*Sigh*... The whole "Simplify/dumb down/5 star" thing again... Same discussion that happened at Gamespy in 2003, when they moved to 5 stars. I am pretty sure nobody will read my boring stuff, but, here is my take on this:
People often defend "simplifying" with the following argument: "since the reviewer ends up guessing that last digit anyway, remove it!". It always sounds to me like "since the last wagon of a train is the most prone to accidents, remove it!". I mean, the last digit is always sort of "guessed"; the thoughtful writer knows it, the intelligent reader knows it, we communicate. Remember basic Physics? Any measurement is made of digits of which you are sure, and a last digit that is just an estimate. So when I read "length of 6,54m", I know it is 6,5m, and some more 0,03-0,06m. That is the correct way to read it!
So of course the 6.6 games and the 6.7 games are very similar and, in a different day, maybe their order would be switched (though I agree with some people here that, if the reviews are done with strict coherence, it is actually possible to be that discriminating; I rate my games in a 0.0 to 10.0 scale and I can tell the difference between the 8.0 and the 8.1 games -- you might not AGREE with my ratings, it is subjective after all, but *I* can tell the difference according to *MY* opinion; of course, it is much harder to keep this strict coherence across different reviewers...).This was never a problem with the measurement SYSTEM, it isan unavoidable component of any MEASUREMENT!
Now, think the other way round:at least, I KNEW THAT, IF THE SCORES WERE 0.1 DIFFERENT, THAT DID NOT MATTER MUCH. Now, the 6.6 and 6.7 games might become 6.5 and 7.0 respectively. If you read it face value, two very very similar games in quality will go to "significantly different" categories.
But wait -- the intelligent reader will not read it face value -- having figured this out, this reader will actually know that 6.5 and 7.0 now COULD be pretty much the same, depending on the games, and that last digit is not that meaningful. All we have done is removing the last wagon; we are back to "the last digit is sort of arbitrary, really".
Another borderline example: I KNEW FOR SURE THAT A 6.6 GAME AND A 7.3 GAME WERE QUITE DIFFERENT; NOW, IN A GOOD DAY FOR THE FIRST AND A BAD DAY FOR THE SECOND, BOTH GET 7.0. So the sparse reviewing system is not uniformly bad, but will screw up occasionally with games that are "borderline" between categories.
Summary: humans are not precise in rating games (and movies, and students, whatever), nor they should be (specially when opinions are involved). Exactly because of this imprecision, we have to ALLOW fine scoring, so these "imprecisions" can be kept small! Sparse scoring FORCES INCREASES in imprecision and, as a consequence, invites raters to be lazier about the scores. (I am a teacher, and a defender of the 0.0-10.0 scale as opposite to A-E or something; unhappily, the PASS/FAIL line has to be drawn somewhere, and that alone leads to perceived "injustice", independently of the rating system)
I add that one of the reasons that I liked Gamespot more than Gamespy is that they gave scores from 1.0 to 10.0 and the scores seemed coherent with the text reviews, at least the PC scores that I read, anyway (note: PC Gamer also gives 0-100 scores, but they don't seem nearly as coherent, so I do not like it as much).
Just my 0.5 dollars... Or should it be 0? :P
Log in to comment