2011-07-07 10:11:43H.T to Dikran for equation of the day
John Cook

john@skepticalscience...
121.222.9.229

Had to LOL at this equation from Dikran:

physics > statistics > chimps & buckets

Full context:

and15 yes, a prediction can be accurate without being based on any particular theory; however predictions made by chimps trained to pick the numbers out of a bucket can be accurate as well. The point is that predictions based on physical theory are more likely to prove accurate, at least that is what history tells us (hence the name of the series of articles).

There is a heirarchy in the strength of arguments:

physics > statistics > chimps & buckets

chimps + buckets provides no insight
statistics provdes insight into correlations
physics offers insight into causal relationships

If you have identified the causal relationships correctly, your predictions will be good, whateve the circumstances the model is used in. You can't say the same for a statistcal model based on correllations; they can only be expected to be accurate in the region of the calibration data. The "chimps & buckets" model is equally accurate in all circumstances, but not in a good way ;o)

Akasofu is somewhere between "statistics" and "chimps & buckets" (as his statistical methodology is questionable).

2011-07-07 10:13:32Also love term 'climastrology'
John Cook

john@skepticalscience...
121.222.9.229

That could be a term used throughout the Prediction Series. Wonder if sphaerica has patent pending on it :-)