Imagine you are a smoker who drives for a living. Every day you throw burning cigarette butts from the car. It's happened many thousands of times without incident. The smouldering stub lands, bounces and dies. Occasionally one may singe a clump of grass. You drive on in blissful ignorance.
This analogy, say virologists, is a good way of thinking about zoonotic spillover events like the one which sparked the Covid-19 pandemic. They are happening all the time but are seldom noticed. Only very rarely, when environmental conditions are just right, do they spark a violent wildfire that spreads exponentially.
"These things happen every day around the world, and we miss them because we don't focus enough on supporting surveillance of unusual illnesses in underserved and often distant rural communities", says Dr Peter Daszak, the veteran virus hunter, and one of the world's foremost experts on coronaviruses.
Working backwards to find the origin of Covid-19 is necessary - but to focus on this is to miss the main point, adds Dr Daszak. If you are very lucky, you might find the smoker who sparked the wildfire, but that won't stop the next one unless something fundamental changes.
"The key problem is that too many ill-informed people are working outside of their own experience to try to 'trace back' from ground zero", says Dr Daszak. "What we need to do is look at it the other way round: from the bats to the people".
The idea that zoonotic spillover events are much more widespread than we realise is not just a theory but a fact, and variants of the Sars-Cov-2 virus are a case in point.
A study published in 2018 by Dr Daszak and Shi Zhengli - Wuhan's famous "batwoman" - documented how they took blood from 218 people who live in close proximity to bat caves in Yunnan province, China. Most were farmers, and 97pc of them had a history of exposure to or contact with livestock or wild animals.
In total, six of those tested (2.7pc) were found to have antibodies to a Sars-Cov variant carried by local bats. The virus did not appear to have caused harm, but even if it had, we may never have known. In the 12 months before the sampling date, only one of those infected had travelled outside of Yunnan. Several of the others had never left their village. This particular spark - like the vast majority of them - briefly smouldered, then died.
But that's not always the case. Sometimes the conditions are right for things to catch fire. And on occasion, the blaze can be vast: one recent outbreak, HIV, has cost 30 million lives to date and is still burning.
Take Hendra virus. Its traits mean it has never triggered a major outbreak - it isn't contagious enough - but every so often, it still sparks a lethal blaze.
The first reported outbreak was in 1994 in Hendra, a small suburb in Brisbane, Australia, when a racehorse called Drama Series fell ill. Within 24 hours she was dead. Drama Series was the first of 21 racehorses to die in that outbreak. But the disease also jumped to two humans, killing one - a horse trainer called Vic Rail. It took years to find where the Hendra virus came from, but in the end it was traced to fruit bats.
A bigger viral blaze was sparked in Yambuku, a village with a Belgian missionary outpost deep within the Congolese forest, in 1976. A mysterious hemorrhagic fever killed well over 200 people in just three weeks before apparently petering out. The virus - later named after the nearby Ebola river - has now sparked more than 28 different outbreaks (each a separate spillover event) and in the process has killed more than 22,000 people over the last 40 years.
The worst occurred in West Africa between 2013 and 2016, when the first spark fell near a major road, enabling its spread. At least 11,323 people died.
Experts suspect the "natural reservoir" for Ebola is also bats - but they are far from the only animal to carry zoonotic viruses. The Spanish Flu of 1918 is thought to have started in North American poultry and then spread through a globe ravaged by World War I.
Middle East respiratory syndrome (Mers), another coronavirus, comes from camels and has killed 858 people since it was first discovered in Jordan in 2012. The vast majority of spillover events go unreported, say experts.
"We're continually exchanging viruses with animals, that's what happens", says Dr David Redding, from the Centre for Biodiversity and Environment Research at Unversity College London.
It is HIV that best illustrates the point, a virus now known to date back to the early 1900s. Its simian version (SIV) is thought to have jumped from monkeys to humans through hunters and butchers in Africa. Dr Daszak describes HIV, which has killed an estimated 32 million people, as the "ultimate" example of spillover.
After many decades of repeated small-scale flare-ups, it exploded as a pandemic in the early 1980s. What had changed was not so much the virus itself - the spark - but the society it landed in. The population boom in Africa, the globalisation of air travel, the sexual revolution in the west all played a part.
"Changes to human behaviour increase the transmission of viruses between people," says Dr Dazak.
Covid-19 may also have been circulating longer than thought. Today there are roughly 300 animal pathogens from 25 "high risk" viral families, which are known to infect people. That more will emerge is inevitable. Researchers estimate that as many as 1.7 million viruses from these same families exist in the wild, including some 700,000 with "zoonotic potential".
That's a lot of sparks.