angelzen

Update: the error belongs, indeed, to the RKI crew: https://twitter.com/Tim_Roehn/status/1477975486764572675

Via Google translate: "Update: The RKI made a significant change in the weekly report 4 days after publication (see below). Now there are 1097 instead of 186 unvaccinated people, so the proportion of those who have been vaccinated is significantly lower. // @world"

Edit: Update of the OA: https://twitter.com/holmenkollin/status/1477970997638676483

Interesting.

* 30.12.2021 RKI Wochenbericht reads "1.097 Patientinnen und Patienten waren ungeimpft, 4.020 waren vollständig geimpft, von diesen wurde für 1.137 eine Auffrischimpfung angegeben."

* 23.12.2021 RKI Wockenbericht reads "186 Patientinnen und Patienten waren ungeimpft, 924 waren vollständig geimpft, von diesen wurde für 161 eine Auffrischimpfung angegeben".

* The OA screenshots a line that reads "186 Patientinnen und Patienten waren ungeimpft, 4.020 waren vollständig geimpft, von diesen wurde für 1.137 eine Auffrischimpfung angegeben." Either they doctored the data, or the 30.12.2021 RKI Wochenbericht perhaps contained an error that was since corrected and the OA panicked seeing the initial version?

One more thing. Using omicron data from 30.12.2021 RKI Wochenbericht and vaccination data from https://impfdashboard.de/en, the omicron incidence rates don't look too good. This is difficult to do correctly in absence of data explicitly broken down by age because <18yo have significantly smaller vaccination rates (thus more "unvaccinated" at virtually no risk) compared to the adult population, but even so:

           total  unvax    vax  2-vax  3-vax
    pop    83.2M  21.5M  61.7M  26.9M  32.3M
    cases   5117   1097   4020   2883   1137
    /100k    6.1    5.1    6.5   10.7    3.5

Similar figures from Denmark, reported as rates and broke down by age groups. Waiting for severe omicron cases stats, which may lag for a few weeks. We are in for an interesting winter...

https://covid19danmark.dk/#gennembrudsinfektioner

https://files.ssi.dk/covid19/omikron/statusrapport/rapport-o...

In a free society the job of the media is to report the facts, not to "keep the raving mad in check".

This leads to 2 twitters, one for each side of the political aisle. Filter bubbles enabled by social media are already bad enough, a complete fracture of the public space along political lines is a distinctly dangerous idea.

Since you are not posting your sources, it's hard to tell what claims of Marjorie Taylor Greene you are referring to. I recently stumbled upon a site that reports on covid data from Denmark. The rate of Omicron infection in both vaccinated and unvaccinated populations >15yo is the same. This corroborates with a recent DHH post surfacing similar Denmark data. There is a distinct uncomfortable possibility that the vaccines targeting a 2 year old variant are, indeed, ineffective as a measure to limit the spread of Omicron. Unfortunately, I can't corroborate with US data because CDC is obstinate in not publishing covid data with a breakdown by date/age/vaccination status/virus variant.

Based in prior variant data, it is likely the vaccines do work to prevent severe omicron cases. Feel free to clarify what is it that the vaccines work for, and which of Marjorie Taylor Greene tweets fell afoul of that.

PS. This post is in no way an endorsement of Marjorie Taylor Greene political activities, of which I know very little other than her being a Republican Congresswoman.

https://covid19danmark.dk/#gennembrudsinfektioner

https://news.ycombinator.com/item?id=29765351

https://world.hey.com/dhh/should-you-vaccinate-your-kids-169...

(2nd attempt, realized the outlier spike is in "unvaccinated and 80+" category). Utter garbage data. Using the breakdown by age we note a huge outlier spike in unvaccinated cases in the 80+ age group, 800/100k vs. 5-10/100k. All other age groups show a decline in case rates for both vaccinated and unvaccinated, with a vaccinated case rate << unvaccinated case rate (5x-10x). I see no way to add up the breakdown by age case rate numbers into total case rate numbers showing vaccinated case rate taking over unvaccinated case rate.

