So are we having a fourth wave of the COVID-19 pandemic in the U.S., or not?

University of Minnesota’s Michael Osterholm has forecast a COVID “Category 5 hurricane” this spring. To some people, a look at hard-hit Michigan could justify this level of alarm.

But Paul Offit of the Children’s Hospital of Philadelphia could also claim to be right with his much more optimistic view that vaccines will mostly tame the pandemic from this point forward. After all, U.S. death rates and hospitalizations are not spiking, and may never return to anything close to the fall/winter peak.

Predictions have become a Rorschach test, where people can find doom or hope or reassurance. But what’s becoming clear is that public health officials don’t really know why this pandemic rises here and falls there.

It’s no longer plausible that it’s all happening because of mask mandates, or tightening or loosening restrictions more broadly. If restrictions and rules were the only factor at play, the current surge wouldn’t be in Michigan but in a state that has reopened more aggressively, such as Texas or Florida. And cases wouldn’t be so high in those previous paragons of virtue, Maine and Vermont. And why would cases be rising now in Canada?

In an interview last month, University of California, San Francisco physician Vinay Prasad brought up a few factors that might influence regional and national waves. One is initial conditions. In the first wave, some states in the Northeast had many more silent cases than others, so cases stayed high in New York and New England despite draconian policies.

Another is variants. Michigan might have been unlucky to have been seeded with more people with the more infectious variant of the virus, B.1.1.7.

Then there could be seasonal effects that are more complex than just fluctuations in temperature. Scientists have been suggesting environmental factors from pollen exposure to air conditioning to humidity.

Regional differences in COVID-19 peaks may also reflect irregularity in the way cases are reported, said Prasad. There are huge differences state to state in how many people are tested, which people are tested, and how positives are tallied up.

And then there’s just plain random chance, perhaps the hardest factor for people to accept.

Prasad said there’s been a consistent bias in the direction of pessimistic predictions, because in the epidemiology community, there’s a sort of moral exemption for making overly dire forecasts. But if you predict fewer cases, hospitalizations or deaths than come to pass, you’re seen as downplaying the pandemic, “and you’re a villain.”

Then there’s the assumption that our non-pharmaceutical measures are a major driver of cases, something that might be true but isn’t necessarily supported by data. Mask mandates mostly affect mask wearing outdoors, where many experts say the benefit is negligible. People in no-rules Florida still wear them in supermarkets, and people in rule-following Rhode Island have been contracting disease in private gatherings.

Prasad sees the bias toward restrictions as part of a broader bias in the medical community that overestimates the power of intervention.

A sort of fibbing has also become accepted practice in the public health community, said risk communication consultant Peter Sandman. The idea is that it’s OK to mislead people in order to encourage healthier behavior. Big and small risks have been lumped together.

That may not be working well with the pandemic. Some people have seemingly decided to ignore public health advice altogether, while others have been scared into being hypervigilant.

Even if we can’t predict the size or timing of a next wave, public health officials do know that the disease will continue to kill as long as some people don’t qualify for the vaccine or can’t book an appointment.

This would be a good time to focus the message on what is known about risks, and be more upfront about what’s still not understood.

Faye Flam is a Bloomberg Opinion columnist and host of the podcast “Follow the Science.”