If you live with a gun owner, you're 2X more likely to be murdered

[Model35mech]

6. Yes, it's based on x times as high as the current rate of occurence. What bugs me is

that many times it's reported without saying anything about what the current rate of occurrence is.

If the current rate, x is small, say 0.0000001 then 2x is going to be, 0.0000002, which, honestly remains relatively small.

But it can still be, and often is, hyperbolized for rhetorical effect as a DOUBLING rate of risk with no reference to the actual magnitude of original risk being doubled.

When someone says the risk is doubled or tripled, or has a 200% or 300% increase, the question begged is what original magnitude or percentage is being talked about.

It's really not a trivial thing. Finding a high slope in the increase of risk can truly be a very important thing, it draws attention to a high rate of change over time. As an epidemiological statistic that is often an early warning signal in the data.

But examination of that statistic requires interpretation. Does it change anything about detecting and/or mitigating the risk? When the real number of events remains small, the events may still make detection unlikely (despite expensive efforts at detection), and possibilities for outcomes of mitigation rather minimal (although they are likely to be more expensive than the cost of detection).