SELECT app, AVG(stars) AS display_stars FROM customer_reviews GROUP BY app;
It’s actually an ML algorithm that optimizes for revenue. SELECT app, AVG(stars) AS display_stars FROM customer_reviews GROUP BY app;
It’s actually an ML algorithm that optimizes for revenue.I'm sure that there are many tweaks to the way apps are ranked, but I would expect that the rating itself would, indeed, mostly be a star average.
I can confirm that the displayed rating on both stores is exactly the weighted average of valid reviews and ratings posted by users. No trickery there.
The only caveat here is that both platform regularly invalidate posted reviews and ratings for various reasons (the user asked for it to be removed, apple/google has reasons to believe it was fraudulent/automated, ...). If a review/rating is removed, it won't be shown and won't be taken into account in the calculation.
So the calculation is really along the lines of :
> SELECT app, AVG(stars) AS display_stars FROM customer_reviews WHERE deleted=false GROUP BY app;
They will probably reenable ratings for the Robinhood app when things calm down, so if you want to give it a 1 star rating just wait a few weeks and try not to forget about it.
Or, even better, high ratings for apps from large corporations encourage them to support the continuation of your store's monopoly power.
There's a massive selection bias at play which tends to naturally inflate review scores without any explicit fudging.