At Heathen central, we have a large jar full of pocket change. It takes a year or so to fill up, at which time we cash it in and have a free couple hundred bucks. Usually, we use the change-counting machines in grocery stores — I mean, who’s got time to count all that change by hand?

Anyway, this morning we wondered how close an estimate of the jar’s value based only on (a) the distribution of each denomination of coin (i.e., in circulation) and (b) the known weight of each coin denomination might be, and further where we might get those data points so that we could make a guess before we turn it in. Anybody care to guess how close we’ll come?