When people ran into problems attempting to use New York's health insurance exchange website on Tuesday, the state released a statement claiming that the site was experiencing problems because it had “received an overwhelming and unanticipated response of 2 million website visits” within hours of opening. By 5 p.m., the Associated Press reported, it was claiming 10 million.

Yet at the same time, federal officials were boasting that the federal website, Healthcare.gov, received 2.8 million visits.

How can both claims be true?

New York is no doubt a big state, with a population approaching 20 million people. But the federal exchange serves 36 states with combined populations of 209 million*, or more than 10 times as many people. Texas alone has a population of 26 million, and a larger percentage of them are uninsured.

What’s more, New York officials said that more than 9,000 New Yorkers had “shopped” for insurance, but they didn’t say how many enrolled in a plan. So, to sum up, New York is claiming more than three times the web traffic of the federal exchange site for a state serving less than one-tenth the population. Yet only a small fraction of these visitors were actually able to shop for insurance, but we don’t know how many people came away with coverage.

On Wednesday, the New York site still contained this message: "Due to overwhelming interest in the NY State of Health — including millions of visits during the first day of site launch — the health exchange is currently having log in issues. We encourage users who are unable to log in to come back to the site later when these issues will be resolved."

Web traffic numbers are notoriously easy to manipulate (unique visitors versus page views, for instance). Given that New York’s website has been extremely buggy, officials have every incentive to blame the problems on overwhelming demand. And given that I have no access to the raw data, I can’t say that’s wrong.

But I do think there is good cause to be skeptical and not to place much stock in the web traffic numbers being thrown around. What really matters is how many people actually enroll, whether enough young and healthy people enroll and whether the exchanges are accurately calculating subsidies. Those are things that we won’t be able to really determine for months, as the enrollment period extends through the end of March.

*Due to the government shutdown, I could not access Census.gov, so I used Wikipedia for the state population figures.