This is not actually true -- I ran some analysis on review data and when controlling for downloads there isn't a strong correlation between ranking and # of reviews.

For example, https://sensortower.com/ios/us/funzio-inc/app/modern-war/468... Modern War had a spike on Feb 9th -- ranking stayed as predicted.

Another example: Facebook https://sensortower.com/ios/us/facebook-inc/app/facebook/284... on Jan 23 or Smule Dec 5th: https://sensortower.com/ios/us/smule/app/magic-piano-by-smul...

A lot of apps can push user messages say via urban airship to prompt for reviews, and we see events like this all the time.