The Endless Needles of Niggle

*Too long for email - so you might want to open in the app or browser

What are we on now? The 8th shot? I’ve lost count. Nobody in their right mind is even remotely interested in getting this shite anymore, but it’s still being pushed by some.

I’m sorry. I got it all wrong at the beginning. I was cautiously optimistic about the vaccines when they were first announced. There were two reasons for this. Firstly, I was still in a mindset of Vaccines are Medical Miracles™ and so I thought they could prevent people from getting covid¹. Secondly, I was so thankful we had a possible way out of the nonsensical shite - things like lockdowns and masks and the rest of the excruciating idiocy we had been subjected to as a result of the CoronaDoom™.

Few things, for me, highlight the crushing cretinocracy we lived under more than this article on something that was purported to be “science”

No idea what this dumbfuckery was, but science it was not.

What were the 4 words they tested? Go. Fuck. Your. Selves?

But, hey, maybe it would be OK if we just drank something with electrolytes² in

We should never have needed this way out in the first place - but there it was anyway; a hope, a chance for us to escape from the totalitarian insanity.

I didn’t get vaccinated because (a) I wasn’t at too high a risk (I’d done my own calculations of risk based on official UK data from the ONS³) and (b) I was somewhat sceptical of a new pharmaceutical product that had been devised, tested and manufactured in something like 9 months.

Thank God I listened to those alarm bells and decided to let others be the Guinea pigs if they so desired⁴.

Since covid, or rather since governments decided to go full-cretin on us because of a mild(ish) respiratory infection for the vast majority, I’ve been in a state of simmering rage. I’ve kept it mostly in check by re-inventing myself as a snarky offensive bastard, but it’s still there.

I did nearly lose it, however, when they pushed for kids to be poisoned vaccinated.

It really did feel a bit like this

I couldn’t understand how any health agency could recommend this for kids when the NNTV (number needed to vaccinate to save 1 life) was something like 500,000 (and higher) and yet just one side effect (out of the many possible) was already known to be a concern - even at that time.

We might, possibly, have saved 1 kid for every half million injections, but we’ve given 50 of them myocarditis (a rate of 1 in 10,000 vaccinations, which may even be on the low side). What kind of fucked up “health” assessment is that?

The NNTV only makes sense when you have an effective vaccine anyway - and the evidence for that is somewhat thin on the ground.

The arguments for vaccine efficacy rumble on and we even have some US colleges, inexplicably, still requiring covid vaccination as a condition for enrolment. One can only assume some portion of their funding comes from organizations whose names begin with the letter P or M.

And here is where I find myself being niggled. There’s clearly something I’m not properly understanding about the basics of vaccines. It may be that I’m just an idiot (highly likely) or it may be that there’s something there I’ve not yet figured out. I’ve been trying for some time to find some simple ways of analysing it all - partly because I like simple stuff, partly because I’m a bit crap at proper data analysis, and partly because there’s a bit of me thinks there’s a simple way of really getting to the bottom of it all.

This re-examination was all spurred on by a recent article, but I’ll get to that later.

Homogenous Population

We’ll start off by looking at a homogenous population. Statistically speaking, there’s nothing to knock us out of kilter and so we’re ignoring things like age or wealth as possible confounders.

Now, for this kind of homogenous population, it’s kind of obvious that if you end up with a higher percentage of vaccinated deaths (from the disease the vaccine is trying to protect against) than the percentage you’ve stabbed, then the vaccine isn’t working and is, in fact, making things worse.

So let’s throw a few squiggles at this and see if our intuition matches up with the math.

What squiggles might we need to analyse this problem? Here’s the list :

• N : the total population

• r : the death rate from the disease (in the absence of the vaccine)

• g : this modifies the death rate in the vaxxed sub-group so that the vaxxed death rate is gr. If g = 0 then we have a 100% effective vax. If g = 1 we have a placebo vax. If g > 1 then, Houston, we have a problem. The parameter g can be thought of as a “death multiplier” for the vax - and we’d like that to be reducing deaths, not increasing them.

• f : the fraction of the population stabbed (f is between 0 and 1)

What are we trying to figure out? Well we need to figure out F - the fraction of disease death in the stabbed sub-group. So this is the the ratio of stabbed deaths to total death (stabbed and unstabbed).

Deaths in the stabbed and unstabbed sub-groups?

Well, there are (1 - f)N unstabbed people and fN stabbed people. To get the deaths we just multiply these by the respective death rates. So

\(D(stabbed) = fgrN\)

\(D(unstabbed)= (1-f)rN\)

The fraction of stabbed deaths is then just given by

\(F = \frac {fgrN}{fgrN + (1-f)rN}= \frac {fg}{1-f(1-g)}\)

Although this looks like a proper equation we should note that, really, we’re dealing with the properties of an average here⁵. In practice there will a statistical distribution about these mean number of deaths.

We can do a bit of algebraic jiggery pokery to re-write this as

\(F = f \left( \frac {1}{1 - (g-1)f/g} \right)\)

What are the properties of this F thing?

