It's interesting because publications tend to also grade movies and music (Filter, Paste, Rolling Stone, et al). I agree that letter or number grades assessing quality are arbitrary, but the reviews themselves are arbitrary for the most part.

I think it's a difference, if, on one side, a music magazine rates a CD overall on a scale of 1 to 5, if Roger Ebert rates a movie overall on a scale of 1 to 4 or, on the other side, if a gaming magazine tries to break down the different parts of a game like graphic, sound, gameplay, etc. and gives all these points individual scores from 1 to 100 and combines them in the end to an overall score.

The movie and CD reviews are clearly subjective and the score is based on the personal taste of the reviewer. This is the reason, why most movies and CDs aren't advertised with the amount of points, they got from the Rolling Stone or from Roger Ebert, but with what the reviewers said about the movie or the CD.

Games reviews on the other side pretend to give an accurate measurement of the game. But it is impossible to rate the graphics of a game or the sound or the gameplay. What does graphics of 85% mean? What does a sound of 20% sound like (except from being a little bit to quiet ;-) )?

It's interesting when you use movies as an analogy as games tend to age poorly in the view of the masses. A movie from 2003 which was heralded as quite moving and well scripted is still an enjoyment to see but a game from 2006, which got 9.5 for graphics, may not be rated the same when played today.

It seems more analysts are reporting that review scores have little impact on game purchases yet people will argue that a game that got 9.2 should have gotten a 9.7 as if this has some difference. The system of a vague review score begins holding more merit than one that nitpicks a myriad of smaller metric values. In the end, do we not want to know if it's a "thumbs up" or a "thumbs down" game and let the text of the review point out the good and/or the bad?

Some reviews are quite biased and others miss the mark for those games that people will still enjoy despite the rough edges (see Operation Flashpoint: Dragon Rising for a good example). Word of mouth from fellow gamers and experimentation with rentals will always allow us to get a good sense of a new game while reviews should bring new releases into the spotlight with some context to help us decide if we should have our interest piqued.

I'm in total agreement with Katz here.

When I did reviews for a major Canadian local newspaper my style changed and by the end I've focused on a few things. I tried to communicate to the reader:

-the genre of the game
-the story/themes present
-what the game does and doesn't do well
-who it was appropriate for in terms of age ratings
-which sort of gamers would like it

There was no scoring system and at best there would be a hearty recommendation for very good games. In my view I was trying to convey the game in an honest fashion to the extent that if a reader read the review it would be as if they rented the game or tried it at a kiosk, and then as a result came to their own conclusion from the review. Unfortunately being locked in at 400 words didn't help in conveying this notion but at the same time it forced me to be concise and stick to the points that mattered.

As Katz alludes to, I do agree that letter grades can devalue a review and magazines to an extent. But some readers want that. They want to know if Avatar for the 360 is any good, and they can't be bothered to read an entire well-written review.

I also agree with Katz's clarification of "casual" games being "family social" games. The entire interview was great.

And I have to add one more thing...just as EG addressed a niche need for a gaming magazine with reviews and news, Gamasutra is doing the same with their sister sites like Indie Games, Fingergaming and GamerBytes. Not many other gaming websites (as many as there are) are taking the same sort of risk.

One thing nice about "scoring" a game is that it's much easier to archive and compile via sites like metacritic. For instance, Filter magazine gives albums a numerical score from 1 to 100. If I want to look back at the albums of 2009 I can see on one page what the magazine and its editors thought of all musical releases for one year.

This works well for games, movies, etc with sites like metacritic. For instance, I tend to agree with Roger Ebert but there are always examples where due to personal taste, life history, etc. Ebert and I are going to be miles apart. However, if I can go to rottentomatoes or metacritic I can get a pretty good general idea of the critical consensus. I might not agree with that consensus, but if a movie that I thought would have sharp dialogue and well drawn charcters has a 08% score on rottentomates, I'm going to read a couple more thorough reviews to see what the issues might be.

For games that tend to cost $60, web sites like metacritic can be invaluable in assisting the consumer in making good choices. "Quality" games, even misfires, tend to score in the 70s and 80s on site like metacritic or gamerankings, but the compilation sites, to me, can still be helpful for people that have time to make a few button clicks but not time to read a half dozen well written reviews.

Also, I don't think assigning a numerical score or number of stars prevents an editor/writer from writing well written reviews.

@Jonathan Gilmore

Actually it's been proven time and again that people don't pay much attention to scores anymore. Too often games magazines and sites have given high scores to games that simply didn't deserve them. Now about 25 -30% of people (and by all evidence the number is growing) rely on word of mouth. They will have one person in the group buy a game and effectively act as the review for the group. I myself have done this for some time.

If anything metacritic is only really used by games companies and naive developers, who then sit about wondering why their sales don't match the metacritic score percentages.

Sooner or later magazines are going to have to abandon conventional scoring or go back to the multiple sum-ups that magazines like crash 64 used to have. People went with the score of the reviewer who best suited their tastes. It worked well because it promoted a genuine honesty. Sometimes you would even get two scores so far apart as to wonder if they were looking at the same game at all but, if you knew that reviewer A liked shoot-em-ups and reviewer B liked strategy games, then you understood and went with the guy who best matched your own likes. It's how music reviews work even today!

