First off, I'd like to highlight the
issues with the numeric scoring system so often employed with any
review in any medium. Many reviewers will use a scale of some sort,
giving a score along a range of (for examples) 0-5 stars, 1-10, or
even 1-100. A main problem with this system is its
confusion/collision with the grading system of many schools. In
school, a 75/100 is considered 'average', with 80/100 being 'good',
and 90-100/100 being great. Any grade below 65/100 is considered a
failure. Having this rubric applied to review scores is somewhat
confusing. This condenses many game scores into a very narrow range.
Ideally, an average score would be 50/100. At this point, it would be
something that is not great, but not bad, just average. This
view has recently put game journalist Jim Sterling into focus. In his
review of the latest Batman
game (Arkham: Origins),
he gave the game a 3.5/10. While the education based system would
mark that as a total and utter failure, Jim merely reviewed it as
somewhat below average. Many of the comments on Reddit's Games
section echoed similar feelings about the review score – initially
it seemed low, but then they read his words on it and saw that he
backed his score up with the written review (the top three
first-level comments are all basically this).
Secondly,
I'd like to take on the issue of Metacritic. Metacritic is a review
score aggregator. They take review scores from many sites on the
internet, put them into a proprietray algorithm, and hand out a
Metascore. The Metascore is,
once again, on a scale of 1-100. They even further perpetuate the
problem from the last paragraph by color coding their scores. Green
is 75-100, yellow is 50-74,
and less than 50 is red. That means half of the possible ratings are
red scores. The vast majority
of games released in 2012 for PS3, Xbox 360, and the Wii U/Wii were
in the yellow category. One
of the most mysterious
parts of the Metascore is that different sites are given different
weights for their scoring, and nobody outside of the company knows
what the weights are. When a
certain site scores a game, some of them affect the Metascore more
than others, with some sites likely not affecting it much at all.
Metacritic is also looked at
by developers and publishers of games. Obsidian, developers of
Fallout: New Vegas
were to receive a bonus if their game scored over 85 on Metacritic.
The game just missed that target, and only received an 84 on
Metacritic. Creative
Assembly, developers of the Total War
franchise actually looked at Metacritic scores and cut content if
they think that it won't give them a better Metascore.
Personally,
I think that reviewers should use and advocate for rating systems
more based around an average game being in the middle than in the top
quarter of ratings. I know that some games writers/journalists are
staunchly in favor of getting rid of the review scores completely,
but I think that they can help provide information quickly and
effectively is used correctly. Metacritic
should also be far less important than it is in many developers'
eyes, but with the marketing departments being so strong in some of
these companies, it would take a lot of work to separate them from
its influence.
No comments:
Post a Comment