize the chances of an unfavourable verdict and
    • maximize the chances of a favourable verdict.

Examples include scientific trial consulting and scientific jury selection (SJS). For instance, the latter uses the scientific method and research to select the most favourable jurors. Harold Fahringer argues that 85 percent of cases litigated are won or lost in the jury selection phase. See “Mirror, Mirror on the Wall …: Body Language, Intuition, and the Art of Jury Selection” (1993–1994), 17 Am. J. Trial Advoc. at 197.

A more accurate estimate may be closer to 100 percent, since jury selection determines the jury, which determines the verdict. 

Qualifications of a Case Consultant

A case consultant should have:

    • A degree in law (i.e., proven understanding of the law and legal process)
    • Experience and success conducting or advising the conduct of trials or motions (i.e., proven authority to consult)
    • A science degree or equivalent (i.e., proven ability to do and review quantitative research)
    • A humanities degree or equivalent (i.e., proven ability to seriously engage with topics such as history, philosophy, religion and politics)
    • General knowledge of human beings (e.g., from history, philosophy, religion, social science, anthropology and psychology)
    • General knowledge of the jury pool (e.g., from demographics, social psychology and social epistemology)
    • Specific knowledge of the judge (e.g., from studying the judge’s attitude and beliefs)
    • Specific knowledge of the jury (e.g., from studying the juror’s attitude and beliefs)
Why Hire a Case Consultant?

As a case consultant, we can:

    • review your case and tell you if it’s strong or weak from a scientific perspective
    • select a jury who is more likely to believe your case
    • tailor your case to the judge or jury
    • assess the predictability of the case based on certain information and methods (see the following sections)
    • suggest the most appropriate procedure to follow
    • study and consult about the judge (e.g., Justice Engels is a Marxist and legal realist who will find a way to judge in favour of the proletariat against the bourgeoisie)
    • study and consult about the jury (e.g., the people of Rapesalot County rarely convict people accused of rape based on the testimony of a single witness)

Methods of Predicting Decisions and Assessing the Scientific Merit of a Case

Using Wigmore Charts to Assess the Scientific Merit of a Case

The Wigmore chart is a device proposed by one of the leading scholars of evidence in the common law, John Henry Wigmore (d. 1943). Unfortunately, this scientific device with a lot of potential has almost never seen a day in court!

The Wigmore chart basically schematizes a legal argument a lot like a Toulmin model of argument, as the Table below shows.

Wigmore chart

The Wigmore chart can be combined with Bayes’ theorem to yield a systematic method, which consists of asking what is the probability group A is wrong given that group B contradicts their testimony?

Bayes’ theorem

Related image

For instance, take a set of 10 claims by 10 persons. Presuming one of them is correct, the prior probability any claim is correct is 1/10.

Evidence may reveal more information about the posterior probability. For instance, if all 10 claims are that the earth is flat, the prior probability the claim is correct (presuming one of them is) is 1. However, the posterior probability is 0.


J Wigmore, “The problem of proof” (1913) 8:2 Illinois L. Rev. 77–103.

J Wigmore, The Science of Proof: As Given by Logic, Psychology and General Experience and Illustrated in Judicial Trials, 3rd ed. (Boston: Little, Brown, 1937).

The Need to Predict Decisions

Most clients will be not satisfied with only an assessment of whether their case is strong or weak from a scientific point of view. Most clients want to know if they’ll win the case.

What Makes It Difficult

The judge or jury decides who wins in court. Unfortunately, a lot of judges and juries are hard to predict because they don’t follow a known or systematic thought process.

Several studies in the US show that legal experts tend to predict US Supreme Court decisions with little more accuracy than a coin toss or worse. For instance, in 2002 legal experts in the US only correctly predicted about 60 percent of the US Supreme Court’s decisions.

This could be because lawyers’ methods of predicting decisions are inaccurate or arbitrary, or because the decisions are arbitrary.

Three Methods to Predict Judges' Decisions

Three methods of predicting judges’ decisions are as follows:

1. Statistical Mechanics

Statistical mechanics seeks a pattern in past judgments. Where there is a pattern, this method will be effective.

For instance, in 2012 physicists Roger Guimera and Marta Sales-Pardo Guimera came up with two successful models to predict the votes of individual US Supreme Court judges based on their voting record relative to other judges’ votes, as the Table below shows:

Predictions of US Supreme Court judges’ votes

Judges’ Opinions Method of Prediction Rate of Accuracy
Split (5-4) 50/50 77%
Majority (>5) Majority rule 83%

2. Computer Algorithms

Another method is to simulate the decision process with a computer algorithm. This method will be effective where an algorithm can approximate judgment.

For instance, in 2017 a competition was held between CaseCruncher Alpha, a program that predicts legal decisions, and 100 lawyers from London’s top law firms. The lawyers and CaseCruncher were given facts of hundreds of payment protection insurance (PPI) mis-selling cases and asked to predict if the Financial Ombudsman would allow a claim.

