AstroManLuca
Lifer
It all just comes down to how many steps you want. A 1.0-10.0 scale theoretically has 91 possible ratings (is this correct math?). GameSpy's 1-5 star rating, with half stars permitted, still has between 9 and 11 steps (depending on whether 0 star or half star ratings are allowed).
Your system of providing "tilt" is basically the same as IGN's. They give regular old ratings to games, like "Decent" or "Outstanding," and like your system, the decimels are meant to provide tilt.
Why not go with a true 4-tier system, with no half stars allowed? 4 stars for excellent games, 3 for decent ones that are probably worth buying, 2 for ones that might be worth renting but probably not buying unless you're a die-hard fan, and 1 for really bad ones. Going with 4 levels instead of 3 gives just a little more flexibility to the proposed "buy, rent, avoid" scheme without losing the original spirit.
EDIT: I just read some comments from when 1up switched from numbers to grades. Something people mentioned was how, in school, an "F" was anything under 60%, and an average "C" was only 70-79.9%. This creates problems because people tend to associate anything under 7/10 as bad, even if 5 is supposed to be an average. So that's how IGN and GameSpot have been rating games. It doesn't really matter if a game gets 5.5 or 3.4 or 1.5, either way it fails and should be avoided. All the good games are clustered in the 7-10 range, so what's the point of having all those low numbers available? If a game rates below 6, do you really need to know precisely how badly it fails?
Using grades is good because then you can say "C" and people will think "average, not outstanding" without having to wonder whether "average" is supposed to be 5 or 7.
Your system of providing "tilt" is basically the same as IGN's. They give regular old ratings to games, like "Decent" or "Outstanding," and like your system, the decimels are meant to provide tilt.
Why not go with a true 4-tier system, with no half stars allowed? 4 stars for excellent games, 3 for decent ones that are probably worth buying, 2 for ones that might be worth renting but probably not buying unless you're a die-hard fan, and 1 for really bad ones. Going with 4 levels instead of 3 gives just a little more flexibility to the proposed "buy, rent, avoid" scheme without losing the original spirit.
EDIT: I just read some comments from when 1up switched from numbers to grades. Something people mentioned was how, in school, an "F" was anything under 60%, and an average "C" was only 70-79.9%. This creates problems because people tend to associate anything under 7/10 as bad, even if 5 is supposed to be an average. So that's how IGN and GameSpot have been rating games. It doesn't really matter if a game gets 5.5 or 3.4 or 1.5, either way it fails and should be avoided. All the good games are clustered in the 7-10 range, so what's the point of having all those low numbers available? If a game rates below 6, do you really need to know precisely how badly it fails?
Using grades is good because then you can say "C" and people will think "average, not outstanding" without having to wonder whether "average" is supposed to be 5 or 7.
