People talk on here a lot about the impact of review scores on how well a game sells. I thought it would be interesting to see the correlation of software sales and review scores.  So I used a random number generator, to pick 25 games out of the top 300 software sellers on a platform (after the top 300 I wasn't sure we would have complete sales information).  I then got the gamerankings rating for the 25 games, and came up with these charts.

Edit: Just relaized I didn't label the axis' on the graphs.  The vertical axis is the review score average out of 10, and the horizontal axis is millions of copies sold of the title.

If you've never taken an introductory statistics course, r is the correlation coefficient and has a value of -1<r<1 inclusive, with 1 being a perfect positive relationship and 0 being no relationship.  So you can see that the PS3 and 360 have similar correlations between review scores and software sales at a fairly strong ~.65.  Wii also has a positive relationship between software sales and review scores, but it is a much weaker relationship compared to the PS3/360.

Again if you've you ever been in a statistics course the first thing you learn is "correlation does not mean causation". I.e it is certainly possible that another factor explains why games with higher reviews typically get higher sales, such as games that get higher review scores also have higher advertising budgets.

So what do you all think? This information goes along with what was commonly believed (that review scores don't impact the Wii as much as the HD systems), but I still thought it was interesting to look at more thoroughly.

Reviewers align to the core market. There's little (or any?) media for the expanded market.

I don't really find it very convincing evidence, the large majority of data points are very close to the Y axis regardless of review score. The high selling games seem to be the outliers in this data and basing the regression line on them seems arbitrary at best.

Why are you missing PS3 titles though? PS3 has 17 titles above 90 and has the same titles close to 10 that the 360 does, not that its important, it just looks like ps3 titles level off at 9 while they actually level off in the exact same location as the 360.

you should have removed the outliers from your top 300 list.... there it seems you managed to get wii fit and wii play.... two badly reviewed games that skew the wii graph...

@inverted, I wasn't just picking the highly rated titles, in the VGChartz Database, I had the games sorted by sales on the platform. And then used a random number generator to pick 25 of the top 300 selling titles on the system, and then of those 25 titles I found the gamerankings score. So for example, one of the numbers that came up was 3. So the 3rd best selling title on the 360 is Grand Theft Auto 4 which has a gamerankings score of 9.7. The 3rd best selling title on the PS3 is Metal Gear Solid 4 with a gamerankings score of a 9.3, so Metal Gear Solid 4 was used on the PS3 not Grand Theft Auto 4.

yet they are wii titles, that are part of the success of the console and define the main audience, why remove them?

@Hephaestos, Wii Fit is actually the 4th highest reviewed game of the Wii titles generated. And at a 6.1, Wii Play is right around the Wii average of 6.5, of the games selected.

@Hephestos

Why remove the outliers? It's a random sample, you don't get to choose your data.

Of course it would be interesting to see this extended on a bigger sample of hundreds of titles per console, but first we should thank the OP for the time he spent.

Okay, then you should point out that the important piece of data on the plots is the r value. Which shows the agreement to sales and better games. hence, ps3 owners are the ones who are most likely to only buy good games, though xbox 360 is basically the same.