Couple of related notes:

* Covid statistics without a breakdown by age / date / vaccination status are garbage. Possibly BMI and/or immunosuppressed status should also be added to the mix. The authorities failure to publish this data breakdown 2 years into the pandemic is going beyond incompetence.

* Focusing on "cases" beyond a coarse "going up or going down" is largely useless. Observed "cases" are more likely to indicate test penetration than actual case numbers, and test penetration may significantly vary between age groups, institutionalization settings, etc.

Not a big fan of nuclear because of security and waste disposal issues, but quoting 3 accidents over a span of 60 years as a reason for a blanket ban on a whole class of energy production technologies is hyper safetyism.

Perhaps this Twitter thread (sorry) may help alleviate some of the fears? The author is a full professor working on computational biology at University College London and had prior to that had a stint working on infectious disease epidemiology at Imperial College London.

https://twitter.com/BallouxFrancois/status/14765074434966364...

"Those results provide no evidence for SARS-CoV-2 being a neurotropic virus, it doesn't inform us on long-covid. SARS-CoV-2 is not expected to be commonly found in the brain of patients except in a subset of extremely severe, often lethal, infections. 3/"

https://www.ucl.ac.uk/biosciences/people/francois-balloux

Your note on misuse of VAERS data caught my attention: "deaths happening shortly after vaccinations, but not necessarily connected to it". Apparently the same kind of sleigh of hand is widespread practice in mainstream epidemiological data collection, according to recent comments by none other than Dr. Fauci.

> "But the other important thing is that if you look at the children who are hospitalized, many of them are hospitalized with COVID as opposed to because of COVID," Fauci continued. "And what we mean by that — if a child goes in the hospital, they automatically get tested for COVID. And they get counted as a COVID-hospitalized individual. When in fact, they may go in for a broken leg or appendicitis or something like that. So it’s overcounting the number of children who are, quote, 'hospitalized with COVID,' as opposed to because of COVID."

https://www.washingtonexaminer.com/policy/healthcare/broken-...

Indeed. A few more examples of fundamental systems advances since 2000: iOS, Android (intents), Chrome (multiprocess), Kubernetes, Cloudflare Durable Objects (collaboration), WASM, Rust, Typescript, Golang, Pytorch, JAX, Bitcoin, Ethereum.

"Python is basically slowly becoming a TypeScript." If only.

"We've measured the position of iron balls thrown at different speeds from a 56m tall tower and found they were 90% likely to be found flying in the air."

This is technically correct, yet fundamentally misleading. The same statement, bracketed by time since triggering event:

"We've measured the position of iron balls thrown at different speeds from a 56m tall tower, 0.1 seconds after throwing, and found they were 90% likely to be found flying in the air."

Even better, show the plot of position by time, or at least clarify you are merely doing a point-in-time measurement. Time is a fundamental component of the vaccine effectiveness equation.

Given that boosters are an attempt to counteract waning vaccine vaccine effectiveness, there are two fundamental questions:

1. Do boosters work at all?

2. Do boosters wane over time, just like the original shots did?

They answer only 1. They didn't even attempt to shed light on 2, which is a serious letdown. Given known waning vaccine effectiveness, at a minimum claims on booster effectiveness should be bracketed by time since boosting event.

From the paper I originally linked: "Infection with one serotype likely elicits lifelong immunity to that serotype, but generally not against the other three." Dengue researchers routinely talk about lifelong original antigenic sin effects in dengue. You seem to want to split hairs and ask instead "but which papers specifically established the duration of the original antigenic sin in dengue". I can't help you with that beyond linking to the seminal Halstead 1983 paper. From there on you are on your own.

https://pubmed.ncbi.nlm.nih.gov/6824120/

Long-lasting original antigenic sin is well documented at least for dengue.

"Original antigenic sin has the advantage that a response can be rapidly mobilized from memory. However, the downside is that in some cases, such as dengue, the response is dominated by inferior-quality antibody. In influenza, original antigenic sin has been shown to reduce the effectiveness of vaccination (13, 34, 51). In dengue, the effect of original antigenic sin has considerable bearing on vaccine strategies. Once a response has been established, it is unlikely that repeat boosting will be able to change its scope, meaning that balanced responses against the four virus serotypes will need to be established with the first vaccine dose."

