Meh.

[Hank Chinaski]

Originally posted by sebastian_dangerfield
You shouldn't be very impressed. Being a doc is no big shakes. Its like being a lawyer.

We're talking past one another. I agree with your point re generall disease management based on blind stats. How else can you predict anything?

Where I get off the train is when people start worshipping stats without taking into account that most of them are not taking the adequate number of variables into account. If I say "X number of people with breast cancer will live 5 years," I'm looking at one factor - death rates. Not a very reliable stat. The statisticians haven't looked into what particular shared traits those people had which caused them to live 5 years, and what traits caused others not to live five years. They just hand people big, fat general numbers. A better analysis would avoid death rates (which are also not very much value because many people get disease are old, and die of something else soon anyway) and categorize survival by disease characteristics. Those better stats are out there, but amazingly, a lot of lazy people still rely on big general studies to select treatment and offer predictions.

But that's just health care... tip of the iceberg. Think of how much of our daily lives are manuevered with general stats (insurance, lending, securities). Stats are wonderful, but I think (and I thought this before delving into books critiquing stats) people worship them far too much. You can't card count your way through life, but people seem to really get pissed and bristle when you tell them to take "the numbers" with a grain of salt.

Stats to me are like a financial report without a cash flow stmt. You get an idea of what's going on, but you can never be really sure. The info's treated.

This is one of the special 10% of your posts; the ones I can understand and agree with.