Statistics are a lie (or at least a "truth" derived dicto simplicitor) we like to believe. After all, most people would say that 99% is a pretty high statistic, however, that is the amount of DNA we share with a banana – AKA 50% of our genes. Conversely 1% seems insignificant... well that again is how different you are to a banana.

Statistics therefore shouldn't be used as a measurement of probability because even though a 99% chance of not dying from a disease, for example, means that there is a 1% chance of 100% dying.

Life is a gamble – don't pay too much attention to statistics. Understand the placebo effect instead.

----------

Edit: 26th Mar 2021

Be aware of sweeping generalisations. Just because something apparently "always" works does not mean that it always will do. The Sun has always risen so it might seem strange to think it might not rise tomorrow, but the "Problem of Induction" (https://bblais.github.io/posts/2014/Sep/22/will-the-sun-rise-tomorrow/) is mutually inclusive with statistical lies. 

Maybe something "always" works, or "always" applies, but that does not mean it "always" will. Paradoxically, there are "always" exceptions.  Seems like a dichtomoy, but it really isn't. 

Essentially, expect the usual but prepare for the unexpected.  

----------

Edit: 29th Mar 2021

Nothing is impossible; just improbable.

Common rebuttal are thus: "if that's so, then why do(n't)... ?"

It's simply because the variables are favourable. 

Theoretically, pigs can fly, however, the genetic mutation necessary for flight is very improbable to be switched on.

As Murphey's Law (which are essential deviations from perfection: the province of the Second Law and are based on probabilities) essentially points out - whatever is probable will eventually occur. Unfortunately, Humans are shortsighted and therefore lack the imagination to understand everything that is probable.

----------

Edit: 3rd Apr 2021

Next step: confirmation bias

----------

Edit: 20th Apr 2021

"Science isn't about truth and falsity, it's about reducing uncertainty." - Brian Nosek

Nothing can be proved at this moment as we lack the equipment to completely isolate a single variable. All causations are just correlations based upon 'negligible factors.

I.e. 99.999... is close enough to 100

argumentum ad ignorantiam 

Argument from ignorance criticises that "everything is true until considered false", but if we do not take a stance, where does that leave extreme skepticism? When nothing can be fully proven at the moment (unless it is theoretics), how do we know anything unless we hold something to be true until proven otherwise?

----------

Edit: 25th April 2021

Data science works perfectly in theory.

Too bad that this is reality...

In practicality, common sense, heuristics, and logic are the most efficacious. 

Reference: https://harvardpolitics.com/stop-using-data-alone-to-make-decisions/