Let me see the death rate and causes for other ex-microbiologists.
We do have that information. The actuarial data for life insurance, and more specifically the morbidity data, would tell us whether microbiologists are at a significantly, insurable, higher risk of death than the general population. I cannot find any evidence that this is the case by conducting a web search; someone please point me to the data if there is evidence for the opposite. Note that I take insurability as a sufficient filter for patterns that rise above background noise; you are free to disagree, but then the onus of proving how insurance companies can base a business model around such a statistical model falls upon you.
On 5/5/02 11:02 AM Pacific, The Great Satan said:
The problem is that no adjustment is made for the vast array of equally improbable and retrospectively striking coincidences that could occur in the universe of observables.
The climatic conditions that affect Portland, Maine weather and subsequent chances of hurricane weather developing there are relatively unique to that city. This is not to say that coincidence is possible: that's why we are talking probabilities. Dismissing the prima facie facts on the ground with the statistical equivalent of hand waving by saying "there is a ton of coincidences that could happen in this universe" without listing the possible coincidences and attempting to quantify them, when presented with a quantified estimate, does not bolster a refutation of the estimate.
The source article and a simple web search will give plenty of details of the circumstances surrounding each of the deaths, and provide plenty of data points to fill the "vast array of equally improbable and retrospectively striking coincidences that could occur in the universe of observables". If the observable event space that could affect this situation is as large as you claim, it should be easy to construct the apparent coincidental event space after the deaths have occurred. I'm not saying you are wrong; I'm just asking you to provide a more analytically rigorous rebuttal. I'm more than willing to admit the estimation is incorrect and learn from this; this is how we advance ourselves.
On 5/5/02 11:27 AM Pacific, Nogbad said:
Just goes to show statistics can be used to prove or disprove anything.
Please see my response to The Great Satan above. Again, I'm willing to concede that the estimation is incorrect, if someone provides an equally quantified statistical analysis that refutes the estimation. The mathematical equivalent of blustering and rhetorical platitudes however, are hardly persuasive against objectively constructed analysis.
On 5/5/02 2:19 PM Pacific, Seeking the truth said:
Thanks for putting a number on it.
On 5/5/02 2:19 PM Pacific, Mad_Tom_Rackham said:
Good Job! You've explained it very well.
All credit is due to the original poster and analyst on Slashdot, I'm merely relaying what I read, and defending its applicability until shown data or an objective and equally quantified analysis that indicates otherwise.