And Happy New Year to Everybody. I'll be updating comments and adding new stuff over the next few days. Hope you all had a safe and happy holiday season.
I'm just an interested generalist, so not an expert at anything but can sound credible at parties :)
"So, let’s assume that her calculations were off by 50%. We personally think the error rate is far smaller than that, but let’s use it anyway. That reduces the odds against chance from 19 billion to one all the way to 9.5 billion to one".
You're assuming the probability curve is linear. If it's logarithmic, for example, a 50% drop in accuracy could account for 99.99999% drop in probability (just making it up to make a point).
My point is that I don't think you can assume a 1:1 relationship between accuracy and probability.
Hi Mike
ReplyDeleteI'm just an interested generalist, so not an expert at anything but can sound credible at parties :)
"So, let’s assume that her calculations were off by 50%. We personally think the error rate is far smaller than that, but let’s use it anyway. That reduces the odds against chance from 19 billion to one all the way to 9.5 billion to one".
You're assuming the probability curve is linear. If it's logarithmic, for example, a 50% drop in accuracy could account for 99.99999% drop in probability (just making it up to make a point).
My point is that I don't think you can assume a 1:1 relationship between accuracy and probability.
Keep on twuckin'
Pete