I'd planned on not posting anything else today, but then this was flagged on LGM, and I concluded it could not pass without comment. Various panjandrums in North Carolina are hoping to pass a bill that outlaws various statistical methods for the prediction of sea-level change.
OK, "outlaws" is a bit strong. Scientists can still use the methods, they just can't use them in anything the state will actually officially look at. And what are these deeply questionable, pseudo-sorcerous approaches that must be refuted as the bunkum they are?
Well, let me ask you a question: what comes next in the following sequence:
1, 2, 4, 8, 16, 32, 64, 128, ...
Wait. What was that at the back? I distinctly heard someone mumbling "256".
BURN THE WITCH!
As any true North Carolinan (and possibly the Spanish Inquisition) can tell you, the answer is 192. Why? Because the difference between 64 and 128 is 64, and 128 + 64 is 192. Job done.
Sure, some people would argue the sea level is rising exponentially; that the more it rises, the faster it rises, just as with the sequence above. Those same people would argue that just looking at the previous difference and hoping it shows up again would have given you the wrong answer every single time since the sequence began.
Those kind of complaints are exactly why we need a law to make sure these gradient-obsessed malcontents are kept at arm's length. Because if there's one thing the course of human events has taught us, it's that ignoring the worst-case scenario means it can never, ever happen.