Named after Tanzanian boy Erasto Bartholomeo Mpemba, the Mpemba Effect is the observation that a liquid, typically water, which is hot can freeze faster than the same liquid which begins cold, with all other conditions kept equal. After many years of debating the phenomenon, experts have finally landed in the conclusion that hot water doesn’t usually freeze faster than cold water, but will under certain conditions occur.

Overall, there are two main debated ways by which hot water can freeze faster than cold water. The first reason is due to the fact that hot water evaporates and turns to gas faster than cold water. This means that if you start with equal masses of hot and cold water, there will soon be less of the hot water than the cold water, and thus the hot water will freeze first because of the lesser the mass. Another way it can happen is if you place hot and cold water in your freezer, and the hot water melts the ice under the bottom of its vessel, leading to better thermal contact when it refreezes.

A central thing about the Mpemba Effect is that it’s very uncertain. There are still no clear answers to why this phenomenon can occur, but rather all we know is that it can occur, with a few qualified guesses to support our understanding of it. The only thing we can actually say with certainty is that ”sometimes hot water will freeze faster than cold water”.

In an article written by my colleague Mikael Palmblad, Mikael speaks about The Value of Uncertainty, and how you can actually find value in uncertainty. Just like the Mpemba Effect, it depends on the circumstances. In the case for uncertainty, the circumstances depends on the relationship between knowing knowledge and usage of knowledge. He exemplifies with listening to a weather forecast, which says it might rain. To mitigate the risk of rain when you go out, you go into the closet, pick up an umbrella, and take it with you. That is value creation.

The gist here is that if you have knowledge about something that is uncertain, but don’t use that knowledge, you’re losing out on potential value, whatever form that might be. Another example he mentions is if you have a milk carton in your refrigerator that has passed its ”best before” date by two days. In general, people know what sour milk smells like. Thus, in order to use your knowledge about sour milk, you would open the carton and smell it, and depending on the smell, you would either drink it or pour it into the sink. This means that if you just look at the date and refrain from using your knowledge about sour milk, you’re losing value, and in this case milk, by not using this information. 

Quite often, we plan our lives according to static boundaries like fixed dates, specific times, or we measure our plans in increments like worst, medium, and best case scenarios. I think it’s fair to say that when you for example agree to meet a friend at a specific time, the only thing that’s almost certain is that you will not get there at the exact agreed upon time. Instead, you will either get there a little bit early or a little bit late, and so will you friend, even if we are just talking about seconds. Thus, mathematically none of you met the criteria of meeting at an exact specific time! Whether this was due to that the bus was early or that the weather was bad, the point is that the static value had uncertainty to it. Of course, for this example the consequences aren’t particularly dire for not meeting the static value, but what happens when the consequences are?

The trick with leveraging uncertainty is to understand the probability spread around the desired static value. Questions to think about should rather focus on the likelihood of meeting that static value, and consequently the likelihood of not meeting it. When using only static values, you run a high risk of becoming skewed and misguided in your evaluation of possible options. For example, imagine you have to choose between two options, Option A and Option B. Even though Option A may seem better than Option B due to labeling it with better static values, learning that the likelihood of Option A actually happening is much less than Option B may cause you to shift your decision towards Option B instead, or make changes the initial strategy you had for pursuing Option A to reduce its uncertainty spread. All as a result of the uncertainty involved. As such, uncertainty can help us value our decisions and add more ”reality relevance” to our understanding of decisions that we make, instead of simply trusting the static value. 

In the end, the real value of uncertainty comes down to the way that you choose to use your knowledge to mitigate uncertainty. Just like in the example with the umbrella or sour milk, refraining from considering uncertainty and acting on it with the knowledge you already have, leaves you with less overall created value. 

But wait a minute? How does this tie back together with the Mpemba Effect? Well, the main take away is to remember that even for a seemingly sure thing as cold water freezing before warm water, there is uncertainty involved. Sometimes hot water will freeze faster than colder water, that’s just the way it is. By simply acknowledging that there is uncertainty involved, you can begin to generate value, i.e, what do you do if the hot water freezes before the cold water? For this case of very little importance, you would probably do nothing. But for something of greater importance, I’m sure you would do everything in your power to minimize the uncertainty involved, and once that’s done, use your knowledge about what the remaining uncertainty entails in order to extract as much value as possible from it. If you can do this, then congratulations, you just found value in uncertainty!