IGN took a poll asking how many people were experiencing problems with Skyrim, of the people who voted 75% had problems, of that 75% a quarter of the people had gamebreaking issues. Pretty damn big numbers
Firstly, that's not a random sample at all. That's 75% of the people that decided to bother with that poll were having issues. And the people that even saw the poll in the first place isn't a random sample either, because that consists of people that are online searching for Skyrim information. Already, that whole group are probably users that need information for one thing or another that's bothering them about the game.
If you want a random sample, you need to start sending out surveys to a selection of players. Maybe hang out at a gamestop for a few days and pass out a survey sheet to everyone that buys the game. That's a random sample
Secondly, what did they use as the definition of gamebreaking? There are several million people playing this game on each of 3 platforms. A few hundred responses to an online poll is not representative of anything.