I mean, it’s the UK. Presumably there is video. Plus witnesses. Plus GPS. Plus he left the country immediately after. Plus he is known hired muscle. Certainly enough to charge someone.
If there is video sure, but the rest is literally “I saw a guy who looked similar in a park” and “the GPS said this guy was in the park a few times before the event occurred”
It is the UK though, so idk, but even in the US people get convicted on less. Just seems open to reasonable doubt
When you start stacking up evidence that by itself is inconclusive, you find that the overall probability starts to get very tight.
For instance:
suspect was in this neighborhood at this time
suspect was wearing a blue jacket
suspect had short blonde hair
suspect was an acquaintance of the victim
Each of those by itself means next to nothing; you can't go around arresting anybody with blonde hair because there are millions of people with hair like that. But if you find somebody who fits all four, they are almost certainly your guy. The chance of there being two acquaintances of the victim with blonde hair and a blue jacket in the neighborhood at the time of the murder is much smaller than the chance of a false positive on any one of those individual conditions.
With enough circumstantial evidence, you can blow right past "reasonable doubt".
This line of thinking is rather dangerous when determining whether to lock someone up for life. Especially in the absence of direct evidence. The math is often tricky, given that we already know some highly unlikely things have in fact occurred (a murder, for one).
There is a huge difference between “here is GPS data, let’s see if somebody remembers seeing that guy” and “somebody reported seeing that guy, let’s see if his GPS data confirms that.”
In this case, it was the latter. An unusual sighting was corroborated after the fact by hard data.
Less circumstantial and less suggestive than “here’s some hard data, let’s ask people if they can remember something.”
Somebody better at stastitcs can probably prove this with some Bayesian formula.