@Peter Dwyer
I disagree with your take on music reviews-I can't name a single music critic. You are a little closer to the mark with movie critics. People know and trust a good number of movie critics to at least give their honest take. Besides Ebert I'd point out Kenneth Turan, Michael Philips and Peter Travers to name a few.

But back to games, when I bought my DS I relied on metacritic type cites to get my second game: GTA Chinatown Wars. That didn't turn out much better than my first game, Suikidon tactics. For that game I asked the people that worked in the Fye for a recommendation-and I played a game from that series on PS2. Anyhow, at least with GTA I could see at least it had high production values, which I think is generally reflected in the meta scores. GTA IV was extremely overrated as a game, but the unanimous glowing reviews were mostly a response to the apparent high production values-it was the best looking open world game by far. I think critics still have influence, but consumers get more and more sophisticated and learn to read between the lines.

On metacritic I read more user reviews and I find it interesting to see how big the gap between the metacritic score and the user score can be sometimes.

What I find most interesting is, that if you check the user score on metacritic and on some national games sites all over the world, that offer a user score for games themself, those user scores are often close to each other and another thing is, that a user score doesn't change that much, if only 50 people voted, or if 500 or 5000 voted.

I don't rely on user scores or user reviews, if I buy a game, but often, I look at them and check my own opinion against them, much more often, then I check my own opinion against magazine reviews.

I find most reviews are not especially well written and they don't try to develop any style, in the early 8 Bit days, I remember some fantastic reviews in C&VG, most of the reviews I remember totally thrashed the game, but it was written with wit and passion, it sometimes made me try out a game, just to see, if it was really that bad. But then it was the time of cheap budget games. :-)

If I write a review today, I try to point out, what's new in this particular title, what might have been the ideas and intentions of the developers and if they succeeded or not. I try to judge every game as a single piece of work (art if possible) and I try not to judge a game by it's technical standards only. I wouldn't want to read a review of the Avatar movie, that's only about the technical aspects and I don't want to read a games review with only technical measurement in it.

There will always be reviews. Reviews are not the problem, it's the content of the review that often is. Art, music, film, literature; etc. gets reviewed as well, with the same controversies

I thought about this again and I must say I have come to totally disagree with this guy's view on scoring games. Review "scores" can be very helpful because it allows the reader to guage just how much critic liked a game or whatever.

For instance, a well written review of a game that has no score at the end might explain how great or how terrible a game is, but a score lets the reader guage how great or terrible it is relative to the other games or whatever that reviewer has critiqued. Sometimes the scores can handcuff a critic-for example, if Edge magazine gave Perfect Dark Zero a 8, how can a different critic for that publication justify giving Bioshock an 7?

On balance though, I think that is a good thing. That comparison gives a pretty good overview of how high opinion the publication or critic has for any particular game, which can be helpful information. It also forces the reviewer to look back at how good a game is versus the width and breadth of what else is available to the reader/consumer.

There are two problems I see.

The first problem is that many people, including many of the posts here on Gamasutra, use terms such as "good" or "well" to discuss this issue, but those terms are subjective to each and every individual. Person A claims that a review that states a game's graphics are "well done" and agrees with that particular reviewer or compiled score cannot understand why Person B feels that such claims are completely incorrect, nor can Person A understand why Person B is very upset after spending their money on a game that is simply not worthwhile in the latter person's view despite "high" scores and praises from vocal proponents of the game (who often are a minority of the total global population, not a majority). This happens with any media, but games have a somewhat unique issue due to the second factor.

To wit, games are unlike many other media due to their complexity. This is true even for a relatively simple game. A movie can be watched in roughly two hours, and a novel can be read in perhaps 10 hours or so (for a longer novel... short examples can be read much faster, of course). This allows multiple passes for any attempt at reviewing a specific work. Games of such a length would be extremely short, and would generally receive very low review marks and comments no matter how enjoyable they may be. Consider how many people play Solitaire on their $1000-$2000 laptops. It doesn't matter that they could do the same thing with a $0.50 deck of cards; they choose to use their laptops, and they enjoy what they're doing, even if a review of such a game gets very low marks with respect to gameplay, graphics, and cost/benefit ratio. This analogy can be extended to any game, of course, but the main point is that many games, and perhaps even most titles, cannot be truly reviewed as far as content and enjoyability, even as far as the subjective tastes of any particular reviewer, because there simply is not enough time to thoroughly play and replay many games in order to offer commentary about the title. Many other media formats do not face this obstacle, or at least not in anywhere near as severe a fashion as games. Therefore, a so-called "review" of a game is almost a contradiction because there's really no way for a person to offer decent coverage of the product within the time and space constraints that such coverage always face.

@ Dave Endresak

The length of a typical AAA Title is 10 hours max. I know many novels, which I read much longer and If I read a review of a DVD Box of a TV series, the reviewer had to watch up to 26 episodes of the series, which consumes much more time, then playing the average game.
You are right about any review being subjective, but that's the same, with any other medium. The problem with games reviews is, that they claim to be objective, the whole score thing is saying, we can measure the qualitiy of the graphics from 1 to 100, we can measure the quality of the sound from 1 to 100, we can even measure the fun it is to play this game from 1 to 100 and if we combine all the scores, an "objective" number appears, that is a real indicator of the quality of the game.
That's totally stupid, for sure, but unless any form of games journalism comes into life, it won't change. To be honest in todays games business, it isn't important, if there is a score under the review or not, at least not here in germany, here is every single review bought by a publisher.