775 predictions were submitted and CaseCruncher won with an accuracy rate of 86.6%, compared to 66.3% for the lawyers.

3. Method of Judgment

This method consists of following the judge’s method. Needless to say, it will only be effective if the judge follows a method. This method is the most effective if the judge is methodical and the least effective otherwise.

Judges fall into one of four categories:

Four types of judges

Methodology Formalist Realist
Positivist Predictable rulings and findings of fact. Unpredictable rulings and predictable findings of fact.
Constructivist Predictable rulings and unpredictable findings of fact. Unpredictable rulings and findings of fact.

For instance, we could only predict 12 out of 23 decisions (52 percent) by Justice McLachlin, who is a realist-constructivist. By contrast, we could predict 19 out of 20 decisions (95 percent) by Justice Paciocco, who is a formalist-positivist.

Limitations of these Methods

The two major limitations of these methods, besides those already mentioned, are (1) they are of limited application to arbitrary decisions and (2) they only apply to judges. This is because unlike judges, juries

    • are not always the same persons,
    • don’t give reasons for their decisions and
    • little information is available about them.

In other words, the jury box works like a black box. Woe to the accused who has not a Henry Fonda on its jury!

Henry Fonda gets the jury to think outside the ‘box’

However, most trials are judge-only, and most judges insist their judgments aren’t arbitrary, which can offset some of these limitations.

Methods to Predict Juries' Decisions

The standard method of trial is by a judge only, which is called a bench trial. However, a trial by jury (aka a jury trial) may be requested in the following situations:

    • a person accused of a crime can request a trial by jury if it faces a sentence of more than five years in prison pursuant to s. 11(f) of the Charter
    • parties to a civil litigation can request a trial by jury pursuant to r. 47.01 of the Rules of Civil Procedure unless it falls under an exception in r. 108(2) of the Courts of Justice Act

A bench or jury trial can be more or less advantageous to either party, as the Table below shows.

Advantages and disadvantages of bench and jury trials 

Method of trial / Party Bench trial Jury trial


  • the plaintiff only has to convince the judge instead of 5 of 6 jurors in a civil trial or 12 of 12 jurors in a criminal trial
  • information about the judge tends to be available (“Better the judge you know than the jury you don’t”)


  • the judge may not favour the plaintiff


  • the jury may favour plaintiff (e.g., if public opinion supports it)


  • predicting jurors is a 6-body problem in a civil jury trial and a 12-body problem in a criminal jury trial, which is more complicated than a 1-body problem in a bench trial
  • little information about jurors tends to be available, so it’s best to ‘play it safe’


  • information about the judge tends to be available (“Better the judge you know than the jury you don’t”)


  • the judge may not favour the defendant


  • the defendant will be acquitted if the plaintiff fails to convince (or the defendant succeeds at convincing) at least 1 of 12 jurors in a criminal trial or at least 2 of 6 jurors in a civil trial


  • the jury may not favour the plaintiff (e.g., if public opinion is against it)

A bench or jury trial can also make the trial more or less predictable. For instance, the standard of proof in civil law (a “balance of probabilities” or “more likely than not”) means the evidence must establish that the plaintiff is more than 50 percent likely to be telling the truth. Blackstone’s ratio, on the other hand, means the evidence in a criminal trial must establish that the prosecution is more than 90 percent likely to be telling the truth.

Before the trial, the judge’s or juror’s prior probability (or base rate) of acquitting the defendant should be equal to the standard of proof. This means that before the trial a civil judge or juror should be 50 percent likely to acquit and a criminal judge or juror should be 90 percent likely to acquit. The prior probability of acquittal based on the method of trial may therefore be calculated as follows:

Prior probability of acquittal based on method of trial

Method of trial Prior probability of acquittal (%) Explanation
Bench trial – criminal 90 P(acquittal) = standard of proof
Jury trial – criminal 99.9 P(acquittal) = 1-.1^12 = complement of the chance all jurors will convict
Bench trial – civil 50 P(acquittal) = standard of proof
Jury trial – civil 89 P(acquittal) = 1-7/2^8 = the chance at least 5 of 6 jurors will acquit, given the equiprobability of all outcomes

The Table above shows that a priori and de jure (i.e., in theory and according to law) a jury trial is more favourable to the defence. The difficult part is to establish the conditional (or posterior) probability a judge or juror will acquit on any given evidence (i.e., the empirical or de facto chance of acquittal). 

