Ooops sorry. I forgot that I had restricted access to the site. It will be OK from the link here now.

The increase in postive rates at the start of October will have been caused by the mass testing of students across the country where in some areas they were reporting 30% +ve rates among this cohort.

Ooops sorry.

it works now,
looking here, when the plot touches the baseline early July.

1)there were 94284 tests on july 6, and 343 positive(0.34%)
2)and on July 14, 132121 test with 398 positive(0.30%)
3)on 21st, 142559 with 445 positive(0.31%)

It never gets lower, hence that should be close to the false positive rate of covid-PCR-test if we assume there was little or no real covid-19 at that time as Mike Yeadon and others suggest.

Then the false positive rate in these circumstances of covid-PCR-test is 0.3%.

By a remarkable coincidence, there is a match to this!

Cambridge University test their students with the covid-PCR-test. They batch them into "pools" for efficiency. They mix all the swabs from students in the same house together and just test that. In the rare event that the test is positive, they test the individuals to find who amongst them has covid19.

Anyway, in doing this they found that In total, across nine weeks, 0.3% of pool tests conducted produced a positive result that was then followed by negative results in individual follow-up tests. In other words, the test was declaring a false positive result in 0.3% of tests done. (No need to talk about negative and real positives. Irrelevant to my case.)

So there you are, without really trying and by accident, your plot seems likely to have fished out the exact same fpr (false positive rate) that Cambridge found. Here it is on full fact:

https://fullfact.org/online/cambridge-pcr-false-positive/

It also lends credit to Mike Yeadon, by suggesting there really was hardly any covid-19 around in early July, he was right, judging by this.

BTW: this is a much smaller false positive result ratio than many sceptics would accept. I expect this post to cause a bit of a flare up, but there it is; I cannot argue with my own finding, although I almost wished the fpr were higher, since it would mean we more right, if you see what I mean. But, up the up side, at times of low prevalence, even 0.3% fpr can create a dramatic distortion. enough, already!!! |So that's it - the operational false positive ratio is finally revealed.

Cambridge University test their students with the covid-PCR-test.

On 14 December, I posted a comparison of the recent Cambridge tests, the NHS tests from the summer and the private labs tests from September onwards here https://forums.lockdownsceptics.org/viewtopic.php?p=6863#p6863

The upshot is that the private labs cause 'cases' to increase by a factor of about 8, the NHS labs by about 1.7 and the Cambridge labs by 1.13.

On 14 December, I posted a comparison of the recent Cambridge tests, the NHS tests from the summer and the private labs tests

you findings are not a million miles from the rough estimate above, so 0.3% fpr is possible to attain at e.g. Cambridge where they take pride in their work, you say the mass testing is much poorer. That is where to attack covid-19 PCR test; on the various procedures used.