The Pacific Institute has developed a highly accurate algorithm to predict violence by individuals and collectivities, the instructions for which are presented below. We also utilize the highly robust Delphi Method in assessing predictive conclusions and in retracing the decision process of the most accurate predictors. The Delphi Method combines the best features of individual versus group problem solving, and comes in 3 steps:  ​

​

• First, a group of experts render independent of each other submit in writing their predictions regarding a specific issue.
• Second, the experts exchange predictions and their basis of reasoning for all their predictions.
• Third, the experts submit final predictions without benefit of consultation with the other experts.

​Using the below algorithm and Delphi method, the onset and characteristics of both the Persian Gulf War and the 2003 invasion of Iraq were accurately prognosticated by Pacific Institute members. The writer in several hundred clinical and forensic cases for individual violence has utilized the algorithm alone. The results produced over 90% combined true positives (correctly predicting violence) and true negatives (correctly predicting nonviolence).  The plain-English version of the algorithm follows:

​

1. HAVE YOU SPECIFIED A DISTINCT TIME SPAN FOR YOUR PREDICTION?  If not, then do not proffer a prediction. If yes (e. g., violence within one year; imminent violence within 90 days), continue with the steps;
2. HAS (UNINTENTIONAL) DISTORTION OR (DELIBERATE) DECEPTION BEEN TAKEN INTO ACCOUNT?  If not, predictions cannot be proffered. Collected information should not be taken at face value.  If yes, first specify the kind of distortion uncovered (e. g., dull normal intelligence; cumulative extreme stress). Utilize standardized testing or structured observations/interviewing to ascertain the presence or absence of deception displayed by the predictee, target and others providing information.
3. DOES THE POSSIBLE PERPETRATOR HAVE A VIOLENCE HISTORY?    If not, the only prediction that should be made is that no violence will occur. If yes, list by date and circumstances.  Look for patterns that suggest a target typology and the usual context of violent interaction.  Acts of violence by non-leaders in a highly cohesive, suspicious, punitive, lethal organization (e.g. terrorist cell; dictatorship) are counted for the violence history of the leader himself or herself. Followers in these types of organizations identify with the leader, or at least comply with the desires of the leader, or they do not survive. Go on to next step;
4. HAVE YOU SPECIFIED TRIGGERING STIMULI, INTENSE IN IMPACT AND SHORT-TERM IN DURATION, WHICH MAY SET VIOLENCE INTO MOTION?  If not, refuse to predict unless the predictee displays violence that does not appear associated with stress and is a function of opportunity (e.g., psychopath for individual violence; note that in organizations, obedience to authority may act as a trigger).  If yes, list them and continue; 
5. SPECIFY OPPORTUNITY VARIABLES WHICH MAKE VIOLENCE POSSIBLE, OR WHICH EXPAND ITS OPPORTUNITY.  If not, refuse to predict. Weapons, transportation, communication, release from a restricted area and target availability are key opportunity variables for both individual and collective violence. List opportunity variables for the predictee and continue;
6. SPECIFY INHIBITORY VARIABLES, WHICH MAY LOWER THE CHANCES VIOLENCE, WILL OCCUR.  If none, refuse to predict.  Inhibitory variables range from certainty that negative sanctions will follow violent behavior to having bonded/familial relationships with the target.  If inhibitory variables have been considered, proceed to conclusions.

_______________________________________________________________________________________________________________
 

BASED ON THE ABOVE, FORMULATE THE FOLLOWING CONCLUSIONS

(Place conclusions in binary or in a quantitative format if possible for easy verification)

1. ARE CONCLUSIONS PROFFERED WITH A REASONABLE DEGREE OF [DISCIPLINE] CERTAINTY?  If not, then do not present conclusions unless referral source will accept a lesser standard of truth such as a reasonable degree of probability. The discipline refers the area of knowledge on which the prediction is focused (e. g., medical certainty, psychological certainty, military scientific certainty);​
 

2. HOW DID UNINTENTIONAL DISTORTION AFFECT RESULTS, IF AT ALL? If they did, either take results into account or do not present conclusions.​
 

