A Bayesian Interpretation of Two Deaths
Stephen “Steve” Chamberlain, 52, was struck by a car in Cambridgeshire, England, on Saturday, August 17 at around 10:10 a.m.
Less than 48 hours later, Chamberlain’s former business partner, Mike Lynch, apparently drowned when his super yacht named Bayesian sank while at anchor off the coast of Sicily.
Chamberlain and Lynch were co-defendants in a U.S. trial over their tech company Autonomy.
Both men were accused of fraud and conspiracy over allegedly inflating their company’s earnings to secure an $11 billion deal with Hewlett-Packard. Both were acquitted of all 15 charges in June by a San Francisco jury.
The Bayesian was named after a statistical method for assessing probability, formulated by the 18th century British mathematician and presbyterian minister, Thomas Bayes.
Bayes’ Theorem is a mathematical formula for calculating conditional probability—that is, the likelihood of an outcome occurring based on a previous outcome in similar circumstances. Bayes’ Theorem is tool for updating existing probability assessments given new or additional evidence.
A probability assessment based entirely on the frequency or propensity of cyclists getting struck by cars in Cambridgeshire and modern super yachts sinking in freak weather events in the Med would probably find that the chance of these two men randomly dying in these unusual ways within 48 hours of each other to be less than 1 in a billion. I’ll leave it to pro statisticians to compile the data and do the proper math on this one.
However, if investigators find grounds for suspecting Steve Chamberlain’s death was no accident, but staged to look like one, this could enable investigators to perform a Bayesian interpretation of the probability of Mike Lynch’s yacht sinking 48 hours later.
If it is determined that someone had an interest in eliminating Steve Chamberlain, this raises the probability that Lynch’s subsequent death was related to Chamberlain’s and was also no accident.
A Bayesian interpretation performed at the time Chamberlain was struck could have warranted alerting Mike Lynch (and his guests on the yacht) that a their lives were in danger and that extraordinary security measures should be taken to secure the vessel.
I hasten to add that the above reflections are more a matter of intuition than a proper understanding of how to perform a Bayesian probability assessment. I welcome the pro statisticians among our readers to chime in with their expertise.
See more here Substack
Please Donate Below To Support Our Ongoing Work To Defend The Scientific Method
PRINCIPIA SCIENTIFIC INTERNATIONAL, legally registered in the UK as a company 
incorporated for charitable purposes. Head Office: 27 Old Gloucester Street, London WC1N 3AX.
Trackback from your site.
Charles Higley
| #
So, considering that such coincidences do not occur, was it really a tornado (water spout) that took out a large yacht? What was it really? There appear to be no witnesses on other craft. What about interviewing the survivors, who might, unfortunately also be part of or coerced into supporting or covering up the crime?
Reply
Ken Hughes
| #
My feelings exactly on reading the news article. This double event is so unlikely, it has to have more connection than just the trial. The whole thing stinks. It seems highly suspicious we are given no information on the car and driver that killed the cyclist. Was it a hit and run? If it was, then the likelihood of a double murder increases dramatically. I asked the key question in my comments online, “Are the police investigating?” I wonder, are they? If not, they should be.
I have some training is statistics but not is Bayesian methods and many years ago. But I do know that the probability of two events happening, is the product of the two, (or more), independently happening. What are the chances of his dying in his yacht from a “freak” water spout, pretty damn low in the first place. What are the chances of the cyclist dying from a car accident at the same time? That is simply the number of cyclist deaths per year divided by the number of cyclist trips per years in total in the UK and this is also very low. What is the probability of these two unlikely events happening together when also linked by the trial? This is also obviously very low indeed.
Let’s assume all three have a probability of one in a million. I don’t think that is unreasonable. So the probability of all three happening is one millionth times one millionth times one millionth. That is ten to the power of minus eighteen.
You stand less chance standing in the middle of a field with a cigarette in your mouth and for a lightening bolt to light it.
That’s without the Bayesian analysis.
Someone should ask professor Norman Fenton.
Reply
James
| #
Probabilities of a cyclist getting hit are quite high, it happens all the time. Of a yacht sinking because the hatches were open and the keel lifted, also. The Med’s not as safe as it looks.
Reply
Herb Rose
| #
I’m confused. The two were acquitted on all charges so this is a motive to have them”murdered” so they could not face any civil suits like O.J. did?
Reply