Not the OP but it's probably a reference to the sudden (?) slippage of effectiveness definitions visible from the start of COVID. There are at least four but ma

Not the OP but it's probably a reference to the sudden (?) slippage of effectiveness definitions visible from the start of COVID. There are at least four but maybe more: 1. Redefining effectiveness to mean the generation of proxy bio-markers, not improved health outcomes. COVID vaccines and treatments are now routinely being waved through trials because they generated antibodies, even if those antibodies don't actually reduce infection, sickness or death (e.g. because they're generated against the wrong version of the spike protein). 2. The Pfizer COVID trial had more deaths in the vaccine arm than the placebo arm i.e. negative effectiveness against death. This was ignored on the basis that the difference wasn't statistically significant. That's a logically incorrect way to use the concept of statistical significance, but a remarkably common problem in science. If more people who take the vaccine die than those who don't, this might be just a fluke if the numbers are small, but it might also be a real effect. The outcome you'd expect from this situation is to gather a larger sample to ensure that you won't accidentally kill more people than you save, but, public health people don't think like that. Instead they said, we can't be sure the problem is real, so we'll assume it isn't. 3. Post-trial, COVID vaccine effectiveness was computed by public health agencies using statistically invalid methodologies. There are at least two sub-problems: 3a. Effectiveness numbers were not the ratio of raw case rates split by vaccine status. They don't like these numbers because they fear bias, e.g. if people who refuse vaccines lead generally healthier lives, it would make the vaccines look bad. So they adjust the numbers, normally using test negative case control study designs which compare test positivity ratios instead of case numbers. Unfortunately this approach requires that people don't control whether they get tested or not, a condition that is clearly very much violated in the case of COVID testing. Vaccinated people are far more likely to be concerned by COVID and thus far more likely to be testing themselves regularly, there were also of course the various testing and vaccination mandates to contend with. In such a scenario the methodology artificially increases perceived effectiveness by massive amounts e.g. -400% measured effectiveness was "bias adjusted" up to +50% effectiveness in the UK. 3b. People aren't classified as vaccinated until some time after they actually get the shot. This should cause all stats to split into at least three categories: unvaccinated, vaccinated pending, vaccinated activated but they don't do this. Instead people are described as unvaccinated even if they aren't. This creates a form of time bias due to slippage in the denominator - if you take the vaccine and get sick during the 'pending period' this gets allocated incorrectly to the unvaccinated bucket but if you don't, they count that time without infection as a benefit for the vaccines. You could make water look temporarily effective with this approach. It gets worse if the vaccine actually makes you more likely to get infected, as in this case the methodology will turn negative effectiveness into positive effectiveness. The bias introduced by this approach wears off with time (unless you get boosted and it begins again of course). It's possible that this is part of the explanation for why real world measured effectiveness numbers started falling almost immediately from their initial highs. 4. Public health agencies simply made up some claims of effectiveness, e.g. the claim that vaccines would stop the virus in its tracks by blocking transmission was entirely made up, with no scientific basis whatsoever. It was done to manipulate people into taking it even if they weren't in any risk group. When asked why they told people the vaccines blocked transmission, the answer from Dr. Debora Birx was "I think it was hope". There's the other related issue that effectiveness alone is not an interesting number. Only the effectiveness:safety:cost balance is interesting. Cyanide is a cheap and 100% effective vaccine for any known disease, if you ignore adverse events. There is a lot of evidence that COVID vaccine AEs are ignored entirely by the entire public health system. For example, the US CDC promised at the start it would conduct weekly reviews of the VAERS data to detect a "safety signal". Side effect reports went vertical, such that COVID vaccines have far more AE reports than all other vaccines combined. Later on in response to a FOIA request they claimed they'd actually never done any analysis or data mining of VAERS, and didn't consider that to be their responsibility in the first place. Given this bold new regulatory environment, claims of effectiveness for vaccines must be taken with a large helping of salt. They use techniques that can and do convert data showing a vaccine backfiring into data showing a great success.