[derekp7]

The way I understand it, you can analyze it mathematically. Say one piece of circumstantial evidence puts the likelihood of guilt at 20%/ not guilty at 80%. Now another one has a 15/85 ratio. Take that 15% from 80%, and you now have a 68% not guilty / 32% guilty. Keep going until you have a good enough number, and you have your verdict. Now what that number should be? 90% guilty? 95%?

[dragonwriter]

> The way I understand it, you can analyze it mathematically.

Not meaningfully, for a number of reasons.

> Say one piece of circumstantial evidence puts the likelihood of guilt at 20%/ not guilty at 80%.

Problem with that is that it is rarely easily quantifiable.

> Now another one has a 15/85 ratio. Take that 15% from 80%

This, of course, assumes that the pieces of evidence are independent, which is not necessarily true.

Applying math to the issue isn't really helpful, because that's not the way jurors process evidence.

And if the system were changed to force them to work that way, it would just be a giant mess.

[sjg007]

In my interpretation it is all about what happened and how you spin the story. We see in books and movies where someone appears to be quite guilty but then in the twist they are not or vice versa. Now, life might be messier in some sense that such is less likely in theory. But even forensic evidence can be manipulated and the science of forensic evidence does not have the same rigorous scientific underpinnings as depicted on CSI. In a digital world, a lot of things can be manipulated, and in the world of espionage probably are.

[derekp7]

In that case, it is probably a good thing I've never been on a jury before (sat through jury duty once, never got called). I'll have to definitely keep your points in mind if I ever have to serve on one. Thanks.