Reasonable on 25 July 2009
| astrosmash said: The problem with any investigation of the impact of reviews on the sales of games, is that it would be difficult and extremely subjective to factor out the inherent quality of the game. That is to say, if you are comparing the sales of two games one of which received a 90% and the other received a 70%, how do you know whether the sales were impacted by the better review or if they were impacted because the 90% was a better game? I think the only thing you could eliminate the impact of game quality would be to only look at day one sales; though then I should think that the review would be just one cog on the hype machine. |
That's a standard forecasting issue. In the end you don't know for sure, you just have very strongly implied correlations.
There are various approaches, some are simpler, some are more complex. Some look at each item individually, others look at 'cross-elasticities' to also include how items may affect each other - for example if two FPS launch at the same time will they affect each other, and how? What about the latest FPS from a popular franchise vs a new FPS IP?
It can be done though, you just need the correct algorithms and either patience or an actual forecasting engine to undertake the grunt work for you.
Try to be reasonable... its easier than you think...
(Former) Lead Moderator
(mostly)