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3014204

Network effects. As long as the other kids socialize on social media, not using social media becomes a drastic socialization penalty. The problem cannot be solved individually, only by collective action. Most parents must ban social media (and electronics in general) for their kids. Which indicates a need to pass laws to ban social media for minors. Bans and teens don't usually work as intended, we probably also need a wide campaign of stigmatization of said social media, akin to anti-smoking campaigns.

The point of 1/ 2/ 3/ ... threads is to enable out of context deep linking. They are not going away, lest Twitter becomes yet another run-of-the mill blog platform nobody engages with.

old.twitter.com, I hope.

Not really. Team X uses Java? Someone on the team takes 1 hour to create YamlPlus.java, submits to the monorepo. Now team X + all the other 100(0) teams that use Java have access to a yaml+ parser. Same with Golang, JS, Python, C++, cli check, presubmit check, and we are running out of supported languages.

The argument that no incremental change is ever possible because it has to update every related legacy usage is a recipe for organizational paralysis.

Coding is values, arrays, dicts, ifs, loops and functions. An expressive set-like operations library over arrays (& dicts) gives superpowers.

Overreaction? Let's hypothetically consider yaml+, with the definition:

  yaml+ = yaml "\n" "."

* easy to wrap a yaml parser and create a yaml+ validating parser.

* easy to create a simple script to validate the end-of-file dot line.

* easy to create a presubmit check to validate yaml+

* easy for an editor to syntax check the end-of-file dot line.

Assuming there are other redeeming qualities of YAML, the truncated YAML problem can be solved with minimal effort.

The vaccine is not the virus. If you take a stochastical approach to understanding immunity, that raises questions whether one can be substituted for the other in the long term. Derek Lowe has a good elementary introduction in the subject.

"It's important to realize, though, that all immune responses to a viral infection generate a mixture of neutralizing and non-neutralizing antibodies. That's one of the things about the immune system - it revs up production of a wide variety of antibodies, selected from the untold billions of them circulating around in your bloodstream. Some of them bind to one part of the pathogen, and some to another. And they bind in different conformations, sticking to different parts of the surface of the invading virus from different directions."

https://www.science.org/content/blog-post/antibody-dependent...

Vaccine mandates enforced by heavy fines and/or prison are criminalizing being unvaccinated with the number of "boosters" the authorities deem necessary. Justified by group "crime" statistic differentials. We may argue that it is done in service of a good cause, but it is still criminalizing. Our weird reality.

This is probably the right course of action. However, it is unfeasible to derive a fair tax level in the current climate. Too many people look to force conformity and/or exact punishment, instead of engaging in a good faith tradeoff negotiation.

True, unvaccinated people are being hospitalized at a higher rate and for a longer duration.

However, we are not talking about "the unvaccinated", rather about "a tiny unfortunate fraction of unvaccinated people". There are 4,690 covid patients in ICU beds and 25M unvaccinated people. Assuming all covid ICU patients are unvaccinated, the covid ICU incidence rate within the unvaccinated population is 0.02%.

For an analogy, we could reason that because the prison population skews 90/10 male/female, all males should either undergo a vasectomy, or pay a recurring 10k euro fine. I hope this is obviously unethical.

True. Thank you for acknowledging the point and adding further qualifications. Public discourse is rather disheartening. The options are either 100% vaccination or 0% vaccination, either strict house confinement or nightly clubbing, either masking while hiking in the woods or no masks ever. Life is a rich tradeoff of a myriad factors, somehow we reduced it to a handful of black/white binaries.

As I already replied elsewhere in this thread, as of 12/02/2021 German CDC estimates 7-day R-value as 0.92. That is below 1.

https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus...

As of 12/02/2012, Robert Koch Institute, roughly the German CDC, estimates 7-day R-value at 0.92

https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus...

Index page: https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus...

The vaccines, while not as effective as classics like polio vaccine or measles vaccine, are definitely a net positive and help significantly in reducing the epidemic burden. Perhaps instead of internecine finger pointing we could blame covid itself?