When it comes to science, I think bias, in general, is not a good thing. When it comes to people and day-to-day living, I think biases are awesome and what makes the world go round, so to speak.
Without the mish-mash of biases that make up your average human being, the world would be irredeemably dull. It would be the society-level equivalent of hearing aid beige, a featureless sludge of terrifying tedium.
I like my biases.
I’m hugely biased against things like mushy peas, or brussels sprouts. I might even choose to get injected with the Bourla Bounty rather than be forced to watch a Pinter play1.
Another word for biases, in this context, would be preferences. Shock! Horror! We discriminate all the time. Our daily lives are pretty much made up of continual discrimination. Do I do the dishes, or laze on the sofa with a good book? You know what? I’m going to discriminate against the dishes.
Of course, the word discrimination has another context - one in which some prejudice (bias) is unfairly, or unjustly, applied2.
But what if the applicant looked like this?
If this, erm, thing, turned up to an interview would you be biased?
Too bloody right you would be!
The lights on your weirdo meter would be lighting up like a Christmas tree having a seizure.
I mean, anyone who wears long socks like this these days has to be considered as a weirdo. On the plus side, at least this freakish apparition isn’t wearing sandals too.
However much we try to “be kind”, to make allowances for difference, there’s that little voice inside that says fuck no, never in a month of Sundays am I going to hire that thing.
In the official report you will, of course, point out minor shortcomings in the candidate’s experience, or some other officialese to hide the fact that you’re biased against people who wear long socks.
In short, there are some biases (preferences) that are considered to be acceptable and others that are not. Having a preference for pineapple on pizza is, in my view, probably grounds for a custodial sentence, but it’s considered to be just about OK. Not hiring someone because they’re the wrong skin colour is a bias (preference) that is definitely not acceptable and ought to be subject to legal penalty3.
When it comes to science, however, bias4 is not a good thing.
Science, even theoretical science, is data driven. The theories we construct don’t come out of nowhere - they’re inspired by a desire to explain things, the world we see. Even the Greeks who, allegedly, sat around thinking deep thoughts and theorizing about how many teeth horses had, but never thought to actually check, based their ideas on their experience (observations). They just didn’t always check their new ideas and predictions against reality, or so the story goes.
If the data we base our understanding/theorizing on is biased in some way we’re going to come to the wrong conclusions, if we’re not careful and don’t attempt to deal with the bias.
Physics, perhaps, is a bit less subject to the effect of data bias than some other disciplines. Experimental physicists, in general, try to control all of the variables that might affect the results, letting only the variable they’re interested in vary. That’s the ideal, anyway.
This is possible because, again in general, physicists deal with relatively simple systems where such control is feasible.
Scientists trying to understand, for example, the factors that influence human behaviour, or the effect of some new pathogen on society, do not have such luxury. They’re dealing with trying to understand almost impossibly complex systems. The best they can do is to collect data and try to tease out some insight whilst recognizing there are going to be a whole bucketful of uncontrolled variables that may be influencing the result.
One of the unexpected benefits of the whole covid farce is that a lot of people now have a far greater understanding of all sorts of things. The absurdity of governments and their pet poodles ‘experts’ during the last 3 years have forced many people to investigate things for themselves.
We’re now aware of things like
epidemiological models tend (with very high probability) to be shite5
the whole scientific endeavour of research and publishing, particularly in the medical arena, is horribly compromised
vaccines are not universally the medical miracles they are proclaimed to be
antibodies are not the only component of our immune systems and more is often not a good correlate with better
scientific principles can be ditched at the behest of governments
any modest increase in a curve can (erroneously) be described as ‘exponential’ by the pointy heads
official health institutions and government statistical departments cannot be trusted
And lots more. It’s fascinating that the people largely derided as “anti-science” in the media and by governments were the ones behaving like actual scientists. They demanded data, and data transparency, and based their thinking on their examination of the data and the arguments.
The scientists we saw on our screens were, largely, a bunch of unquestioning muppets who refused to countenance anything other than the ‘official’ description of things - despite the data indicating their understanding was paddling up a smelly creek without the benefit of any water-moving implements.
The litany of false claims made by the ‘experts’ was staggering.
One of the things we’ve all had to come to terms with, myself included, is the notion of bias in the context of statistical data. We’ve all had to take a crash-course in basic stats and to understand why you can’t take statistical data at face value - particularly not when we’re dealing with complex systems.
Phrases and words like “correlation is not causation”, or “multivariate” now readily trip off our tongues. Would you like a helping of age-adjustment with your p-value, Sir? No problem - whilst we may not understand these things quite as well as a professional statistician, we are no longer anywhere near as ignorant and innocent as we used to be.
