Reviews are an interesting subject and I've enjoyed what you have here, Jason, even the memory stuff. As I've stated a number of times in the past, Chessa and I thought long and hard about how we'd be scoring games on this site before it was opened.
At a point we had discussed even the Buy, Try, Avoid system, which we actually thought was pretty good - but there were a few problems. #1, a recommendation of that sort is a little ambiguous for one (who does it apply to, does every good game deserve to be bought, etc). #2, a recommendation is just that, a recommendation. It's not a critical evaluation of the game, it's just what you think someone should do. While your text should absolutely reflect your critical opinion, like it or not, it's a critic's job to provide a rating. #3,you can't get picked up on Metacritic. Despite anyone's feelings on the site, they are a key player in this industry and being aggregated there is extremely important.
Of course, this has resulted in the 5 star + Recommendation system we have here. We never quite go with the Buy, Try, Avoid wording, but try and steer recommendations to tell different players what's worth their money. Some games can hit all three recommendations given their content - Demon's/Dark Souls comes to mind.
Naturally, I agree with you in terms of using 5 stars. I think it's one of the best mediums to dispel a score. It has enough granularity to properly classify a game and leaves little to be ambiguous.
Now, let me share a story. Back when I was a kid, I also read Nintendo Power. If you remember correctly, for a long time they used the 100 point scale to rate games. I was really into this rating scale. In my mind, it helped categorize and order every game out there. The granular scores meant, to me, that I could definitively say one game was better than another and if you were to make a list, x game would come before y game because x got a 95 and y got a 93.
If you remember still, Nintendo Power switched their rating system to the same 5 star scale we use today. I was infuriated. How on earth could I possibly get an idea if I should buy one game over another? Both x and y now have a 5 star score - which one is better?
The fact of the matter is, it doesn't matter, it all comes down to preference because they're clearly made from the same stuff. What I didn't understand and maybe Nintendo Power didn't at first either, is that the 100 point scale may give you the flexibility to score one game over another by a minute amount and that works great for a while, but what happens when two games get the same score and you clearly state one is better than the other? Well the whole ordering system goes out the window, that's what.
Fact of the matter is, it's not our job, in my opinion, as critics to figure out where each game sits in the grand scheme of things. Metacritic should be worrying about that one. It's our job to score the game based on some quantifiable and well defined scale. I appreciate that you use our policy guidelines, as I consult them each time I write a review to help clearly designate how I feel about a game. I can't imagine having the burden of choosing a score out of 100. How do you justify it? What do you say to the inevitable ranking questions?
I think any system, outside of the 100 point scale, can work great as long as the guidelines are in place that define each score. I'll probably write a staff blog of my own on why I think scores tend to skew high and an interesting cycle I've begun to notice just within the PixlBit ecosystem.