Yesterday, while I was enjoying a cup of coffee at one of the allegedly hipster coffeehouses I frequent, the woman at the next table struck up a conversation. She and her husband had just returned from a bike ride about ten miles up the coast. I objected that the bike paths were closed by order of our scary-looking county public health official and assorted lesser authorities. I was informed that they were now open, whereupon I set off to explore.
Indeed, it was true! I was overjoyed. Everyone was delighting in a beautiful day by the beach, reveling in their new-found freedom. I was grateful to be able to resume some normal activities. Only one day before, there were orange barricades blocking entry and at many points along this path, and rent-a-cops enforcing the no-go zone. [click on photo to enlarge]... [Read More]
In 2005, epidemiologist and physician John Ioannidis published the controversial and famous paper, Why Most Published Research Findings Are False in PLOS Medicine. He showed there were sound reasons to be skeptical of most published research, especially those results that rely on statistical analysis to draw conclusions about the relationship between two phenomena. Familiar examples:
• How effective is a drug?
• Does a substance in the environment increase the incidence of cancer?
• What is health effect of a food?
As he suggests, the science in most of these cases is far from settled.
By 2014, Ioannidis’s fame had grown to pop-science level. He was invited to do one of the Talks at Google, where he was given a fawning introduction by the earnest and callow member of the “People Operations Team.” He can do no wrong.... [Read More]
A Port of Seattle police officer posted a video on YouTube on May 5 decrying the enforcement of “tyrannical orders against the people.” It was primarily directed at other police officers and was initially praised by his superiors. Within three hours, they contacted him again asking him to remove it, citing higher authorities. He refused and was threatened with sanctions (“documents will be found”). He was fired. Independent of the subsequent controversy, his original video is worth a look. This is a re-post of the original video, which he claims had over a million views and much support from other law enforcement officers. He raises the issues of the consent of the governed and of the dangers for LEOs under the loss of public trust.
Every evening, people gather by the beach near me to enjoy the sunset. The mini-Mussolinis in my town have decided to make it illegal to stand on the ocean side of the street and have closed off the sidewalk. For a few days the police were leaving folks alone to stray across the street but last evening they drove down the street yelling at the citizens to get out. Tonight, the people came back and the police have backed off. We had our own tiny version of the protests seen across the land. This is a view from the (forbidden) ocean side of the street looking inland.
Thousands of cars, trucks, and commercial vehicles joined an #OperationGridlock protest organized by a conservative group called the Michigan Conservative Coalition Wednesday, blocking streets in Michigan’s state capital, Lansing. The march is a demonstration against the state’s coronavirus quarantine “Stay Home” order.... [Read More]
More than 60,000 Americans died. About 45 million were infected and 800,000 were hospitalized. News coverage of the event: crickets. Of course, way back then we didn’t have the careful, dare I say heroic, news media we have today. And to be fair, they were busy with more important things like Russian collusion and OMB* to bother covering the flu season of 2017-18.
The overall burden of influenza for the 2017-2018 season was an estimated 45 million influenza illnesses, 21 million influenza-associated medical visits, 810,000 influenza-related hospitalizations, and 61,000 influenza-associated deaths.†... [Read More]
Scientific models can be useful in understanding complex phenomena and to make predictions. All is well so long as everyone, modelers and users, understand the limitations of the model. It all goes wrong when the modelers become enamored of their models or the users treat the model as gospel, or both.
This drama is playing out on a global scale under the WuFlu panic of 2020.* Imperial College (London) epidemiologist Neil Ferguson predicted† about 500,000 deaths in Great Britain, which was scaled to 2.2 million deaths in the US. The UK and US political leadership were using this model to make policy decisions on March 16:... [Read More]
Models are judged by their predictive skill. During today’s COVID-19 press briefing,* Dr. Birx presented the latest modeling that is informing decisions at the federal level. The model makes specific predictions about the number of deaths expected under the current “full mitigation” scheme. Within about two weeks, deaths are predicted to peak at 2.2k per day and it should be clear if the model has predictive value or not. The advantage to using deaths instead of confirmed cases is that the latter are subject to testing bias; corpses are easier to count.
This graph plots the number of deaths per day (vertical scale: each division is 500) versus date (February 1 through August 1). Click on the graph for a full size version. The area under the graph is the cumulative (total) number of expected deaths. The dashed curve is labeled as “projected,” which I take to be the principal prediction within the wider band of uncertainty. The projected curve is approximated by
\(deaths per day=120(x-3.5)^4 exp[ -4(x-3.5) ]\).
The total number of deaths obtained by integrating under the curve is about 90k, which is in good agreement with an eyeball estimate. The lower and upper bounds are 40k and 140k, respectively.... [Read More]