Is David Bain guilty of murder? No. That’s what a jury of our peers said on far better information than we now have. Did he kill people? Possibly. Does it matter? No.
Our justice system is designed to acquit people when we cannot be sure of their guilt. In its design there is an implicit judgment that errors of wrongful conviction are worse than errors of incorrect acquittal. Let’s think about that judgment. If you acquit someone who committed an offence then the cost to society is only the cost of their recidivism. Given uncertainty about their guilt, the cost is P(guilty) x P(reoffend) x (damage from reoffending). If you wrongfully convicted someone then the cost is P(not guilty) x (damage from imprisonment).
Suppose P(guilt) > P(not guilty) by a little, so the person is not convicted but we think they probably did it. Given that P(reoffend) is probably ~30%, that means that the damage from re-offending would have to be a lot higher than the damage from imprisonment to make it right to imprison the person. Most murderers do not reoffend by committing another homicide so we can class that outcome as highly unlikely. Given the cost to a person of imprisonment it does not seem unreasonable to make the value judgment that we have implicitly made in the design of our justice system.
On the basis of this approach, the question of whether Bain killed his family is only a curiousity. It has little bearing on whether he should be in jail, since any normal person’s answer to that question is based on a ‘balance of probabilities’ approach. It is commendable that the jury took their responsibilities seriously and did not fall into the trap of going with their gut.