That pandemic came in waves that were much more distinct than what we have experienced. The first wave was extraordinarily mild. The French Army suffered 40,000 hospitalizations but only about 100 deaths. The British Grand Fleet had 10,313 sailors fall ill but only four deaths. Troops called it three-day fever. It was equally mild among civilians and was not nearly as transmissible as influenza normally is.

Like SARS-CoV-2, the 1918 influenza virus jumped species from an animal to humans. As it infected more humans, it mutated. It became much more transmissible, sweeping across continents and oceans and penetrating everywhere. And as it became more transmissible, it caused a much, much more lethal second wave. It became the worst version of itself.

The rapid spread of covid-19 in the United States began in early 2020. A lot has changed in our day-to-day lives since then, including the use of face masks. (Allie Caren/The Washington Post)

In that second wave, the 1918 virus had an overall case mortality in the West of 2.0 to 2.5 percent, but that average is meaningless because it primarily killed select age groups: children under 10 and adults 20 to 50. Metropolitan Life found that, of those aged 25 to 45, it killed 3.26 percent of all factory workers and 6.21 percent of all miners; and yet it barely touched the elderly.

U.S. Army training camps routinely recorded case mortality over 10 percent; at Camp Sherman in Ohio, case mortality exceeded 21 percent. In 13 studies of hospitalized pregnant women, the death rate ranged from 23 to 71 percent. In a few isolated small settlements in Alaska and Africa, it killed everyone.

Virologists expected SARS-CoV-2 to mutate more slowly than influenza, and between its emergence and November 2020, the virus did seem remarkably stable.

Full coverage of the coronavirus pandemic

Thats why last year, when I was repeatedly asked whether I worried that SARS-CoV-2 would, like the 1918 virus, become more lethal, I always replied that, even during 1918s mild first wave, that virus had on rare, isolated occasions demonstrated its potential to kill in, according to an Army report, from 24 to 48 hours. Since the SARS CoV-2 virus had not shown any indication none of increased lethality, I was not concerned.

But in the past several months, different variants have surfaced almost simultaneously in Britain, South Africa, Brazil, and now in California and New York. Each of these variants has independently developed similar and in some cases identical mutations and achieved greater transmissibility by binding more efficiently to human cells.

A virus that binds more efficiently to cells it infects would, logic suggests, also be more likely to bind to a larger number of cells, which could, in turn, increase disease severity and lethality. On Wednesday, BMJ, formerly the British Medical Journal, reported that Britains so-called U.K. variant was 64 percent more lethal than the virus it replaced.

There is not enough data to evaluate the variants first identified in South Africa and Brazil, but whether or not they are also more lethal, one thing is certain more variants will arise. Mutations are random. Most either make the virus so defective it cant function or have no impact at all. But this virus has already demonstrated that it can become more deadly and evade some immune protection, making vaccines less effective. If we allow the virus additional opportunities to mutate, it will have more opportunities to become the worst version of itself.

There is no reason to expect that this virus will suddenly turn into 1918. There are limits as to how far it can mutate. But the more people who abandon masks and social distancing, the more infections can be expected and the more variants will emerge.

In gambling terms: If you roll the dice once, yes, there is only a 2.77 percent chance you will hit snake eyes. But if you roll the dice 100,000 times, it is virtually certain snake eyes will come up several thousand times.

Right now, policymakers are making decisions that will limit or expand opportunities for the virus to spread and mutate. Most proposals will require weighing costs, benefits and risks, such as when and how much to reopen the economy or delaying second doses of vaccines.

Wearing masks requires none of these calculations.

We know masks decrease transmission. Lifting a masking order not only means more people will get sick and die. It also gives the virus more rolls of the dice. That is a fact.

The variants we have seen so far do not worry me much. The variants we have not yet seen … yes, they worry me. To increase our risks is, simply, foolish.

Read more:

John M. Barry: The coronavirus is mutating. Will our vaccines keep up?

Leana S. Wen: The CDC is missing a critical opportunity to get Americans vaccinated

Matthew Kavanagh and Madhavi Sunder: Poor countries may not be vaccinated until 2024. Heres how to prevent that.

Kenneth Osgood: For cancer patients, trying to get a coronavirus vaccine is a dark comedy

David Von Drehle: The Johnson & Johnson vaccine works great for everyone. Its a perfect match for young adults.