Here's what I meant by my USA Today quote.

Yesterday afternoon, after doing a little victory lap over nailing the First Week enrollment total via HealthCare.Gov (I was off by 1.2%), I was asked to comment by Jayne O'Donnell of USA Today:

Of the 543,000 people who submitted applications and chose plans, 34% were new customers and 66% were re-enrolling, CMS said. Nearly 1.2 million people submitted applications for coverage.

Health and Human Services Secretary Sylvia Mathews Burwell called the numbers a "solid start." But she warned that It’s difficult to directly compare this year to last year's start of open enrollment as it started on Nov. 15 last year.

"We know from past experience that people are deadline driven, so we don’t expect to see deadline effects for a few weeks," she said.

As for the numbers, Charles Gaba, a data expert who runs ACAsignups.net, says it's "hard to say what it means yet." The more telling dates will come right after Dec. 15, which is the deadline for people who want to have plans Jan. 1, and during the last week of January, which is the final deadline for enrollment for 2016 plans, Gaba says.

When signups for the state exchanges are added, Gaba estimates the total enrollment so far to be about 714,000.

Don't get me wrong, I'm obviously happy on a personal level to have called the number so closely, and it's certainly encouraging that the first week ended up being about 17.5% higher than last year's 462,000.

Now, if that 17.5%-higher rate were to hold at the same pace for every state for the entire Open Enrollment Period, the final tally on 1/31/16 would end up being 13.7 million...higher than the HHS Dept's projection, but a full 1 million lower than the 14.7 million which I'm projecting. Bummer for me.

However, aside from the fact, as Burwell notes above, that the enrollment period started 2 weeks earlier this year (thus skewing the urgency factor a bit...more time before the 12/15 deadline presumably means more time before the numbers start to ramp up), determining whether 66% of the first batch of enrollees being renewals vs. 34% being new additions is a "good" or "bad" thing really depends on your point of view.

Last year, out of the 462K who signed up the first week, only 52% (240K) were renewals; this year it's 66% (358K). In terms of hard numbers, that's 49% more current enrollee actively renewing their policies (or choosing a different one) in the first week.

However, here's the punch line: The actual number of current enrollees who can renew is also much higher this year: Around 9.4 million (give or take) vs. around 6.3 million last year...or roughly 49% more to choose from.

Put another way, last year, 3.8% of those already enrolled actively renewed/switched during the first week. This year...3.8% of those already enrolled actively renewed/switched during the first week. Considering that, again, the "first week" is 2 weeks earlier, I'd call this a pretty good early sign.

OK, so what about the new first-week additions? Again, last year, 222,000 new folks signed up the first week (48%). This year only 185,000 new people signed up via HC.gov...17% fewer. Oh no!! Panic!!

However, let's assume that this pattern holds true for the entire open enrollment period. Last year there were 4.67 million new enrollees added via HC.gov. 83% of that would be about 3.88 million, which would extrapolate out to roughly 5.1 million nationally.

If you assume around 9 million current enrollees renewing (assuming 9.4 million or so who could potentially do so), that would add up to 14.1 million total...still 600K lower than my projection, but 400K higher than the "17.5%" scenario above.

As you can see, last year's numbers give a loose framework to make projections for this year...but only a loose one. There's a dozen new factors which could skew not just the final total but the milestones along the way higher or lower.

As for my own accuracy, it's worth remembering that last year, even though I was mostly dead on target all the way up through the 12th week of Open Enrollment, I still ultimately ended up overshooting by 6.8% in the critical final week (remember, I had bumped up my projection to 12.5 million from 12.0 million). Basically, I was expecting a whopping 2 million person surge in the last week which didn't materialize (the official 2/15/15 tally ended up being 11.4 million, plus another 300K tacked on during the week-long "overtime" period, for at total of 11.7 million).

So, what happened? Well, a big part of it may have been due to a dramatic (and horribly timed) technical outage at HealthCare.Gov on the 2nd to last day of the enrollment period which could potentially have been responsible for several hundred thousand lost enrollments. However, the larger reason is that I simply overestimated the "final surge"; it just didn't end up being nearly as dramatic as it was in late March of 2014.

In other words, this year, even if I end up being 100% accurate all the way up through January 23rd (where I'm assuming the total will be around 12.8 million), if that final week surge doesn't end up panning out, I could still end up being off by up to 1.5 million or so.