The thing is that in the very beginning we must set a goal about: What is the intention of such an award?
Is the goal
A- to have as many people participate as possible?
B- to motivate game companies to link to this site (vote for us over here!)?
C- to promote indie games?
D- to have indie games compete with other games?
E- to reflect the staffs opinion?
F- to give a realistic reflection of the RPG watch community's favorites?
G- to give a reflection about what the RPG watch community has actually played?
Now the most Gaming Sites either go for Community participation which of course is commercially the best option or for reflecting the staffs opinion as this is less effort than the communities opinion and gives a very mixed result as well where some underdogs will show up.
Once you are sure about what you want it's time to find a method how to vote.
It's just logical that B will not work with F for example.
Personally I would prefer F (to give a realistic reflection of the RPG watch community's favorites). But to get a realistic reflection will be complicated and there is no way around it.
I think the suggestion HiddenX did was pretty good and will already reflect the reality quite well. To make the reflection even better you would also need to include the question whether you consider playing a game or not.
If 5 people love a game and vote it as favorite, but 95% don't even try it because it's obvious they don't like it, this game shouldn't get a 100% score.
To get a result which truely reflects the reality it's probably best to set up a survey via surveymonkey for example.
There you can make checkboxes besides each Game and let people fill out one check boxes like
Would result in a easy to do (on the community side) survey with very representative data.
Now all you need is a key to interprete this data.
For example +10 for liking it, -8 for not liking it, +3 for wanting to check it out, -2 for not wanting to check it out.