This is a slow day, with most off the internet, as is right and proper. So I will leave us with this repetition: All models, good or bad, inexpert or Expert, ephemeral or computerized, only say what they are told to say.
This is not a bug, but a feature. Parroting their instructions is what models are supposed to do. Think: if a model about, say, blood flow in yaks suddenly decided to spout facts about interarrival times of patrons in the Library of Alexandria, you’d be right to suspect something has gone awry.
Models are sets of instructions that are like this: “If X, then Y”. That’s it. It is then no surprise if the model is fed X and Y pops out. Of course “If X then Y”: that’s in the code! Even if the path from X to Y is long and complex, so complex it is perhaps beyond the ability of the programmer to see the path.
All this means is it that is always absurd if modelers or propagandists announce, “IF X THEN Y!” as if a discovery has been made. As if Science has come upon some new thing.
Enter the JAMA article “US COVID-19 Vaccination Efforts May Have Prevented More than 1 Million Deaths, 10 Million Hospitalizations“, sent in by alert reader Stephen Shipman.
The key paragraph is this:
To estimate the effects of the vaccine program, the researchers used a computer model to analyze features of the coronavirus (of variants that emerged before the Omicron variant, including the highly infectious Delta variant), its transmission, and its effects, accounting for factors such as waning immunity and changes in population behavior over time as travel increased and schools and businesses reopened. They then compared the trajectory of infections, hospitalizations, and deaths occurring between December 12, 2020, and November 30, 2021, with those estimated under a scenario without the vaccine rollout.
All that is “If X then Y”, and nothing more.
What is the purpose of saying “computer model” instead of just “model”? In part because it sounds cooler to say the model was done on a computer, and computers are good at calculations, therefore, they want you to falsely imply, the calculations here are necessarily correct.
Of because, perhaps, of the hope it will be implied the computer has done a little thinking on its own, beyond that which happened in the modelers’ heads.
Which is, of course, impossible. An abacus doesn’t think: turning its beads into bits and plugging it in doesn’t turn the mechanics into thought. It merely makes the sliding happening faster than the eye can follow.
Anyway, nothing has been discovered here. The model was told to say the vaccine would say Y lives under condition X, it was run under condition X and said Y lives were saved.
None of this is any help at all in discovering whether the vaccine did in fact, or did in falsity, save Y lives. That’s a much harder problem, one that is even impossible, at this point, to know with any certainty.
If a guy is vexxed then gets covid and recovers, was it the vex that let him live? Or was it the weakness of the bug? Or the strength in his own body? How can you know?
If a guy is vexxed, gets covid, and dies, was it covid that killed him? Would he have lived if he didn’t get the vex, which weakened his immune system?
If a guy is not-vexxed and gets covid and lives or dies, same question.
If a guy is not-vexxed and doesn’t get covid and lives, well, we seem to have a clear cut case. Only he might have got covid and never knew it, which happened often.
The vex was rushed into use, then pushed with a fervor greater than Amway. Careful, long term testing, that would allow us to give decent answers to these questions, was never done. It’s been nothing but panic lathered with gross over-certainty since the beginning.
Buy my new book and own your enemies: Everything You Believe Is Wrong.
Subscribe or donate to support this site and its wholly independent host using credit card or PayPal click here