Predicting Launch Title Review Scores

Next Generation is hosting an article attempting to prognosticate the spread of review scores for the PS3 and Wii. Author Matt Matthews does this by examining historical precedents for previous system launches. From the article: "Next month Sony's PlayStation 3 will launch with only two games which will get an average review score of 90% or better. On the other hand, Nintendo's Wii will have three games which will average scores of 90% or better. And it is almost certain that each of those consoles will launch with two absolute stinkers, games which consistently score below 60%... How can we know this? Because history tells us that this is what happens with console launch titles." For even more analysis of the data, Matthews has additional charts on curmudgeon gamer.

Comments

Scores = joke

[daeg]

The problem with game ratings is they aren't fair or useful. What, exactly, is the difference between a 92% and a 93%? In order for them to be accurate and useful, game reviewers should be using the entire scope of their scale, not just the middle and top. A really bad game shouldn't get a 60% -- it should get a 10%. An average game shouldn't get 85%, it should get 50%.

Similarly, various review sites use category-based ratings, for instance, graphics are rated separately from sound effects. The game might have a low graphic, sound, and gameplay, but get a good overall grade. That doesn't make sense and they usually justify it in their review text by saying "We felt the average score of 80% just wasn't enough, so we're giving it an overall score of 90%." Why? If everything was so bad, your scores should back it up.

These and many other reasons are making game reviews less relevant and more like blog posts than concrete reviews.

Re:Scores = joke

[DrXym]

There are sites such as metacritic.com and gamerankings.com which tot up all the ratings for a game from dozens of sites and give an average. Just like with Rotten Tomatoes and films, it gives you a fair idea on whether a game is any good or a stinking heap of shit.

It can be useful to look at especially if some site has a weird rating system, or Nintendo or Sony buy a good review. Naturally you still have to read reviews to see if the game appeals to you but these site help there too since you have links to dozens of reviews all within a single page.

Questionable

[Nasajin]

I think this approach to analysing computer game review scores is pretty flawed. Naturally, a percentage of games are going to be better, and a similar percentage is going to be worse than the average. All this shows is that there's a consistent mean score, and that the rest of the scores operate in a bell curve. It may be that the reviewers for GameRankings are required to work within a certain spread of figures for their reviews.

If anything it probably says more about three or four people working at GameRankings, not an industry wide standard.