If we look at the bit on the bottom in the big bracket we can see that if g < 1 we’ll end up with 1 + something positive : which means that the big bracket is 1 divided by something bigger than 1. The big bracket is less than 1. This means that when g is between 0 and 1 (or equal to 1) we should have

\(F \leq f \space \space \space \space \space \space (0 \lt g \leq 1)\)

So that, at least on average, we should find that for an effective vax the fraction of stabbed death is always less than the fraction stabbed. This establishes an upper bound on the fraction F when we have an effective vaccine (remember that the closer to 0 that g is, the more effective it is - g is a “death multiplier”).

If the vax makes things worse (g > 1) then we find that F > f.

You can also establish a lower bound (I won’t go into the boring details, but it’s quite easy to do) so that we have our F function, the fraction of vaxxed deaths, lying between

\(fg \leq F \leq f \; \; \mbox { when} \: \; 0 \lt g \leq 1\)

So when you see data, as we did in the UK, that 90% of covid deaths were occurring in the stabbed population when maybe only 75% of the population had been stabbed, our alarm bells start going a bit doolally.

Non-Homogenous Population

The “get out” clause for the passionate puncturers has always been that we don’t have a homogenous population and that you can’t simply translate the results from a homogenous population to a non-homogenous one.

This is true.

But when you have over 90% of covid death occurring in the stabbed when only 75% of your population has been stabbed, the eyebrows start to head upwards a little bit at the “non-homogenous” explanation. Could it be true that we still have an effective vax with these kinds of figures?

This is the real source of my niggles - and I’ve written about it before⁶ - but it does seem that you really can find some fairly special parameter values where the vax is still (partially) effective for sub-groups despite looking terrible at an overall (total) population level. The question is whether this “non-homogenous” explanation is still tenable when we input parameter values consistent with covid.

By far and away the biggest factor here is that of age. The risk factor changes by orders of magnitude between young and old. You also have the possibility that fewer young people than old people (as a percentage of their respective population sizes) have been stabbed. But other factors, such as the hypothesised healthy vaccinee effect, might also come into play.

But we need to do a bit more than waffle. It’s not good enough to just say “the data looks bad but, mumble, mumble, mumble, old people are more at risk, mumble, mumble, mumble, healthy vaccinees, mumble, mumble, mumble, more old people have been vaxxed, therefore the vax is really effective still”.

No, you have to be able to show that the data still allows there to be significant vax efficacy when you input parameters consistent with what we knew about the risk of covid and the vax uptake percentages across the age ranges.

If you try to do an analysis by splitting the total population into “young” and “old”, as I’ve done before to try to get some insight, you end up in algebraic hell (even for me - and I love algebra).

We can do a sort of “back of the envelope” job by splitting our populations into two; young and old, where 1 denotes young and 2 denotes old. We assume g is the same for both groups and that both groups are of equal size (this helps the “it’s the older people dying that make the figures only appear to be bad” argument). We can work out our F for the total population as before and we get this ‘orrible looking thing.

\(F = \frac { f_1 g r_1 + f_2 g r_2} {(1-f_1) r_1 + f_1 g r_1 + (1-f_2) r_2 + f_2 g r_2 }\)

We can get an idea of what this might look like by letting the fraction of young vaxxed be 50%, the fraction of old vaxxed be 90% and the death rate in the old be 10,000 times that for the young⁷. With these figures (not too far off what we had for covid) and with the assumption of a very good vax so that g = 1/9 (nearly 90% effective) then plugging the numbers in and ignoring the first term in the numerator (it’s small compared to the 2nd term) we get

\(F \approx \frac{1,000} {5/9 + 2,000} \approx 1/2\)

In other words, quite a long way from anything looking like 90%

The “it’s age wot dun it” argument looks a bit thin to me.

Of course, we could settle this by looking at the data for the “old” on its own - if we ever get access to detailed enough data. I say “settle” but I’m being overly optimistic. The unbelievably cavalier approach to data collection (the deaths within 28 days of a positive test thing, for example) makes a mockery of any attempt to do an accurate analysis.

It’s infuriating - and another source of rage. How could they do this? It’s bordering on fraud and almost criminal negligence in my view (recall that we were, allegedly, in the midst of a deadly and uncertain pandemic - so the obvious thing to do was to make things even more uncertain by collecting shit data? The mind boggles).

Basically, when any analysis fundamentally relies on a measure of “covid death” you might as well just stick your finger in the air and guess⁸. The data is nowhere near fit for purpose. Absolutely sickening malfeasance, in my view.

The Excess of Hubris

And so we come to the article that prompted me to take another wander through algebraic hell.

Given the atrocious approach to data collection the best way to come at the problem is through excess death. A death is a death and it’s hard to fudge the numbers too badly.

The issue here is that excess death relies on some model - because it’s a prediction about what we should normally see vs what we actually saw. We have to have some way of calculating a decent expectation from historical figures (taking into account a changing demographic profile also). Joel Smalley has done some fantastic work in this direction and really nailed down an accurate methodology/model for figuring out excess death.

The typical way you’ll see excess death calculated is to take an average of the deaths in the previous 5 years. Although this is used a lot, it’s only a kind of “first approximation” method; probably good enough to show us possible trends but not really adequate for any truly accurate comparison. Fortunately in what follows I’m not going to need any specific model, merely the existence of some excess (however calculated).

