To be fair, I recall officiating accuracy being regularly reported as 90+% prior to VAR being implemented. Which would be expected, as they wouldn’t be reporting that about 1 out of every 5 decisions are incorrect, would they. Especially given they had been resisting implementing video review for decades after it had been used and continually refined elsewhere in the world.
It’s fairly common for past data points to be revised when an entity needs to provide proof of efficacy for a new policy. Happens all the time in non-sport industry. In fact, there’s an entire industry built up to support such endeavours.
I know, because I used to audit it. ;-)
*44% of those that have responded. It’s probably not actually 44%, though.