C2. Probability vs. Value
Probability vs. Value
Cigarette and gaming companies are amongst the most destructive forces in our society, still legal and making impressive profits. These companies are generating profit from addiction amongst their customers leading to deteriorating health and finances, generally from those not very well off to begin with. One of the drivers behind why we allow this to happen is our inability to judge probability and risk. This inability allows us to buy cigarettes that we know increase cancer risks by several hundred percent while also adding a lottery ticket, knowing it’s a statistical loss.
Acting on probabilities has proven difficult for society as well. Climate change is only one example where a very well-established and significant risk for humanity is still rarely acted upon, as politicians imagine the impact after their reelection. While often challenging to estimate probability and risk, in this case, we can conclude that robust models indicate an immediate need to act, which is collectively ignored.
Taking the future into account becomes even more difficult in those situations where the probability is very low while the value/risk may be very high. In the field of existential risk, future generations inhabiting earth over thousands, or millions of years are considered as a potential loss if today we would fail to mitigate an end of our civilization. While the risk of such an ending is likely and hopefully low, the emergence of AI, resistant viruses, and continuous close calls with a growing nuclear arsenal reminds us that the risk is not negligible.
However large or small we may think this risk is, the value at stake becomes staggering when multiplying the low risk with the potential extinction of all future generations of humans on this planet. When applying this expected value, it is remarkable that virtually no resources in our society are dedicated to mapping out and mitigating these risks. It seems human extinction should be something worth paying attention to.
The concept of time has always confounded us. If we were faced with our future right now, we would likely not have picked up that cigarette, lottery ticket or ignored climate change. Still, we are. In an abstract sense, though, we exist in the past, present, and future all at once. Our present is built entirely on the past, of which events and experiences guide our every moment. Equally so, the future is a deterministically determined path of what is to come.
Hence, it’s counterintuitive to discount future experiences for the passage of time in itself. Future always comes in the form of now. We understand this intellectually but bridging the perceived gap between the now and the future will be key to our survival. We understand that the child waiting for the second marshmallow will use its abstract mind to imagine the future value of two marshmallows to the urge of consumption here and now. We just need to realize the same goes for our societies need to move from quarterly earnings and four-year politics towards a long-term vision that may, in the process, give us the purpose we so long for.
In our own lives, there may be numerous reasons for and against taking a leap of faith towards a risky and exciting new path in life. Here we can imagine the future value, even though often simplified and naively. To better understand the likely value, multiply it with the realistic chance of obtaining that value. This method can be applied to money invested, insurance bought, or happiness from a successful project.
When estimating probability, be conscious about our biases and triangulate by asking people with relevant experience to better evaluate unbiased probability. Imagine there is a true answer to what probability to put on certain events and that you must bet a month’s salary on getting it right. While always difficult to assess risks and probabilities, the simple exercise of an imagined high stakes bet may give us a more transparent idea of the value and risk of different options.
Pay extra attention when a choice involves several unknown decisions in sequence as it quickly adds risk and diminishes the expected returns. If a new project could give us €100k in equity, given that funding is acquired (50 percent chance) and that we eventually become a partner of the firm (40 percent chance), the real value of that happening would be €100k *50 percent * 40 percent = €20k
For low-probability, high-impact risks in our lives, such as a traffic accident, it’s often very difficult to emotionally relate to hypothetical outcomes far in the future. Micromorts, introduced by Ronald A. Howard , measures a one-in-a-million risk of death for different activities. In practice, riding a motorcycle for one hundred kilometers over one hour equals ten micromorts and a statistical risk of death of one in 100,000 hours. It seems low, but when we consider that you may have more than 400,000 hours left to live if you crash when thirty-five years old, this would equal “being dead” for four hours for every one hour you ride your motorbike. Now, this insight is something we can emotionally relate to and use for decision-making.
In the end, these models should be used as inspiration where we need to translate complex data into something intellectually and emotionally digestible. It should not overshadow the great power in us as active agents changing these parameters to our advantage. If you love riding a motorcycle, consider practicing how and where to drive safely. Simply staring at the bleak numbers of statistical probability won’t show the whole picture. These are external factors to be aware of, but given this awareness—we as free agents can shift those numbers and accomplish the unexpected if we put our mind to it.
SYNTHESIS:
- - Dare to quantify different options to reduce complexity between multiple options.
- - Challenge our biases and check data to determine probability.
- - Multiply impact with probability to better relate to abstract risks.