From the outset, there was a predictable pattern to the coronavirus pandemic. People would get sick. Some of them would need to be hospitalized. Some of them would die.

The timing of the pattern varied a bit. It generally took a few weeks to go from infection to death, but because of reporting differences and variations in when people got tested, there wasn’t a well-delineated time frame. But the order was concrete: infection, decline, death.

There was no period in which that pattern unfolded more steadily than in the United States’ third surge, which began in mid-September. By early January, a quarter-million Americans were being confirmed as having contracted the virus each day. (The figures below are seven-day averages of each metric to smooth out reporting discrepancies.) The number of people in the hospital with covid-19, the disease caused by the virus, surged to almost 140,000. The number of people dying of the disease each day stayed above 3,000 for nearly a month.

On Dec. 14, 2020, though, the first coronavirus vaccines were rolled out. The process was slow; it took until Jan. 11 before more than 10 million doses had been distributed. By about that point, all three metrics had already begun to decline.

The drop in cases and deaths was sudden and dramatic. Beginning on Jan. 12, the number of new cases began to drop at an average rate of 4,500 new cases for the next 40 days or so. The number of deaths began to trend downward starting Jan. 17.

We certainly can’t say that the vaccine rollout prompted that downward trend in cases. But that drop did look different from the other two surges. For one thing, all three metrics dropped relatively evenly. In the previous two surges, the number of deaths peaked at least a week after. In the third surge, that peak came after only five days.

What’s also interesting is that the uptick in cases over the past two weeks has not yet spurred an uptick in deaths. It may simply be too early; it took nearly a month for deaths to head upward after the third wave began last September. But there’s also reason to think that the number of deaths won’t increase at the same rate relative to new infections as has occurred in the past.

We’ve repeatedly looked at the ratio of new cases to deaths. For months, there was a consistent relationship between the two, with the number of new deaths on any given day reflecting 1.7 or 1.8 percent of new cases three weeks prior. This, too, moved around a bit, but it was a good rule of thumb for estimating what the daily death toll would be as cases surged.

In recent weeks, that ratio has weakened. From Sept. 12 to Dec. 31, the ratio of deaths to new cases 26 days prior was 1.74 percent, meaning that the number of deaths each day was usually about 1.74 percent of new cases about four weeks prior. Since Jan. 1, that ratio has been an average of 1.43 percent.

Jan. 1 is probably too early a benchmark to use, given that the Pfizer and Moderna vaccines that have constituted most of the administered doses take five to six weeks for full effectiveness. It’s also the case that there are a lot of ways to generate those averages. Use a 21-day gap and the gap is narrower, 1.64 to 1.54 percent. But the standard deviation for a 21-day period is higher than it is for a 26-day period, meaning, broadly speaking, that it’s a closer fit to the actual data.

All of these graphs and numbers obscure the point: Fewer people appear to be dying of the virus. What’s shown above is another wobbly metric, but potentially an important one.

Of course, this is what we’d expect as vaccine distribution continues. The vaccines are very effective at preventing serious illness, so vaccinating more people would presumably reduce the number of people who progress from infection to death. Data from the Centers for Disease Control and Prevention indicate that three-quarters of Americans ages 65 and older, the group at highest risk of death, has received at least one dose of a vaccine. About 57 percent of that group has been fully vaccinated.

Interestingly, those 75 and older — a group which constitutes more than half of the coronavirus deaths recorded in the United States — is only slightly more likely to be fully vaccinated than those ages 65 to 74. For every group under 75, the density of the population which has been fully vaccinated is higher than their density among deaths.

But it is still the case that more than half of Americans over the age of 74 have been fully vaccinated. Those most likely to die of the disease are the group that is currently most protected against it. That’s a relationship that should, over time, keep pushing coronavirus deaths lower.

While everything related to the virus feels eternal and heavy, we’re still fairly early in the vaccination process. Even though more than 167 million doses have already been administered, it’s been less than four months since the first American was vaccinated. It’s a disease which progresses over weeks, and only 16 weeks have passed since the first dose was given. Those whose Pfizer doses are just now hitting full effectiveness got their first doses about five weeks ago, when the number of vaccinations that had been administered was only about 77 million.

The pattern we’re seeing, though, offers tantalizing hints that we’re going in the right direction. It’s a new, much better pattern: Vaccinate and push down the number of infections, hospitalizations and deaths all at once.