Post by marchesarosa on Feb 25, 2011 18:50:23 GMT 1
Three experts, a geologist, a geophysicist and a geostatistician are sitting in a pub in Wales, drinking and gazing out of the window at a field full of white sheep. After a while the geologist says:
“Look – there’s a black sheep in all the white sheep. Just think – there must be black sheep in flocks all over Wales, thousands of them”.
At this comment the geophysicist snorts into his beer and says “Typical geologist. You see one data point and then extrapolate it to the whole of Wales. The only thing we can really say is that there is at least one black sheep in Wales”.
The geostatistician thinks about this for a while and then says “Actually I think we can do better than that. If we count the number of sheep in the field then we can propose a probabilistic model for the proportion of black sheep within the overall sheep population. Estimating the number of fields, we can then propose a probabilistic model of the number of black sheep”.
The three experts fall to arguing amongst themselves about their various proposals. After a while a farmer, who has been standing at the bar listening to them, puts down his beer and walks over to them.
“Excuse me” he says “I couldn’t help overhearing your conversation. Is that my dog you’re talking about?”
And the moral of the story is…wrong model.
here bishophill.squarespace.com/blog/2011/2/25/freeman-dyson-interview.html#comments
“Look – there’s a black sheep in all the white sheep. Just think – there must be black sheep in flocks all over Wales, thousands of them”.
At this comment the geophysicist snorts into his beer and says “Typical geologist. You see one data point and then extrapolate it to the whole of Wales. The only thing we can really say is that there is at least one black sheep in Wales”.
The geostatistician thinks about this for a while and then says “Actually I think we can do better than that. If we count the number of sheep in the field then we can propose a probabilistic model for the proportion of black sheep within the overall sheep population. Estimating the number of fields, we can then propose a probabilistic model of the number of black sheep”.
The three experts fall to arguing amongst themselves about their various proposals. After a while a farmer, who has been standing at the bar listening to them, puts down his beer and walks over to them.
“Excuse me” he says “I couldn’t help overhearing your conversation. Is that my dog you’re talking about?”
And the moral of the story is…wrong model.
here bishophill.squarespace.com/blog/2011/2/25/freeman-dyson-interview.html#comments