rjbuffchix: How do you know the reason they abandoned the game is because they are stuck or having trouble?
amok: I said " where most players are stuck
or when they abandon the game" - emphasis on
OR.
I'm failing to see the difference? How does the developer know the player got stuck, versus abandoned the game? How does the developer know the person didn't abandon the game for a different reason perhaps unrelated to gameplay/skill level?
rjbuffchix: ""Our metrics show that 90% of players quit our RPG after the 20 hour mark, even though there are 50 hours of content. So, our next RPG will contain 15 hours of content and naturally will retail for the same price".
Without these metrics, people might actually make deeper and longer games, designed to be enjoyed by single players/locally...imagine that!
amok: No, again you are not understanding what is happening. What a game designer would most likely think is "Why do the players stop at 20 hours, I would like them to expereicne all of the game I have made! So i need to figure out why they stop there, so i can improve the game so they progress further than that mark". What the metrics tell the developers is where the problem areas are so thet they can improve the game.
What happens in their next game? Most likely, exactly what I suggested.
rjbuffchix: A designer can somewhat anticipate what players want without needing to harvest data. After all, there was a time before achievements where designers did exactly that. Were games commonly being quit after a couple hours back then? I'm not sure we are currently seeing anything different, meaning the great advantage supposedly given by these metrics may not be that extensive
amok: and there was also a time when devlopers just.... did not.... and they made the same mistakes over and over again. all this is doing is letting designers know exactly where in a game a problem arises (which they could not do before) and use this knoweldge to improve and make 'better' games. But off course, they can just guess instead of know.
Yes...but the time you mentioned was primarily before the time I mentioned. Meaning that people have had the ability to learn from these multiple points in time both of us have mentioned. I don't see how the metrics logically connect to full knowledge the way you seem to suggest. The developers can't
know. They are making inferences from the data of the metrics to make what is at most an educated
guess. If we are in the realm of educated guessing, as I think we necessarily have to be no matter what, then the metric-gathering is not necessarily any better than observations that were freely available before all the datamining.