Review Scores the "Least Important Factor" When Buying Games

A recent report from a games industry analyst suggests that among a number of factors leading to the purchase of a video game — such as price, graphics and word of mouth — the game's aggregated review score is the least important measure. Analyst Doug Creutz said, "We believe that while Metacritic scores may be correlated to game quality and word of mouth, and thus somewhat predictive of title performance, they are unlikely in and of themselves to drive or undermine the success of a game. We note this, in part, because of persistent rumors that some game developers have been jawboning game reviewers into giving their games higher critical review scores. We believe the publishers are better served by spending their time on the development process than by 'grade-grubbing' after the fact."

Comments

Let me be the first to say...

[thijsh]

maybe it's the other way around... You only have to buy a sucky game *once* based on a raving review to *never* trust those reviews again. While your friends can comment properly on the game without some obscure metric like '8/10 overall'.

Except you just illiustrate the problem

[Moraelin]

Except you just illustrate the problem: something that's just "good enough" (which really just means "mediocre") gets an 8 out of 10. I'm sorry, but in a perfectly linear scale, "mediocre" would mean a 5. That's the kind of a number you could punch in a formula and get a correlation or anything else.

Plus, if it were just a case of a honest review and the bad ones being already cancelled, the results would look much like the right half of a bell curve. You know, the curve with the below average ones removed. For virtually any sitze out there, it doesn't. It looks like a bell curve centered on 8 or 9, and which pretty much starts at 6 or 7. Sorry, that's not a case of the bad ones being already removed, that's a clear sign of an offset scale. It's what you get when the occasional "something that's truly bad" means you get to give a 5 or a 6, not a 1 or 2.

And then there is the occasional reviewer whose curve looks like two spikes. The kind who churns 90% to 99% scores all year long, and then occasionally picks up some 10 year old freeware game so he can give _something_ a 5% score and fix his street cred. Or publishes a yearly smack-talking "top 10 worst games of all time" -- conveniently all 20 years old and from publishers which are no longer in business -- just to show that he's that unbiased and can give a low score too.

But again, that's not being unbiased and fair at all, it's just trying to compensate one crap (or dishonest) job with another one skewed in the other direction. If it were a real fair and unbiased and non-skewed job, you'd get one bell curve centered in the right place, not two spikes centered near the extremes of the scale.