A comparison of the ratios of good games to bad games on 6 consoles

[Sqrl]

Kasz216 said:

Sqrl said:

The X axis is unlabeled and unexplained, the methods used for selecting games is undisclosed with the exception of top 10 chart, and in general there isn't enough information given to do anything but guess what its trying to say. So in essence we are left to take it on good faith that the graph has meaning.

But beyond all of that they are review scores and as I pointed out review scores are worthless unless you have the exact same tastes as the reviewers....considering metacritic is a composite review score its pretty much guaranteed to that the scores aren't born out of your tastes.

I'm not trying to be mean here but there are flaws beyond just my disagreement with the review scores. The effort was well-intended I'm sure but the result isn't that great to be brutally honest. With some adjustments it probably can be made useful but that would just get passed the technical issues with the graphs and we would still be left with the subjective nature of the data.

Which is really the point. Lots of subjective data does not create objective data. It is still subjective and by that virtue alone the data isn't meaningful unless you share the views and tastes.

I would argue lots of subjective data gets objective when you get more data and it's evenly distributed... of course that's a problem with this data since reveiwers are likely to be nowhere near evenly distributed as score consistancy rarely even makes sense in site.

Of course also based on this method I'm sure Dane cook movies would rank way too favorably overall.

Subjective information never becomes objective. It can be the predominate feeling and it can be the general consensus but neither of those things is objective. What you are thinking of is anecdotes becoming data. Which is true, if you use proper sampling and use a sample size of sufficient size you can take several anecdotes and produce data. But if, for instance, those anecdotes were reviews (subjective) then what you have is hard data about people's opinions, a breakdown of how many people believe what (depending on how you organize the data). But at no point are you ever able to say that one of those reviews or opinions is more correct than any other.

The same can be said for sales data, each person accounted for is essentially saying "I think this is interesting and its a fair/good value.". But I don't think you can say that the Wii's high sales prove that its a better console than the PS3/360. You can say more people like it, but it is still a subjective opinion held by a group of people, albeit a large group.

In the case of the sales data at least there is random sampling in that anyone can purchase the system. With composite metacritic reviews we are talking about an extremely bias sample, by even the most leniant of sampling practices.