Originally Posted by Brandonub

An anecdote doesn't really say anything about relative risk; if I linked a story of a cyclist being killed by someone rear-ending them while they're sitting idly at a red light, that wouldn't indicate that we should obviously all be running red lights to avoid the rear-ending menace. Risks are correctly estimated with data, not with anecdotes.

Even with good data, that only tells us so much about how risky a behavior is for an individual. It could be conclusively demonstrated that rolling red lights is generally riskier than waiting for them, yet this wouldn't prove that a particularly attentive individual is at more risk rolling a red than sitting at it. This is why individuals should be making context based judgment calls rather than there being laws that attempt to enforce a one size fits all solution.