During the World War II, the US Navy was facing a rough situation. Many aircrafts used for bombing the enemy - called Bomber planes - were being hit in-flight from the enemy lines. And the Navy had no idea how to reduce the casualty.
The US Navy started analyzing where exactly and at what specific areas of planes were being hit based on the bullet holes they found on the returning planes. They realized that most of the damage was on the wings of the aircraft and the central body. The surviving planes rarely had damage on the cockpit, engine or tail portion of the planes.
They concluded that the wings and the body should be reinforced with extra armor and thus reduce overall casualty.
But a mathematician named Abraham Wald had different ideas. He told them they were interpreting the Data wrong. "Dead man don't tell tales"
Instead of checking the surviving planes damage, they should consider planes that have never returned. Those planes probably would tell a different story. It's very likely that the planes that returned to the base, survived because they were not hit on the cockpit, engine and the tail. And it will be better to rather reinforce these areas of the war planes instead of winds and body.
This was a complete 180 degree shift in interpretation of the available data. And that proved right. The US Navy followed Wald's advice, and sure enough they started seeing positive results. One simple change in idea made a huge difference in the outcome
The above example has been studied by behavioral psychologists and scientists and it has been often referred to as "Survivorship Bias". There's a human tendency to overlook "what is not presented or is not a visible data" and base their conclusion on what is only the available data.
A simple example can be - Both Bill Gates and Mark Zuckerberg - are success stories and both of them are college dropouts. And this can lead for many to believe that college degree is not relevant for success in life and Gates / Zuckerberg's success validates that interpretation based off of a limited data.
Survivorship Bias is a tendency to get carried away by rare success story without considering similar examples of failures.
For every day-trader who has made millions out of a bull run on the stock market, there are thousands who have failed trying to make big bucks fast on the same market. Yet we end up being appealed by that rare success story and end up investing on short term gains..then long term, albeit boring index funds for example.
If S&P 500 has given 8% return year over year for past 30 years that is less appealing and very boring compared to try and sling one big one from the Bitcoin rush! no? Who's going to wait for years to see a positive gain in your portfolio when you can buy a digital painting for 69 million dollars because you are a big crypto investor?
Data that is not visible to us is equally important as the data that is presented and given to us. It was important for the US Navy during World War II and it is important for us now. Taking a holistic view and not limiting your understanding to partial data size is very important.
We know millions are affected by Covid today. But in my opinion the bigger pandemic is and will be Mental Health. And data for that is yet to be seen.