A recent article by HART examined the issue of excess deaths since vaccination in a way I hadn’t thought about before, and it’s this I want to take a quick look at.

The scandal of excess death reporting (or lack thereof)

The bit in the article that caught my attention was the following :

To use an analogy, imagine if 90% of cars were given a new device claimed not only to be effective at preventing extra accidents but also super safe. Afterwards, there's a 10% overall rise in car accidents from 100 a month to 110. If all the extra accidents were happening in the 10% of unaltered vehicles that would mean that among this group were the 10 accidents that represent the background rate in 10% of the population plus the extra 10 accidents. That would imply the accident rate had doubled in the cars that had not been altered. If, however, the new innovation was responsible, then the 10 extra accidents would be added to the 90 background accidents in the altered group, which would make for an overall increased rate of 11%. The latter scenario is much more likely and is much easier to believe, mathematically speaking. 

The approach here is along the lines of :

OK - we’ve got excess death. What does this actually mean for the stabbed and unstabbed sub-groups?

As usual, I’m going to throw some squiggles at it and hope something useful comes out the other end.

The situation we have is that before some event (like a “pandemic”) we have a background death rate of r in our population N. Then something like the CoronaDoom™ happens. We stab the fuck out of as many people as we can and pay for Bourla’s new yacht hope that we’ve done the right thing. When the dust settles and the pandemic has put on its mask and risen from the restaurant tables of society and headed towards the toilets of insignificance, we take a peek at the death rate and find that it’s now bigger than before.

We’ve stabbed a fraction f of the population. The death rates (in a given time period) can be estimated by taking the ratio of the number of deaths to the population size. For the vaccinated sub-group, therefore, we’re looking at deaths in this group divided by the population in this group. Similarly for the unstabbed. The new (total) death rate is going to be given, then, by

\(r' = r_1f + (1-f) r_2\)

where 1 is the stabbed group and 2 is the unstabbed group, which might have different death rates.

Obviously we’d expect that once the virus has buggered off, everything will go back to normal so that r = r'. But it didn’t - and that’s a puzzle those doctory Experts™ and the MSM seem incapable of unravelling. What could possibly have changed between then and now? Dunno - hard to figure that one out.

Let’s assume, as the covid loons do, that it’s all covid wot dun it and that being unstabbed was a jolly silly thing to do. Obviously, then, the stabbed group goes back to its normal (pre-Doom) death rate and we have

\(r' = r + \Delta = r(1+p)= rf + (1-f) r_2\)

where the delta is a positive number and I’ve rewritten it in terms of a fractional adjustment to the pre-Doom death rate using the fraction p.

The increase in the unstabbed death rate can then be found (after a bit of algebra) by taking the ratio

\(\frac {r_2}{r} = 1 + \frac {p}{1-f}\)

If we use the figures in the HART example analogy we have p = 0.1 and f = 0.9 and we find that the death rate in the unstabbed group has doubled⁹.

HART chose these figures because they are roughly what we currently have - maybe 90% stabbed and an excess death of 10%. The point they make, and it’s a very important one, is now that covid has become a nothing burger, why would the death rate in the unstabbed be twice the pre-Doom rate?

It strains all credulity to suggest this is what’s actually happening.

Of course, what we actually need is a comparison of the death rates between the stabbed and unstabbed - but the Government Data Gremlins have fucked that one up for us with their shoddy approach to data collection.

One thing is clear - it isn’t the unstabbed that are (primarily) responsible for this uptick in excess death.

TL;MMMJ;DR¹⁰

Once again we can see that the case for GOD (the Goo of Deliverance) is on somewhat shaky ground. It shouldn’t be soooooooo difficult to actually see any Goo benefit; it should be repeatedly kicking us in the nuts. We shouldn’t have to be using a microscope to see it.

1. Remember, this was back in the good old days when we all thought vaccines stopped you from getting infected in the first place, rather than merely reducing symptom severity.

2. If you haven’t watched the movie Idiocracy, you won’t get this

3. This data significantly overestimates the impact of covid - but even with this exaggerated data my risk calculation came out OK for me

4. And it’s not all their fault, either. Many people were coerced Guinea pigs. I was lucky to be in a position where I could refuse the jab. I had just about enough saved to be able to say “fuck off”. Luxuries, these days, are a bit thin on the ground, but I’m not in danger of starving just yet.

5. And we also note we have to have r > 0 otherwise it all falls apart (you get 0 divided by 0)

6. I’ve written over 400 pieces on Substack. I’m sorry, I don’t have a clue which previous articles addressed this

7. This gives 70% vaccine coverage for the total (combined) population

8. It’s not quite as bad as this, but it’s not good, either. There is one thing, though. The clear exaggeration of covid death means that the crappy data we have is some kind of upper bound. Often it’s a case of even with this crappy data we can see that (a) things weren’t as bad as they made out and that (b) there is a very reasonable case to be made that the vaccines were a huge net negative

9. It’s always a good sign when the numbers work out as expected

10. Too long, more math mumbo-jumbo; didn’t read