• by Erik Sass , Staff Writer @eriksass1, January 11, 2013

Say what you want about the New York Post, it doesn’t shy away from a fight. New York City’s newspaper of discord appears to be locked in an epic statistical pissing match with Facebook over the number of active users at Instagram (note to self: try to fit “epic statistical pissing match” into more conversations).

The NYP previously reported that data from AppStats showed Instagram’s active daily users dropping by about a quarter following its terms of service privacy fiasco, from 16.4 million on December 19 to 12.4 million on December 26. Facebook fired back that the data was “inaccurate,” but declined to specify how or why this was the case.

Now the NYP says that according to AppStat the number of active daily users has actually declined by about half, from 16.35 million on December 17, when the TOS change became public, to 8.42 million this week. In case anyone has trouble connecting the dots, the NYP also quotes AppStats CEO Sebastian Sujka: “The main loss will be most likely due to the terms of service changes, given how much attention and controversy the terms of service change has brought, and seeing how clearly the Instagram app dropped after the terms of service change.”

And this is where we come to the maddening thing about Web and mobile measurement: a set of numbers might seem to clearly suggest some sort of trend, but there’s always the possibility that they’re wrong. And not just a little wrong, or misinterpreted or massaged -- they’re somehow, like, diametrically opposed to reality according to another interested party (usually the company which stands to lose by the negative publicity).

Thus, Facebook insists that “We continue to see strong and steady growth in both registered and active users of Instagram.”Another good example of this is Google’s ongoing dispute with various Web measurement firms about how much traffic there is on Google+: back in April 2012 Google CEO Larry Page said he was excited about “impressive growth” at Google+, while comScore data showed that users only spent around three minutes per month on the network (which in fact isn’t a social network at all, according to Google execs -- but I’ll leave that for another time).

Anyway, I think the problem is that the science of Web and mobile measurement is still so new that anyone who disagrees with numbers can simply call “methodology” and claim to have won the argument. It would be nice to know if the number of active daily users at Instagram dropped by half following the privacy uproar -- for one thing, it would be a good warning to Facebook and other social media sites that, despite appearances, consumers do care about privacy. But instead we’re left with speculation… or not even that, actually, since Facebook has now blocked access to Instagram’s user figures via its API.