@Gamerace: That's what i was thinking about, although, don't forget the multiplayer.
The thing is, that there's no noticeable difference, aside from sales numbers, between PS3, 360 and Wii. So, even in this case, the results should be in line with PS360.

 Yeah.. but.. we know there's some very casual Wii players who do in fact play fairly irregularly.  But for the most part all PS3/360 owners would be fairly heavy users.   For the Wii to avg the same despite having a large casual contingent would make their 'core' owners pretty fanatical users, for every girl like my daughter some obsessive player to counter balance?  I find that hard to shallow as well.

Bottom line:  There's not enough info here to make these numbers meaningful.

@Gamerace: Look at the first link in my post with links, it has similar table from june 2007 and all three fit into about 10% differences. Then, if you look at the article about 2005-2006 research, it says that "active gamers" play average of 13 hours a week on console and 17 hours a week on handheld. Now we know that, according to Nielsen, PS3 and 360 have playtimes/habits similar to Wii (along with other consoles), the numbers don't add up, unless there's "outsiders" to count in.
And yes, there are heavy users to offset the not-so-heavy ones.

And one other thing, is the usage time the time actually used or the time used in games.

The "casual" is wrong approach here, notice how popular for example Madden is on PS360 and it's a game that fits perfectly into the "casual" mold, from which the "Wii userbase consists of".

lol, I definitely use my ps3 more then 5-6 times a month for 50 min.

First - let me give you props for digging all that info up.

Second - Okay, we'll go with that, sounds okay to me.   I do tend to forget the Madden/PES players.  Good point.

As I thought, the info gathered is done by survey. Internet survey at that, with no actual qualifications except for clicking the "yes" box. I knew there was something wrong with these charts.

I think the farce has gone on long enough. I seriously thought that someone would actually point out the blatantly flawed methodology behind these numbers. Thus I could actually play along, but it is apparent now that nobody is getting whats wrong. That is just damned scary.

The Nielson group is not conducting a scientifically rigorous survey. More to the point this data is secondary to their purposes. To be a Nielson family you must meet specific guidelines. One of those specific guidelines is you must watch a certain amount of television on a weekly basis. What gets in the way of watching television why gaming of coarse. So if your viewing is going to be monitored they want to at least have it be worth their time. So if your watching less television, because your gaming you are probably not going to be included.

Now if this data was not collected secondarily, and was the principle of a study then you have to be aware that the observation does alter the outcome. Thus people can play more or less, or they could even fudge the results. After all if your asked how much you gamed this week you are inclined to have gamed some amount of time. Even if you have to fabricate time played, because you obviously have a desire to provide some data.

Basically the study is flawed either way. They artificially alter the average, because they exclude participants. They alter the outcome by polling which encourages participants to alter their behavior. The best way to gather this data is actually through subterfuge collecting the data without the persons involved knowing.

I think the best statistical study I heard about was the bait and switch poll. Where the interviewer asked a person a series of questions about one subject, and then created a situation where they talked about something conversational. A example would be someone being interviewed about politics, and then for instance the laptop freezes up. So while the interviewer reboots the computer they engage the interviewee about some unrelated subject conversationally.

Which was actually the point. Since the person being interviewed does not understand that the conversation is actually the survey. Since they do not know its a survey they do not censor their answers. They answer as they really feel, or what their perceptions are. Which is actually the best kind of data the data least likely to be contaminated.

Oh I've agreed from the start.
something was very wrong with the survey.
I liked your post though. You write very intelligently

Unless you compared it to PS2 at the same time i'm not sure how you can tell if it's good or not.

I mean... i'd guess consoles with lower userbases probably have higher time played per week.  Cause well.. only people who play every day own them.