Fast And Slow
Jun. 6th, 2013 10:32 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
peterbirksI am very much enjoying Kahneman's Thinking, Fast And Slow which neatly runs through all of the "lazy" thinking methods that humans use to avoid doing any proper analysis.
I'm currently at hindsight bias, which I think is one of the more ubiquitous and more damaging biases demonstrated by the general population.
There is of course a problem with pointing out to people in the pub or in Facebook posts or at a poker table, or wherever, instances where they are guilty of such biases. I can guarantee that the one response that you will not get is something along the lines of:
Kahneman runs through the gamut, a few of which I may have mentioned before (e.g., anchoring). But hindsight bias is one of the most dangerous because it worms its way into many major decisions that are, in terms of expected value, plain wrong.
In simple terms (and Taleb has also covered this matter), hindsight bias is where everything that has happened seems far more likely in retrospect than it did at the time.
To this extent, I've rather given up on currency predictions and so on. If you get it right, people think that it was inevitable. if you get it wrong, people think that you are a dick, because the thing which actually happened was inevitable. And even if they don't think that it was inevitable, they think it was more likely than they thought beforehand.
In court cases for damages, this leads to juries and judges interpreting reality wrongly. Should an extremely unlikely event occur, causing serious injury, the chance of that event occurring will be seen, with hindsight, as being higher than it actually was. This means that the person deemed responsible for that event occurring will be assigned far more blame than he or she actually merits. The most high-profile recent example of this in the UK was the Baby P case.
Because third parties (e.g., surgeons, social workers, health and safety officers) are aware of this, their work attitude automatically (and, from their own point of view, correctly) gravitates towards "is my arse covered here?" decisions. It is not they whom we should blame. We should blame ourselves, and the judges and juries who operate with hindsight bias.
If we had a few more judges and juries (and newspapers) saying "yeah, it happened, but the chances of it happening were minuscule, so no blame can be attributed", then third parties would be less worried about being found negligent in cases where they were not really negligent.
But, as we know (and here other instinctive non-rational biases come into play) people need to find a narrative; they need to find causes, and this also leads them to need to find someone to blame, even if no-one was to blame.
And, as Taleb observed, no-one wins a medal for taking an action that leads to a major (but unlikely) disaster NOT happening. That's why nearly all major initiatives seem to take place "after the horse has bolted".
++++++++++++++
I'm currently at hindsight bias, which I think is one of the more ubiquitous and more damaging biases demonstrated by the general population.
There is of course a problem with pointing out to people in the pub or in Facebook posts or at a poker table, or wherever, instances where they are guilty of such biases. I can guarantee that the one response that you will not get is something along the lines of:
"Oh Em Gee !! You are right! The scales have fallen from my eyes!".
Kahneman runs through the gamut, a few of which I may have mentioned before (e.g., anchoring). But hindsight bias is one of the most dangerous because it worms its way into many major decisions that are, in terms of expected value, plain wrong.
In simple terms (and Taleb has also covered this matter), hindsight bias is where everything that has happened seems far more likely in retrospect than it did at the time.
To this extent, I've rather given up on currency predictions and so on. If you get it right, people think that it was inevitable. if you get it wrong, people think that you are a dick, because the thing which actually happened was inevitable. And even if they don't think that it was inevitable, they think it was more likely than they thought beforehand.
In court cases for damages, this leads to juries and judges interpreting reality wrongly. Should an extremely unlikely event occur, causing serious injury, the chance of that event occurring will be seen, with hindsight, as being higher than it actually was. This means that the person deemed responsible for that event occurring will be assigned far more blame than he or she actually merits. The most high-profile recent example of this in the UK was the Baby P case.
Because third parties (e.g., surgeons, social workers, health and safety officers) are aware of this, their work attitude automatically (and, from their own point of view, correctly) gravitates towards "is my arse covered here?" decisions. It is not they whom we should blame. We should blame ourselves, and the judges and juries who operate with hindsight bias.
If we had a few more judges and juries (and newspapers) saying "yeah, it happened, but the chances of it happening were minuscule, so no blame can be attributed", then third parties would be less worried about being found negligent in cases where they were not really negligent.
But, as we know (and here other instinctive non-rational biases come into play) people need to find a narrative; they need to find causes, and this also leads them to need to find someone to blame, even if no-one was to blame.
And, as Taleb observed, no-one wins a medal for taking an action that leads to a major (but unlikely) disaster NOT happening. That's why nearly all major initiatives seem to take place "after the horse has bolted".
++++++++++++++
