returning to the DNA issue*
Dr. Wibble is pretty much echoing my thinking, about the birthday paradox (the recent Challenger anniversary is a stark reminder of how easily statistics can be misunderstood).
However, it goes a lot deeper than that. Starting with the problem that until such time as the UK has a credible life/death/in country register, there's no way to CLEAN the fucker.
So slowly, and surely, the database will grow.
So, fast forward to 2026, when the DNA database contains (say) 60,000,000 records. Including some dead. Some left the UK. Some moved.
(Remember, the DNA database is a *hash* of the genome, not the entire genome itself).
So a crime is committed, and DNA recovered. SOP runs it through the database, and - horror of horrors - 5 matches are returned. (The real horror is that it means the police now have to do some fucking police work). Of course the 4 spurious matches are a god-given gift to a lawyer who remembers that "reasonable doubt" is all that is needed for an acquittal. So the police have to try and eliminate the 4 spurious matches.
My explanation of the DNA database it that it's like a surname, house number, and area-level postcode. Which means there will be lots of SMITH-16-NW in there.
*presumably, the ANPR system has a link with the DVLA so scrapped cars are archived off. Or will it face the same degradation issues ?
All this, in the face of my assertion that *any* non trivial database of personal data can never be more than 95% accurate. A rule of thumb I devised after seeing thousands of man-hours in a commercial organisation devoted (over summer) to calling every customer to update their details, and seeing the accuracy improve (from 90%) to 95%, only to slip to 90% within 2 months. Databases being snapshots of a transient reality.
None of which is understood by the bottom-feeders in parliament, and their lickspittle civil servants who all dodged anything vaguely technical as "for the proles".
We'll all be well fucked if it becomes possible to make DNA from scratch.