It’s a tricky business, odds. Even when people know them, half the time they want to deny them. Personally, I am all for that. It puts me in the inimitably luxurious position of always being right, which lends the correct balance to the universe. There are some things that even I have a hard time denying. For instance, time is kind to some, but, on average, it has its wicked way with most people. This undeniable effect that time has on humans explains why rocker Axl Rose, of Guns ‘N Roses fame, now looks more like former US Secretary of State Madeleine Albright, or even me, the Climate Bandit, than he does his young self. Ok. Photos of Madeleine and I are more attractive and cheerful. Still.
This also raises the issue of how to communicate scientifically quantified probabilities to non-scientists. Scientists tend to communicate probabilities in numbers. Non-scientists tend to use phrases in a way that corresponds to the numbers that scientists use. Sometimes using phrases works well, sometimes not.
Below are results from a survey by authors writing for the Harvard Business Review. Some phrases are more useful than others. These results suggest that the phrases “serious possibility,” “real possibility,” “possibly,” and “might happen” all give the idea that the chances are distinctly above zero and below 100%. Other than that, these phrases are not particularly useful. In comparison, the phrases “almost always” and “not often” are more precise and respectively mean greater than 75% and less than 40%. Bringing us to, what is up with “always” and “never,” which, according to these results, seem to really mean maybe, or just possibly?
The probability (0% to 100%) that survey respondents associate with a word or phrase.
The above findings concerning how people react to “never” do cause me to understand why my Great Auntie Aphra thinks that the pop star Harry Styles might like to go on a date with her and view 500 or so photos she took of her 1977 summer vacation to Fort Lauderdale, Florida. Another mystery solved. What a relief. That said, the findings pose quite a difficulty for scientists regarding “always,” implying that, even if confronted with a probability of 100%, people may still interpret that as meaning “but there is a possibility it will not happen.” This is a result of something called optimism bias.
Optimism bias—when one has difficulty accepting a high probability for a negative event or outcome, even if the probability of a positive outcome is very low. My Great Auntie has this bias. So can you!
This brings me to my main point, which, speaking of probability, you might have begun to think was “really never” going to happen. When it comes to climate, i.e., long-term weather, a “100-year flood”—is not. A 100-year flood describes a flood that has a 1 in 100 probability of happening in any given year and is calculated according to past statistics on flooding. Past. What happened a while ago. Events that my Great Auntie remembers because she lived through them. Maybe even events that her Great Auntie lived through. It’s a statistic based on events that are likely not fully modern, if by modern we mean “since Harry Styles first rocked the holy-swinediving-piglet-cute out of a bubblegum-pink tutu”. That would be just fine, if the world were constant and stayed the way it was when Queen Victoria of England was young.
(left) Great Auntie Aphra (right) Harry Styles wearing a pink swan ballerina costume
As examples of how climate is altering the probability of flood events, the city of Austin is updating its floodplain maps to reflect that some areas previously designated in the 500-year floodplain are now in the 100-year floodplain, Houston has recently endured three 500-year floods three years in a row, and Central Texas has seen three 100-year flood events over five years between 2013 and 2018.
Even if we use statistics that pre-date Great Auntie Aphra’s Great Auntie, over the payback period of a 30-year mortgage, a home in a “100-year” floodplain has a 26% probability of being flooded over 30 years. This is a probability, that is, the mostly likely “chances” as we know them. Calling something a “100-year” event, an approach that reflects the way probabilities are sometimes expressed for other “rare” weather-related events, such as wildfires, tornados, hurricanes, etc., could be just the fuel optimism bias needs.
Thinking about these ideas, how do you think scientists should communicate probabilities to non-scientists? Put your best foot forward now—if Harry Styles is reading, your answer might help him decide how to communicate “always” when it comes to a pink tutu and “never” when it comes to acquaintance with photos of certain past events in Fort Lauderdale.
Readers are referred here for an intelligent and fully scientific version of this article.
Further Reading:
- If You Say Something Is “Likely,” How Likely Do People Think It Is? (Harvard Business Review)
- Return Period Calculator (GFDRR)
- The ‘100-year flood’ fallacy: Return periods misleading in communication of flood risk (Earth Magazine)
- The 100-Year Flood (USGS)
- What’s a 100-year flood? A hydrologist explains (The Conversation)
This is something I’ve been thinking about, how to identify terms used in scientific, academic, or medical settings that have different meanings in regular conversation. The English language doesn’t make it any easier, when the same word can have a significantly different meaning in a different context. (Is it technically a different word then, just spelled the same? I don’t know, it’s confusing.)
When speaking to the public, scientists need to define or clarify terminology every time it’s used, especially when it’s something that sounds obvious. In the case of a 100-year flood, it needs to be stated, “100-year flood, or a flood with 1% chance of occurring.”