Daniel Kahneman

And it was roughly 50%, 5-0.

So that's, you know, that's what made me curious about noise.

That and the fact that the company was completely unaware that it had noise.

It took them completely by surprise.

So now we're writing a book because there's a lot of noise.

So wherever a rule is that wherever there is judgment, there is noise and more of it than you think.

So that's the pattern.

Well, we call that noise is useless variability.

I mean, variability can be very useful if you have a selection mechanism and some feedback.

So evolution has built variability, and of course it's useful.

But noise among underwriters is useless.

There's nothing.

Nothing gets learned.

There's no feedback.

It's just noise, and it's costly.

The first advice, of course, would be

algorithms as i said earlier so algorithm are better than people on judgment that's not intuitive but it's really it's really true and and after that then you know the procedure that i mentioned earlier for making decisions in an orderly way by breaking it up into assessments and

That's the best that we can do.

And there is one very important aspect that I haven't mentioned, and this is training people in what the scale is.

So there is one piece of advice that you'd have for underwriters, that they should always compare the case to other cases.