One of the main problems with predictions of dangerousness is that the indicators often employed are both static and dynamic.  Static factors are those such as age, offence history, childhood and family factors [Adair].  These factors are static because they are more or less permanent and cannot be changed.  Dynamic factors include such things as attitudes, progress in treatment, and relationships.  These factors are dynamic because they can be changed, or are susceptible to change depending on the offender's situation.  When an assessment is made, static factors may play an influential role in determining the offender’s status, though dynamic factors might counterbalance any adverse effects from the static factors.  An example of this would be an offender who has an extensive criminal history and a poor upbringing in a family replete with inter-generational crime, who now has a stable relationship with a caring partner and the promise of employment on release.  Any heavy weighting afforded the static factors may unduly influence the decision to keep this individual incarcerated even though the dynamic factors would suggest he is more likely to succeed on release.

In the past, predictions of dangerousness relied heavily on the subjective interpretations of individual clinicians that proved highly inaccurate owing to the lack of guidelines [Ward], and the differences between clinicians on theories behind the nature and causes of dangerous behaviour [Montgomery].  Also, the theoretical orientation of the practitioners might cause them to place a greater emphasis on some factors over others simply owing to their professional orientation. For example, a Freudian psychologist might consider the relationship with the mother as a paramount consideration, while a cognitive psychologist might consider the offender's thoughts and perceptions as a more important factor.