3. HOW DID ANY DECEPTION REVEAL ITSELF IN ONE OF FIVE PRIMARY STYLES—HONESTY, FAKING GOOD, FAKING BAD, INVALIDATION AND COMBINATION STYLES? If deception is present, take the results into account or do not offer conclusions, as the risk of error may be unacceptably high.  A cross-validated deceptive style is supported by at least 6 indicators, which corresponds to a low probability of chance occurrence;

4.  WHAT IS THE PROBABILITY VIOLENCE WILL OCCUR WITHIN SPECIFIED TIME RANGE? State whether probability is negligible, minimal, mild, moderate, considerable, or almost certainly. Specify the probability range denoted by each level of probability if supported by statistical data or actuarial measures of violence);

5. WHICH ARE THE KEY FACTORS IN COMBINATION THAT FORMED YOUR CONCLUSIONS?  Some events in combination may weigh more than others and should be specified (e. g., written credible threats to kill combined with victim availability and a multiple history of violence; for collectivities, massing of troops on border of target-country combined with a stated intention to attack the enemy within a certain time frame).

6. WHO OR WHAT ARE THE MOST LIKELY TARGETS?  The specific likely target of violence (e.g., spouse, work supervisor) should be identified if known.  Legal reporting requirements should be checked for compliance.   Classes of targets (e.g., coworkers, enemy soldiers, ships) should be specified when appropriate;

7. WHAT IS THE ANTICIPATED USE OF WEAPONS?  Type of weapon(s) based on history, proficiencies, availability and current training should be noted. The effectiveness of weapon use for an individual or in tandem with a weapons platoon, as an example of institutional violence, should be considered. Failure to identify weapons and armaments of both the predictee and the likely targets renders the entire prediction useless;

8. WHAT IS THE ANTICIPATED USE OF OTHER OPPORTUNITY VARIABLES?  Likely communication methods and instrumentation, transport, target availability, and others should be noted and factored into planning;

9.  WHAT IS THE ANTICIPATED HARM OF PREDICTED VIOLENCE? This ranges from no harm, for negligible risk of violence, to threats, minor injury, moderate injury, death, and multiple deaths. Anticipated total harm includes all that is administered by subordinates of a leader in a violent organization;

10.  WHAT IS YOUR CERTAINTY LEVEL FOR THIS PREDICTION? State the confidence you have in your prediction ranging from negligible, minimal, mild, moderate, and near certain.

11.  WHAT IS YOUR FEEDBACK MECHANISM TO REASSESS CONCLUSIONS? Recommend the type of evaluation you or independent evaluators to reassess the accuracy of your prediction, even if repeated measures are utilized;

12.  POSSIBLE CHANGES IN PERPETRATOR, VICTIM OR CONTEXT, OR OTHER FACTORS WHICH WOULD CHANGE YOUR PREDICTION. For example, the imprisoned high-risk predictee may be transferred to solitary confinement, which would obviously lower the chances violence to others would occur.

Two sets consisting of several dozen predictions for the US in 2012-2013 are now presented. Generally, the first set of predictions for 2012 if verified describes a continuation of serious economic and other domestic problems in this country but no catastrophic changes such as an economic crash, end-time scenarios, or general war.  President Obama is predicted to win re-election. These predictions are described in the .PDF below entitled " Predictions That Americans - And Congress Should Care About."  A second set in the .PDF entitled "List of 62 Predictions" forecasts that in 2012-2013, possibly starting in the next several months. It is more likely than not the US/Israel will attack Iran along with predicted characteristics of the war. In a PDF to be developed, a third set of predictions span 2013 to the 2020s during which time it is predicted that the US will experience economic hardships, including a crash, trade wars, currency wars, more erosion of the middle class, continual conflict in Iran (the war having started in 2012), further loss of civil liberties, and increasing anger, stress and disappointment at home.  A final set of long-term predictions for the 2020s will be developed in another .PDF, during which decade a Great Transformation for American society is predicted.

​

PLEASE GO TO BUTTON ON LIST OF 62 PREDICTIONS FOR MORE INFORMATION

 

​​​PDF “Predictions That Americans—And Congress—Should Care About”   <---- CLICK TO VIEW .PDF