There are two main parts to most game reviews. One part consists of hundreds of carefully-considered words, precisely arranged to paint a complete picture of the gameplay experience. This outline of a game's good and bad points often delves deep into a reviewer's thought process and explains, sometimes excruciating detail, everything an informed consumer and game fan needs to know before making a purchasing decision.

The other part is usually a single number.

Try and guess which part is better at catching a reader's attention.

"People are always asking me what I would rate a game, expecting me to blurt out a number and thus convey my opinion of a game," says Joystiq blogger Ludwig Kietzmann. "Though I suppose brief attention spans and the expectation of quick answers are mostly to blame, I find that the inclusion of review scores in articles often overshadows all the words before or after it. The review becomes the score; it becomes a number."

Indeed, the whole concept of condensing a work as complex as a video game into a single number can be a bit ridiculous. Yet a shorthand score has become a de facto part of the large majority of game reviews, mainly because the readers demand it.

"At the tail end of Computer Gaming World's run, we tried removing review scores, because we really felt that people were focusing too much on the numbers and not enough on the reviews themselves," said Games For Windows Reviews Editor Ryan Scott. "Our audience was largely disappointed when we did this, to put it mildly. I think that, at this point, if you publish a game review in an enthusiast publication sans score, you're gonna get smacked by your readers for essentially taking something away."

For better or worse, readers have just been trained to look for that summary judgment. "We hid our ratings in hopes that readers would take a greater appreciation of the text," said GameCritics.com owner and founder Chi Kong Lui. "But the reality is, unless games are more thought-provoking and conceptually challenging, gamers won't look to game reviews to better understand the game experience. They will think as consumers and expect reviews to be something you find in Consumer Reports as opposed to something in the arts section of the New York Times."

So if getting rid of scores isn't really a viable option, maybe getting rid of some of the scoring options can help. "Thumbs up/thumbs down leaves zero room for ambiguity," says freelancer Greg Sewart. "The reader doesn't have to figure out the real-life value of a particular number score that way. The only real purpose of a game review is to tell the reader whether you think they should buy it or shouldn't buy it. ... Is the game worth the MSRP? Yes or no?"

While some journalists see value in a finely-graded scale, most reviewers I talked to agreed with freelancer Kieron Gillen's assessment. "Marking is an art, not a science," he said. "The more 'definitive' a marking scheme gets, the more it's pretending to be in some way objective, and lives in denial about the squishy human stuff glooping around inside our heads. I often talk about doing a mark-scheme out of 72,384 or something to just really push the fact marks are ridiculous -- yet fun -- to the forefront."

Even then, though, the question would become whether a midpoint score of 36,192 out of 72,384 is really an "average" game. "The whole 'average score' thing is such a huge can of worms," Games For Windows' Scott said. "The 70 percent/C-average mentality is drilled into our heads at a young age. It's a weird sort of Pavlovian conditioning -- 'anything below 70 is terrible!' -- that doesn't make any sense when you actually sit down and examine the logic behind it. Yet many writers and publications slavishly defend it."

"I feel a 7/10 average doesn't make good use of a 10-point scale," said Electronic Gaming Monthly Editor-in-Chief Dan Hsu. "So you can have three scores for 'good'...and seven scores for 'bad'? That seems so unbalanced to me. At that point, what's the difference between a score of a two or a six? It's all 'fail' when it dips below a seven. "

And when a game does fall below the average, there's more hanging in the balance than a simple individual buying decision. A PR rep's job can hang on an aggregate review score, as detailed in a recent Gamasutra article. "The score would never live up to the expectation," former Rockstar PR rep Todd Zuniga said in the article. "If it scored a 99, the expectation was for every other review to be 100."

Sewart thinks blaming PR people for low scores is ludicrous. "The score is (or should) be based solely on the quality of the game, which the PR reps have absolutely nothing to do with," he said. "To use average scores to judge marketing effectiveness is the same as saying they're trying to figure out whether the 'payola' worked or not."

More than just PR jobs, though, review scores can have a profound effect at the retail level. "Aggregate scores are being used to determine re-orders at retail, to greenlight sequels, and as payment bonuses," says former Computer Games editor Steve Bauman. "It's a depressing trend. A collection of arbitrary numbers, when added together in a rather arbitrary way, becomes an even more arbitrary and meaningless number. While they provide a good general indicator of quality (or a lack thereof), they're an overly blunt instrument."

So how can we make readers focus less on the scores and more on the text? A few journalists suggested that making the actual writing better would help, but others doubt how effective that would really be. "No one reads text," Bauman said, "so if a reader can't be bothered to care about the 'why' of a score, there's little a writer or publication can do."

One interesting solution might be doing away with numbers and simply adding more words. "You can say a lot about a game with a single word," Joystiq's Kietzmann says. "Epic. Miserable. Bubblegum. Moving. Stick a nice, bold word at the end of the review that captures how you felt about the game."

Or, if you prefer, just stop worrying about it so much. "If readers want to base their decisions off a random number, then so be it," Sewart said. Or, as former Tips & Tricks editor Bill Kunkel more bluntly puts it, "Anyone who buys/rents ... any game based solely on somebody's star or number-based rating rather than the review itself is pretty much a tool anyway."

— —

Got something you'd like to see on Media Coverage? Send it to kyle.orland@gmail.com.

Media Coverage is an opinion column. The opinions expressed in this column are solely the opinions of the columnist and are not necessarily the opinions of GameDaily.com.