Regression to the Mean

by Andrew Porter

If something varies normally between two far extremes,
It usually swings back naturally to values in between.
From sport to crime and illnesses, we see this common theme,
An effect we call, statistically, 'regression to the mean'.

When Brucie plays his cards right and he's holding up a queen,
You know that next a lower card is likely to be seen,
Because there are so many cards much lower than a queen.
It's simple probability, it's 'regression to the mean'.

Random fluctuations of performances in sports,
Befuddle sports professionals, who use gimmicks of all sorts,
Crystals, magnets, copper bracelets they esteem,
But improvements in achievement are 'regression to the mean'.

A man with awful backache, that sometimes gets much worse,
May turn to herbal remedies and swear his pain's reversed.
Perhaps it has, but not because the herbalist intervened,
The pain will ease quite simply through 'regression to the mean'.

Evidence-based treatments that doctors should assign,
Use tests that will be randomized, controlled and double-blind,
Stopping self-deception before those test-results are seen,
It stops them being fooled by 'regression to the mean'.

Big international drug firms are rightly criticised,
For concealing contrary evidence and other sorts of lies,
So how come homeopathy and herbalists have been,
Permitted to bamboozle us with ‘regression to the mean'’?

While scientific medicine cannot cure all ills,
It does do so much better than those homeopathic sugar pills,
That a child should die through want of proper treatment is obscene,
When wishful-thinking parents fall for ‘regression to the mean'’.

When much that was untreatable is curable, it’s tragic
That patients turn from medicine to phony cure-all magic.
The alternative to thinking is delusionary dreams,
Alas, from death there will be no ‘regression to the mean'’.