Huh. I just realized you are - and I was - misinterpreting the Pfizer analysis.

You should have started by dividing 275 by 30 million, not 42,086.

Now that I've read the introduction I realize that 42,086 was the number of adverse events reported, not the number of subjects in a trial.

About 30 million doses of Pfizer's vaccine were used between its first authorization - the temporary one - on 01 December 2020 and the end of the analysis period on 28 February 2021. https://ourworldindata.org/grapher/covid-vaccine-doses-by-manufacturer

In that time 42,086 adverse events were reported through a variety of protocols - see section 2. Methodology on page 5.

So the percentage of reported stroke victims is (275/30,000,000)*100. Something like 0.0001%.

And that's without knowing if any of the strokes were even caused by the vaccine.