I wholly agree with that interpretation. I just don’t see the difference between PPOs and PD in that sense. PD implies that we’re already OK with locking people up on some probability of future offending.

You could try to argue that we should lock people up before they commit a crime if our certainty passes a certain threshhold (ignoring my earlier points about the problems applying this to an individual), but western society has generally frowned on locking up groups of people because a subset of them will/have/are likely to commit a crime.

Its about the Rule of Law, which includes a principle of no retrospective criminal legislation.

The point Geddis makes is not about retrospective action, it’s about the difference between deciding at the time of sentencing and deciding after sentencing. I think he’s saying that indefinite detention, not linked at the time of sentencing to a crime that has been committed, is worse than PD.

why should a bunch of bureaucrats get to revisit a judges decision in sentencing?

First, it doesn’t need to be bureaucrats, it could quite easily be a judicial decision. Secondly, it is because they have better information near to time of release than the judge has at the time of sentencing. That is particularly the case for the type of criminals who these orders would apply to, since they would most likely have quite long custodial sentences in teh first instance. It seems possible that the risk of them reoffending could change over the course of the sentence. Indeed, the possibility of being subject to a PPO may encourage them to seek rehabilitation while imprisoned.