We have already discussed some empirical methods for predicting judges’ decisions. Predicting the decision of the jury devolves to the question what are the odds the plaintiff’s argument will convince at least 5 out of 6 jurors in a civil trial or 12 out of 12 jurors in a criminal trial. This depends on information available about

    • the relevant population (i.e., the jury roll according to the Juries Act) and
    • the relevant sample or individuals (i.e., how the jury behaves as a group or each juror behaves as an individual)

Without specific information about the sample or individuals, general information about the population (i.e., statistics) may be relied upon. The relevant statistics must correlate personal attributes with dispositions. Correlation should not be confused with causation. For instance, while most Chinese people are atheists, this is not because they are Chinese. Conversely, most Black people will probably sympathize with a Black victim of racial discrimination because they are Black.

As a general rule “like likes like,” meaning that people who belong to the same group or share certain attributes tend to share certain interests. For instance, a pot-smoking hippie is not likely to judge another pot-smoking hippie harshly for possession or trafficking of marijuana.

Demographic and psychological studies can establish some significant statistics and correlations. However, the rule is “you can’t get more than you put in,” which means that if little information is available about the jury, then little can be predicted about it. Accuracy of prediction depends on the information available, and statistics and correlations are a weaker substitute for facts and causes. 

The lawyers involved in the trial receive a list of potential jurors complete with the information they provided to the Sheriff in the Juror Questionnaire that was sent to them (read this article for more details). The Table below shows each question the Questionnaire asks and the critical information it can disclose about the juror and its interests in addition to their name and appearance in court.

Questions in the Juror Questionnaire and the critical information they may disclose about the juror and its interests

Juror Questionnaire Question Critical information it discloses about the juror and its interests
Appearance in court Reveals a person’s race and ethnicity. Along with other information, this could reveal other relevant information, such as their religious or cultural values.
Name: A person’s name, along with other information, can reveal a lot about their history and values.
1. Are you a Canadian citizen? Narrows down relevant statistics.
2. Are you 18 years of age or older? Narrows down relevant statistics.
3.Have you attended court for jury selection in response to a summons in this or the two preceding years? N/A
4. Does your work exempt you from jury service? Narrows down relevant statistics.
5. Do you have any physical or mental disability that would prevent you from serving as a juror? Narrows down relevant statistics.
6. Have you been convicted of any criminal offence that can be prosecuted by way of an indictment for which you have not been granted a record suspension or pardon? Narrows down relevant statistics.
7. Provide your gender and date of birth: Narrows down relevant statistics. Older people have lived longer and there are biological differences between women and men that could be relevant.  
8. Provide current or most recent occupation: Narrows down relevant statistics. Occupation may speak to the juror’s social class and interests, as well as their knowledge and ability to reason scientifically.
9. a) Do you speak, read and understand English?
b) Do you speak, read and understand French?
This can be extremely useful to challenge a judge or an unlimited numbers of jurors for cause, as we note in our article on “Bilingual Legal Tactics” (see, e.g, The Rights of French-Speaking Individuals in the Ontario Justice System).
10. Business telephone and residential telephone: This can reveal the juror’s address. Their choice of residency, their relationship to their neighbours, etc. given other information about them, can reveal a lot about them.

Method to Study Decisionmakers

As the previous discussion shows, we use a scientific approach to study and predict judges and jurors based on their behaviour and methods. This includes developing and testing a theory of the judge or juror.

Three Possible Scenarios for Prediction

Generally speaking, three situations are possible:

Three possible scenarios

Situation Prediction
The judge or jury is rational (which means they are predictable) The client will win with a scientifically strong case.
The judge or jury is irrational, but still predictable (there is a ‘method to their madness’) Some cases, like some surgeries, are pretty routine and have a known success rate. In these cases, the normal protocol should be followed. Other cases, like some different kinds of surgery, are more peculiar and their outcome is harder to predict. We can tell if a client’s case falls in the first or second category and if it has a known success rate. If not, we can study the judge or jury and provide our assessment.
The judge or jury is neither rational nor predictable The best the client can do in a situation like this is ‘crapshoot,’ meaning they must try to argue, within the allotted time,  whatever the judge or jury might believe and hope to God that something sticks.

Sample Predictions

The following Table lists some of our predictions of ongoing causes célèbres.

Our predictions of ongoing causes célèbres

Date of Prediction Case Anticipated decision Reasons
January 16, 2017 The first election of Donald Trump He will win. Our theory of Trump, based on a lot of evidence, is that he is not an ideologue but a power-monger. He can afford to purchase information to say and do what it takes to gain public support. We predicted he would win when he made his infamous comments about Mexican immigrants and building a wall based on the theory this was a ‘bought’ statement the politically correct establishment would not express. 
May 1, 2019 The second run for presidency of Donald Trump He will win again. Trump is the only one making politically incorrect statements and taking politically incorrect initiatives to gain critical support. As long as this is so, his chances of success are probably very high, otherwise he wouldn’t be wasting his time or money in politics.
Take Our Surveys!

About Our Surveys

Our surveys are anonymous and don’t ask for any personal information. We use them to compare with other research.

Views on Trusting Strangers

Views on Science

Views on Legal Interpretation

Views on First Impressions