If I were a reviewer, I wouldn't use any ratings at all. Instead, I'd post a small "Buying Guide" section at the bottom of the review that states:

1) The average primary (or secondary, if the game is out of print) market value at the time of review
2) The price that you should buy it at (equal to or lower than the current market value; or a simple "Don't")
3) Any factors that might influence your purchasing decision one way or the other.

That way, the reader has both a detailed review of the game, along the lines of a movie or music review, and a recommendation as to whether or not to buy the game, all in the same review. Someone who doesn't care about one section or the other can just skip the part that they don't care about. And, best of all, there's no arbitrary "review scores" required.

A game is either good or crap in my eyes.

I don't pay any attention to sites that use a 100 point review scale. How can the difference between a 73 and a 74 be anything other than arbitrary?

mine is taken from the old 16bit era "mean machines sega" and "nintendo power" if you was around in the uk at that time.

Presentation

Sound

Graphics

Playability

Lastability

Overall

each with a % score rounded up with a overall % on how good the game is

Games I'll pay full price for, games i'll wait to drop in price, and games I won't buy.

that's it.

It's better than:
It's worse than:

If it's better than a similar game that I already own, I'll pick it up. If it's worse to a similar game that I already own, I won't pick it up.

Of course it doesn't always directly pan out like that, otherwise I'd end up with very few games that I'd be very bored with.

Mine goes like this:

or

Mine uses a 5 point grading system as well, but essentially goes from 0 - 4 as it's a Letter based grading system.
Basically A games are MUST HAVES, B games are worth buying, C games should be rented first to check, D games might be worth picking up at a hefty discount(or if gotten for free as a gift) if bored, but F games should be avoided at all times.

For further detail:

F - This game is a failure. It has to either be unplayable or broken in some manner - Ninja Bread Man, Big Rigs: Over The Road Racing, etc.

D - There's nothing that's actually broken about the game, it's just designed badly and isn't really fun to play - can't really say for certain what games fall here since I generally avoid anything that looks like it even might be lower than a C, and none of these games would be bad enough to mentioned as an outright BAD game

C - It's a good game, playable, and might even be enjoyable, but doesn't really have any personality to make it stand out versus other similar games - Wii sports knockoffs, maybe Mario & Sonic Olympics, DDR games after DDRExtreme, Rent first to see if you like it before buying.

B - Two options here, either it's an incredibly fun and well designed game despite not having much innovation, or it's an incredibly innovative game but still shows rough spots from where the innovations may not have been implemented quite perfectly, so it's fun, but still is lacking just a bit - Wind Waker, Wii Sports, DDRMax, In The Groove, Sonic Adventures DX

A - For what it does, it's the best game available. It implements all of its features well and is innovative, has lots of personality and usually has good replay value on top of that - Brawl, No More Heroes, Elite Beat Agents

My review system would consist of me typing a lot and putting absolutely no scores or numbers on anything.

I would probably do the same, or have a 1 to 5 star system, with 5 stars being anything that would get a 90+ on most other system. I would avoid the 10 controversy that comes with today's system.

Not that I find the review systems today broken, I more find the general gaming fan broken for seeing a score of 10 on a game, and somehow equating that to mean something higher then "hey, it's a great game, go buy it".

I liked the old PC Zone 100 points system, with special awards for the best games and the special "Pants" award for extremely crappy games.