I appreciate your thoughts. I don’t think it needs a major re-working as I put enough caveats in my language there to show that this wasn’t scientific, and I was referring to another pub’s initial report.

Even Nikke Finke’s data is outdated at this point as boxofficemojo reports the total for Basterds is now up to $38.1 million. (Though the film did dip on Saturday and Sunday).

Per Nikki Finke, the box office take for the movie was HIGHEST on Friday ($14.3M), down a bit on Saturday ($12.9M) and further down on Sunday ($10.3M). Fewer people saw it each day, yet the volume of Twitter mentions kept going up…because more people total had seen it, and thus more were talking about.

Never mind causation, it appears there isn’t even a correlation between box office performance and tweets.

On top of that, I’m not even sure you’re reading the graph right. If the pink line is measuring PER DAY, then you’re talking about 3 or 4 data points total, one per day. The line that’s been drawn connecting the dots doesn’t represent data or even a trend…it’s just connecting dots. So it’s not correct to say that it “climbed through most of Saturday” before tapering off. All the pink line shows is that Saturday’s total was a lot higher than Friday’s, and Sunday’s was down a bit from Saturday’s.

Judging by the area under the green curve, volume peaked late Friday night/early Saturday as the most eager Quentin fans and other early attendees tweeted. (That’s the spike above the “u” in “22-Aug.”)

(The pattern on Saturday looks like what I would assume is a more normal tweet distribution, basically a flat bell shape – people tweet less during the US sleeping hours. But here I’m speculating.)

If I were you and your editors, I’d pull this article down and do a major rework on it, lest the headline continue to mislead and misstate the facts.

http://en.wikipedia.org/wiki/Correlation#Correlation_and_causality