I’m sure we all pride ourselves on our trustworthiness. But there’s a part of all of us which is a lot more suspect than we normally realise.
Billy Wilder once said “hindsight is always 20/20”. It’s a great line. It’s not necessarily correct. The weight of evidence and academic study shows that hindsight is far less reliable than you might imagine.
It’s galling that, when it comes to recollection and predictability of events, the director of Double Indemnity and Some Like It Hot is wrong and US politician Donald Rumsfeld is right. Unfortunately, his much-mocked ‘known unknowns and unknown unknowns’ speech does seem to have had a strong basis of truth.
It comes down to a phenomenon known as “creeping determinism” or, more pleasantly, “hindsight bias”. This boils down to believing that a result was predictable right from the start, despite this not being the case. Effectively, when we review an event whose outcome was difficult to predict, we often decide we “knew it all along” even when this wasn’t the case.
This may sound a bit odd but it’s been proven to occur in laboratory experiments and in the real world. In health care, a much-studied example, medics frequently overestimate their ability to foresee a patient’s outcome, for instance. Odder still, it’s been proven people demonstrate hindsight bias even when they’re told all about and asked to avoid it.
Why does this matter for us? Well, as project management is all about achieving best possible outcomes, we can see that the predictability of results has to matter. Good project management involves a constant effort to learn from experience, avoid mistakes and seek improvements. And hindsight bias shows that that may be more difficult than we might think.
If a project goes well, believing that this was inevitable and we did everything ‘right’ means we may not spot instances where we avoided problems more by luck than management. Similarly, when problems arise it easy for client or agency to decide “it should have been obvious” that the other party’s actions would lead to the wrong result, even when this is not actually the case.
So what can we do? There are a number of things that we do at Nudge that aim to avoid hindsight bias or reduce its effects on our work.
- Firstly, and most tangibly, a good project process will identify and plan for as many potential issues and outcomes as possible. A detailed requirements list, functional spec and, most importantly, risk register allow all stakeholders to not only plan for the best possible result but also to have a record of what was predicted before hindsight bias came into effect.
- Secondly, every project should be approached in a collaborative spirit. Where teamwork is superficial, success has many parents but failure is an orphan; strong teams, on the other hand, take collective responsibility for all outcomes and avoid apportioning blame wherever possible.
- Finally, a good project process will always involve a final review meeting or ‘wash-up’. Most will look at what went well, what could be improved and what lessons can be learned. But hindsight bias shows an extra step that could be useful: choose a few alternative and even undesirable outcomes and have the team consider and explain why these results ‘happened’ and what could be done to avoid them. Such a session could tell us more than simple hindsight can share and give us a more nuanced perspective on the job we do.