Living with our assumptions (judgements)

We put assumptions in our spreadsheets and calculators to reject or not reject deals to some degree and relationships to a larger extent. We adjust our assumptions to show potential outcomes of different scenarios. But let’s talk about these scenarios and numbers that go into our assumption-making process.

These are today’s forecasts – that is, tomorrow’s proxies. Research shows that most of us don’t think about our forecasts, most optimally. Most of us would say we would expect the value of X or Y is likely or Z occurs most of the time. Decision experts would say that these exact values (e.g., X, Y or Z) may be precisely wrong from the onset as forecasts. (For the purpose of this blurb, let's call them assumption experts rather than decision experts.)

What then? According to super forecaster Warren Hatch (from the Good Judgement project) in 'Superforecasting': The people that predict the future – BBC REEL, he said, “When we are thinking about forecasts, all we're saying is we're thinking probabilistically about an uncertain future. So rather than shrugging our shoulders and going with a hunch, we're finding uncertainty that we can quantify and we're finding that we can quantify more than we thought and we're finding that we can quantify it with more precision than we knew. There's a process that anyone can apply to improve their forecasting accuracy to make better [assumptions].”

I would encourage digging deeper into how to make better judgements from Philip E. Tetlock et. al. Some hot takes are, as follows:

(1.) Start with outside views, end with inside views. For example, get the base rates.

(2.) Think about forecasts or our views of the future as probabilities. For example, get the ranges of numbers behind what are likely/unlikely about events.

(3.) Get different perspectives, neither cloned answers nor affirmations per se. For example, do true updates.

(4.) Tests our beliefs. For example, learn about and avoid mental traps.

According to Warren Hatch, a former Wall Street prognosticator, “A good forecaster, they're always testing their beliefs. So rather than protecting them inside a lock box, they're always challenging them. ‘What am I missing’? A bad forecaster will have a belief about the world. A model of the world and defend it against all challenges because they're convinced this is right. So that's the idea of a hedgehog [See Note 1] who just really grabs on to something and does not let go versus a fox where he'll [she’ll] be agile and eagerly seeking out new information. If you're looking for people to put on your team, you can ask them: 'Is it a good idea to change your mind when new information comes in?' Hedgehogs will say ‘no’. Foxes will say ‘yes’."

As we know now, our spreadsheets, a.k.a., our models, containing our point estimates, are wrong or will be wrong. To borrow the words – a phrase that held true when I validated models in banking industry – from George E. P. Box, “All models are wrong; some models are useful.” All models, including the models that rest behind our eyes, between our ears and inside our heads, can be more useful when we improve upon our abilities to process uncertainties. It turns out that super forecasters are people – mind you, people, still yet, in an era of big data, machine learning and artificial intelligence – who are better at making judgements and predicting future events. Super forecasters would live better with and rest more confidently with their assumptions because their forecasts turn out to be more accurate in reality.

In closing, Warren Hatch said “It sounds esoteric, but it's actually quite simple. It's a world of [assumptions] and thinking more clearly is actually easier than it appears. And is easier than not doing it in the first place.”

Note 1: If you were tracking with the video at frame 6 min.,15 sec. of 10 min. total, then, look out for that hedgehog and think like a fox.