ajqtrz
Chef - loquacious Old Dog
In another post, recently, someone suggested that the forums draw a larger proportion of unhappy players than are in the actual game. I thought it an interesting idea and wondered about it. And maybe that player is right. But maybe not. So I asked myself what it would actually take to know this was true...that the percentage of players in the forum who are "unhappy" is greater than the percentage of the players in the whole game who are "unhappy." Turns out, it would take a bit of work.
First, you'd have to define "unhappy." A broad definition might be: "I'm unhappy with at least one aspect of the game." And even that would mean defining if you mean "a little irritated," or "truly unhappy and ready to quit altogether." And at the narrower side you might define unhappy as "I'm right on the verge of rage quitting because every aspect of the games angers me greatly!" Thus, "unhappy" might include both a measure of the intensity of the emotion and the range of causes for that emotion. And that's just problem one.
Another problem is concluding the forums draw more unhappy players than happy. Not sure, but yes, a survey could be done. Unfortunately, it runs into the same problem as above...the definition of "unhappy."
And finally, you would have to do the same survey with the general player base. Unfortunately, the whole thing may be skewed because who answers surveys? I suspect, but have no actual data on the matter, that those of us who go to forums and express our opinions might be the types who also complete surveys more often. All of which takes me to the subject at hand: psychological statistics versus actual statistic.
An actual statistic is a count of something. It is a set of definitions, a set of recorded observations, and a mathematical analysis of that data. The definitions define what you are counting as X, Y, and Z. The collection of data says how many of X, Y, and Z you measured (via sight, sound or some other sense, directly or indirectly), in the set of observations you made. In the end you present the analysis as an analysis of that set of observations and expect your conclusions to be the same to some degree as anyone using the same definitions, recording the same observations under the same circumstances. In other words, you expect to some degree, that if the experiment you did were repeated many times there would be a high degree of agreement between the analysis you did of your observations and those done by others on their observations.
But of course, in the vast realm of human experience most things are never that carefully observed, so we rely upon an intuitive sense to measure most things. We say, "likely," "probably" and so on. These are not precise words exactly, but they do have statistical weight. "likely" and "probably" both say, at minimum, the predicted outcome is at least 50% plus some tiny fraction of the percent, going to be X rather than Y. But they can also mean 80% of the time it will be X rather than Y. Or 70% or 65%... or whatever, so long as it's greater than that minimum of 50%+a smidge. And that's why psychological statistic words are useful. They tell us not the actual number, but our sense of it. They are, in fact, not a measure of the thing, but a measure of our experience, and that, once removed, exactly because we don't usually record what we are experiencing.
So the idea that the forums draw a higher percentage of "unhappy" players than are in the general population of players may be true. Or it may not. The author of the sentiment may be right, but he may be wrong as well. What usually happens is that people either agree with him or not, and if they agree with him they support his conclusion. And if they don't they sometimes offer something like, "I find the people in the forums complain a lot less than the people in the game," an implied psychological statistic of their own!
The take away from this is simply to recognize that not a lot has actually been measured and even if your sense of something is strong, it may be you may not have the average experience. A bit of humility is always a good thing in discussion.
Just some thoughts triggered by a players comments.
AJ
First, you'd have to define "unhappy." A broad definition might be: "I'm unhappy with at least one aspect of the game." And even that would mean defining if you mean "a little irritated," or "truly unhappy and ready to quit altogether." And at the narrower side you might define unhappy as "I'm right on the verge of rage quitting because every aspect of the games angers me greatly!" Thus, "unhappy" might include both a measure of the intensity of the emotion and the range of causes for that emotion. And that's just problem one.
Another problem is concluding the forums draw more unhappy players than happy. Not sure, but yes, a survey could be done. Unfortunately, it runs into the same problem as above...the definition of "unhappy."
And finally, you would have to do the same survey with the general player base. Unfortunately, the whole thing may be skewed because who answers surveys? I suspect, but have no actual data on the matter, that those of us who go to forums and express our opinions might be the types who also complete surveys more often. All of which takes me to the subject at hand: psychological statistics versus actual statistic.
An actual statistic is a count of something. It is a set of definitions, a set of recorded observations, and a mathematical analysis of that data. The definitions define what you are counting as X, Y, and Z. The collection of data says how many of X, Y, and Z you measured (via sight, sound or some other sense, directly or indirectly), in the set of observations you made. In the end you present the analysis as an analysis of that set of observations and expect your conclusions to be the same to some degree as anyone using the same definitions, recording the same observations under the same circumstances. In other words, you expect to some degree, that if the experiment you did were repeated many times there would be a high degree of agreement between the analysis you did of your observations and those done by others on their observations.
But of course, in the vast realm of human experience most things are never that carefully observed, so we rely upon an intuitive sense to measure most things. We say, "likely," "probably" and so on. These are not precise words exactly, but they do have statistical weight. "likely" and "probably" both say, at minimum, the predicted outcome is at least 50% plus some tiny fraction of the percent, going to be X rather than Y. But they can also mean 80% of the time it will be X rather than Y. Or 70% or 65%... or whatever, so long as it's greater than that minimum of 50%+a smidge. And that's why psychological statistic words are useful. They tell us not the actual number, but our sense of it. They are, in fact, not a measure of the thing, but a measure of our experience, and that, once removed, exactly because we don't usually record what we are experiencing.
So the idea that the forums draw a higher percentage of "unhappy" players than are in the general population of players may be true. Or it may not. The author of the sentiment may be right, but he may be wrong as well. What usually happens is that people either agree with him or not, and if they agree with him they support his conclusion. And if they don't they sometimes offer something like, "I find the people in the forums complain a lot less than the people in the game," an implied psychological statistic of their own!
The take away from this is simply to recognize that not a lot has actually been measured and even if your sense of something is strong, it may be you may not have the average experience. A bit of humility is always a good thing in discussion.
Just some thoughts triggered by a players comments.
AJ