So how do we recognize bias? It’s not always easy, but sometimes the official bodies just make it easy for us to spot. The recently updated data on health outcomes by vaccination status published (only 7 months late) by the ONS in the UK (the Office of National Statistics) is a great example of this.
The ‘data’ presented by the ONS ‘showed’ that the MagiVaxx had powerful properties; it protected you from dying in traffic accidents
OK, it didn’t quite say that - but it did make the claim that people who had been vaxxed were dying less frequently from non-covid causes than people who had steadfastly hesitated.
7 months to produce this shite? This the best they could do to promote GOD6?
When you get a result like that from your stats, you know the bias fairy is spreading her fairy dust of derangement over everything.
Matthew Crawford makes a powerful argument that HUB (Healthy User Bias) is responsible for almost all of the alleged vaccine efficacy that has been claimed for the Bourla Bounty.
I haven’t fully understood all his argumentation in these articles (yet), and there’s still a part of me that finds it hard to accept that HUB is as widespread or wholly sufficient to explain the data, but it’s an example of a bias that must be examined.
HUB would explain what the ONS data is telling us. The people who got the MagiVaxx were, in general, healthier than their un-jabbed counterparts (according to the HUB hypothesis) and so there’s an illusion of efficacy going on. It’s an illusion that is generated by an underlying bias.
It’s fairly easy to understand why bias is a problem. Let’s take a look at the standard formula for vaccine effectiveness
The expression depends on a ratio of two things.
The hope is that the intervention reduces the death rate in the intervention group compared to the control group. If the death rate in the intervention group was 40% of that in the control group we’d find an efficacy of 0.6 (60%).
But there’s an assumption here that we need to pay careful attention to. Let’s suppose the ‘intervention’ consisted of making the treated group wear long socks (in other words, a placebo).
The assumption, not usually made explicit, is that the intervention and control populations do not differ (in a statistical sense) before the intervention is performed.
If the control group was less healthy than the intervention group (the HUB hypothesis) before the treatment then we’d see an apparent efficacy emerge. It would be entirely illusory if the intervention was a placebo.
This is why clinical trials try to ensure that the two populations (control and intervention) as are matched (statistically) as possible. Or at least that’s what they ought to be doing.
Another bias relevant to the claimed efficacy of GOD is the so-called survivor bias. It was very evident in the data from Alberta, Canazida, for example, that there was an elevated risk of a covid death in the first two weeks after vaccination. It is unlikely that Alberta is a part of the universe in which the standard laws of physics fail to operate, and so we can reasonably entertain the hypothesis that this happened elsewhere, too.
If you compare the survivors of this two-week baptism of Pfire with the unjabbed then, of course, you’re going to see an ‘efficacy’.
Lots of people have remarked on these (and other) sources of potential bias in the data that is used to determine vaccine ‘efficacy’. The problem with all these “anti-science” people (who seem to be the only people concerned with rooting out bias) is that they, somewhat inconveniently, demand full data transparency.
One very good reason for not making data fully available is if you want to hide some bias. But our governments and health institutions would never do that, would they?
They have legitimate7 reasons for hiding the data. The reasons for hiding the data are also hidden, but we can trust them can’t we?
So I’m very biased against biases in science, but very biased in favour of biases in our daily lives.
The effects would be the same. Pinter-induced myocarditis is a well-known disabling condition. At least with the jab the pain lasts only a few minutes instead of what seems like an eternity.
In today’s hiring environment this graphic illustrates the process of selecting candidates for the discard pile.
Unless the applicant is white (or Asian in some cases). Then it’s perfectly OK to discriminate to your heart’s content. Have at it, you fucking racist.
I’m not talking about a personal bias here. I have a preference for trying to figure out fundamental quantum stuff. Other scientists prefer to work on other things.
This has led to a greater awareness of the fallibility of ‘models’ in general. Not before time, in my view. Models can be great - but just like with vaccines, not all models are good.
The Goo Of Deliverance
Legitimacy is a subjective thing. No one worthy of the designation ‘scientist’ should EVER argue there is a legitimate reason for hiding/not releasing data (particularly in the field of public health). If you’re interested in maximising the financial kickbacks for promoting the Goo, then it is legitimate (your subjective legitimacy) for you to hide data that would operate against your goal.
The problem with reading Rigger first thing in the morning is that I am unlikely to find anything as wise on the internet the rest of the day. Well, in my biased opinion, anyway.
Over Here: Semantics...