The ‘green batches’ clustered around the green line have a moderate or moderately-high level of adverse events associated with them. In the discussion with Preradovic, Gerald Dyker takes the example of the green point furthest to the right.

As he explains, it represents the batch that was the used the most in Denmark, with somewhat over 800,000 doses having been administered. These 800,000 doses are associated with around 2,000 suspected adverse events, which gives a reporting rate of one suspected adverse event per approximately 400 doses. As Dyker puts it, “That’s not a small amount if we compare to what we know otherwise from influenza vaccines.” According to Dyker’s calculation, the green batches account for more than 60% of the Danish sample.

There are then the ‘blue batches’ clustered around the blue line, which are obviously associated with an extraordinarily high level of adverse events. As Dyker notes, no more than 80,000 doses of any of the blue batches were administered in Denmark – suggesting that these especially bad batches may perhaps have been quietly pulled from the market by public health authorities.

Nonetheless, these batches had as many as 8,000 suspected adverse events associated with them. Eight thousand out of 80,000 doses would give a reporting rate of one suspected adverse event for every 10 doses – and Dyker notes that some of the blue batches are indeed associated with a reporting rate of as high as one suspected adverse event for every six doses!

On Dyker’s calculation, the blue batches represent less than 5% of the total number of doses included in the Danish study. Nonetheless, they are associated with nearly 50% of the 579 deaths recorded in the sample.

Finally, we have the ‘yellow batches’ clustered around the yellow line, which, as can be seen above, barely gets off the x-axis. On Dyker’s calculation, the yellow batches represent around 30% of the total. Dyker notes that they include batches comprising some 200,000 administered doses which are associated with literally zero suspected adverse events.

As Dyker puts it, “malicious” observers might note that “this is how placebos would look”.

And malicious observers might be right. For Dyker and Matysik compared the batch numbers contained in the Danish study with publicly available information on the batches approved for release, and they made the startling discovery that almost none of the harmless batches, unlike the very-bad and not-so-bad batches, appear to have been subject to any quality-control testing at all.