Kottke has a discussion here (a year ago, but roll with it) on the number of Starbucks within a five-mile radius of any given point. Bash your postcode into that store locator and see what you get – in my case, a maybe-the-world-hasn’t-gone-mad-after-all ‘8’, but there are places in the world where the number rises to over 160. No, I didn’t add an extra nought by accident.

Next up, some chap called Cory has calculated the centre of gravity of all Starbucks in Manhattan. That is, he’s found the spot that’s equidistant from all Starbucks franchises on Manhattan island (between 5th & 6th and 39th and 40th, since you’re wondering). At least, I think that’s what he’s done – the methodology isn’t entirely clear.

See, I’m wondering if Starbucks averaging should be calculated by a straightforward distance measurement, or if what you should really be considering is coffee flux – which presumably follows the usual inverse-square law? In which case, the centre is likely shifted. Indeed, there may even be several zero-points distributed over Manhattan island.

This, dear reader, is a breakthrough concept.

Coffee flux appears to be a vector quantity, with the initially-surprising property that Starbucks stores are – by inspection – sinks rather than sources. Further, coffee flux appears to interact with humans, drawing them along flux vectors towards the aforementioned franchises.

So if we can calculate the local minima for Starbucks coffee flux, based on a (presumed) inverse-square distance function from each sink point, we will have found positions of zero net Starbucks coffee flux. Effectively, Starbucks Lagrange points, where the pull of each nearby Starbucks store exactly balances that of all others.

I present the hypothesis that such coffee flux minima would be excellent locations to establish a chain of tea shops. Now: how do I present this as a business plan?