Hindsight Bias and IT Projects
by Andrew Bilsdon, Delivery Lead – Altis Sydney
‘Understanding of a situation or event only after it has happened or developed.’
So why would we bother writing a blog about this, how can we understand something before it happens?
My thoughts are not related to the order in which we do things, I am going to explore our ability to understand and then the bias when we apply what we think we learnt.
I will set out an example of hindsight bias so we can explore the concept.
Let us imagine the traders in a stock market who, despite their best efforts are caught up in a crash that wipes out all of their positions and leaves them all jobless. Despite what you think of capitalism, this is a bad outcome!
After a few months, new traders are employed and they start cautiously to buy and sell stocks. They also look at the tell-tale signs that should have been seen in the previous crash, they are motivated to avoid the same fate that befell their poor predecessors.
Now the bias comes in, as humans we are great at reviewing a situation that has occurred and looking for the data that predicted the outcome. The problem is that we are only good at this when we know the outcome (retrospective), we are not good at this before the outcome (proactive).
The new traders will look at the circumstances of the previous crash, and with a hint of arrogance convince themselves that “they would have seen it coming”.
What this means is that they will convince themselves that they have learnt a lesson which will help them avoid a similar fate. They may even feel like they have an advantage, right up to the next crash.
The boom crash cycle of financial markets is often held up as a classic example, other cases include hindsight bias in:
• Medical outcomes – “Of course the procedure shouldn’t have been carried out, the patient was bound to die, given their weight/age!”
• Engineering disasters – “Why would they have built a bridge there? Of course it would collapse if there was a freak storm!”
• Natural disasters – “I could have told them there was going to be a major bush fire this year. When I told them to back burn, they didn’t listen!”
In IT, we are all about learning lessons, we have Post Implementation Reviews (PIR), Retrospectives, Lessons Learnt, etc which are all fantastic and absolutely appropriate, however, in 15 years I have never been at one where someone has tried to discuss or coach the group on hindsight bias.
I suspect it is because it comes across as being negative or unhelpful, we like to feel that next time will be different, our project will be better as we are not going to fall into the same traps as last time.
My recommendations are to:
• Don’t give up on learning lessons and retrospectives, however, have a conversation on hindsight bias with the group.
• Encourage de-biased thinking by ensuring that they don’t omit important information about what was known at the time of the event. This doesn’t mean making excuses, it just means being open and honest.
• Recognise that at best we will reduce issues, will not stop them, we are unfortunately destined to learn some lessons many times, just like those poor traders