The problem’s that we plan looking forward but we can see more looking back.
Imagine a time just before the discovery of electricity. How would a forecasting model for the future have looked? It would have been about trends in household utilization of lamps and wicks and wax and oil for the next X years. We would have simply assumed that that trend would continue. But look at what happened. Electricity came along, and because of it, society transformed. The same goes for other inventions like the toaster (more breakfast, more productivity, more factories) or the elevator (taller buildings, denser population, more cities).
_We tend to overestimate how much things will change in the short-term and underestimate how much they will in the long-term. _That’s because change at scale is not linear. It is exponential.
What if you could imagine the end–end of a project, business plan, career switch? What if you could, as Shane Parrish from Farnam Street says, make the hindsight of your future self the foresight of your current self? This is where a little reverse engineering helps. The general approach is called backcasting.
New knowledge elicits new behavior. This undermines the benefits of a predictive model, which cannot assume new knowledge.
To sum up, if we believe that people will have new ideas and new knowledge in the future, we cannot hope to be able to predict the many ways in which that will be used to build anything of non-trivial size (a business, a community, a society) in the long run.
So instead of predicting the future, what if we tried explaining the past in a way that we could follow? We look back at events that have already occurred and we provide a satisfactory explanation for them. Instead of asking what the future will be like, we try and answer why it will be what it’ll be.
Backcasting doesn’t bother with predictions. Any meaningful future (career, relationships, business, health) is long-term and uncertain. So instead of asking ‘what will happen’, try working out ‘how something you want will happen’. What do you need to do to make it happen? Reverse-engineer.
- Place yourself in your ideal future. Capture detail. What do you feel like? What does the world around look like?
- Write down what you did to get there. List specific actions/behaviors/decisions. Don’t say ‘I have to do this’; say ‘I did this’. Saying so brings ownership.
- Now that you’ve a rough path of actions, you need signposts to hit. Where will you be:
- Three-quarters of the way
- Halfway (when will halfway come timewise)
- In the near future from now
Here’s how I reverse-engineered a plan to prepare for and run 6 marathons south of 4:30 hours. I didn’t know the approach formally as backcasting.
#1: My desired future state
I’ve run 42.195km in four hours. It’s warm, sunny, after 10am. I’m tired but I’ve got enough to savor the moment. I can walk and talk.
#2: Behaviors repeated to get to the end
* I mirrored run times with the race-day time (only morning runs)
* I ran 4 times a week for 12 weeks
* I did my longest weekly run over the weekend
* I upped my weekly mileage by 15-20%
* I took no fluids for any sub-10km run
* I will not miss any Sunday football matches
#3: Signposts along the 12-week program
- Week 9: I did 31km in 2:45 (three-quarters)
- Week 6: I did 21km in 1:45 (halfway)
- Week 3: I did 10km in 0:50 (near future)
Though I aimed for a 4h finish each year, my times across six years were spread between 4 and 4:30. Looking back, I didn’t think much about nutrition and recovery. But I knew what was most important to me, and that was enough to bring me close to my goal each time.
You can make better decisions to get to your desired future if you fully understand all that is required to make it happen. Backcasting brings that out by placing you at the summit and having you look down.