According to industry experts, the percentage of people who actually see a video game's ending is "10-20%", a staggeringly tiny number once you start to consider the sheer amount of stories worth experiencing across all of video games, and the amount of hours we collectively sink in.
It creates a whole new way of discussing what makes a "good video game ending" i.e. something worth playing for tens upon tens of hours, to either hit a payoff of seeing credits, or to have some sort of narrative resolution.
With so much on the line - and unless you have the story mapped out in its entirety from day one, the ending will be the last thing that gets put together - it's so easy for it all to go wrong.
Because of this, closing levels, cutscenes or entire back thirds can get retconned or otherwise ignored over time; the very idea of an "ending" meaning less and less in a medium full of DLC, patches and reworks.
A fifth of your playerbase should still very much matter, but sometimes - despite the purpose of an ending being to give closure - they can leave us with anything but.