Yeah I don't like "star" type reviews anyways. It's too hard to really calibrate the level of the game in the review. 20pts per star is just too much either way toward good, average or bad. It's like saying the game is only 70pts overall.
I'd prefer reviews and scores based on the following:
AI strategic/tactical score
AI overall Challenge score on the games highest difficulty
Longevity/Replayability (random maps/random units give this score a bonus of 20pts)
Tutorial score (how deep does the tutorial teach, how simple is it to follow)
Multiplayer score (add 20pts if it has a PBEM ability).
Game length score (how long does it take to play one scenario and one campaign, the lower the times the higher the score)
Complexity score (how long does it take to understand the mechanics of the game and play reasonably well)
Demo value to the actual game (how much does the demo compliment the actual gameplay of the retail version? -20 points for games without demos
Manual score (how detailed does the manual explain the game mechanics and internal rules of play)
Personally I don't think there should be a "major" graphics or sound score these would be complimentary scores but would not count in the final value. Graphics and sounds do not make great games if there is no game to go with it. I see all too often high scores for games because the graphics are innovating or pretty.
Overall "fun" level of the game to the reviewer (reviewer should be an experienced gamer of the genre in my opinion though.)
Then take all of the above scores (excluding graphics and sound scores) add them together and come up with an average score.
Now, if I were using the above example for a game like say "Rome Total War"
5 - AI score
5 - AI score
7 - Replayability
5 - Tutorial (if no tutorial do not add into value, but mention it in article)
7 - Multiplayer
5 - Game length
9 - Complexity (higher means easy to learn)
5 - Demo value (if no demo do not add into value, but mention it in article)
7 - Manual information
(9) - Graphics
(9) - Sound
6 - Overall fun value of the game to reviewer
61/10 = 61%
Adjusted Graphics and sound value score
79/12 = 66%
Adjusted gross value of the game
64%
Of course being an experienced strategy/tactical gamer, has influence on the low score overall. Being experienced it's important for the AI and challenge level to be extremely good.
Now let's compare MTW scores
7 - AI score
7 - AI score
8 - Replayability
8 - Multiplayer
5 - Game length
9 - Complexity
6 - Tutorial value
7 - Manual information
(8) - Graphics
(9) - Sound
8 - Overall fun value of the game to reviewer
64/9 = 71%
Adjusted graphics and sound value score
81/11 = 74%
Adjusted gross value of the game
72%
So you can see MTW is 8% better than RTW overall. That coming from an experienced gamer of the genre.
This is the way I'd like to see games reviewed and scored on websites and magazines, but, they've become lazy and just put "stars" which are really no informative scoring system at all.
WE/I WANT 1:1 or something even 1:2 death animations in the KOIOS PANZER COMMAND SERIES don't forget Erik! and Floating Paratroopers We grew up with Minor, Marginal and Decisive victories why rock the boat with Marginal, Decisive and Legendary?