The full reviews attempt two objectives: to offer both a description of the flavor components found in the beer and to situate that description in an analytical context. After reading a review, one should have an idea about how the beer tastes and an objective sense of how that beer stacks up against others in the claimed style category. The bulk of the review is dedicated to the former, while the final analysis section (and rating) suffice for the latter.
In general, there are two approaches to critiquing beverages. A beer can be rated on a fundamental goodness scale, which communicates how much the reviewer liked the beer (aka "hedonistic" or "subjective"). On the other hand, a beer can be rated based upon how well it replicates its self-described style (aka objective).
When appropriate, I evaluate beers using the latter system, for two reasons. First, the fundamental goodness approach is somewhat flawed in that it is limited in its ability to translate the review into a common language, and that it is also highly susceptible to personal style-biases or other purely subjective criteria. Second, I believe that the brewing industry ought to be held to a certain truth-in-advertising standard. If I pay money for a beer purporting to be an India Pale Ale, I have expectations that the flavor profile of the beer will fall within certain parameters. These style designators should have a useful meaning to the consumer, and not serve as a marketing cachet. Rating a beer against generally accepted style benchmarks obviates the need for much guesswork on the part of the educated reader.
That being said, I believe that either approach has its place in our community of beer-enthusiasts. Obviously, style can be a rather loose concept, and some styles are better defined than others. When objective criteria are not appropriate, the rating is based more on a hedonistic approach, as is often the case with several Belgian or Dutch "styles". In extreme cases, I do not offer a rating at all (e.g. the over 30 year-old bottle of Ballantine Burton Ale). In light of this, I write my reviews with both descriptive and analytical comments. This allows for those interested in merely the fundamental goodness of a beer to flat ignore the score and my concern with style and still come away from the review with useful information.
All of my reviews include a score at the end. This score is based upon a five-star scale. For the reader familiar with the system that Michael Jackson employs in his Pocket Guide, this system is similar in theory and practice, with one addition. As Jackson's system does not have any room for beers that are fundamentally flawed in some manner, I have added such a rating. Below one can find a brief description of the star ratings:
5 -- World Class
4 -- Excellent
3 -- Good
2 -- Fair
1 -- Poor
Typically, if a beer is not procedurally flawed in any manner (no infection, no handling problems, etc.) the floor for the score is two stars. I reserve scores lower than two stars for beers for which I should not have paid money. In all cases, the type of the flaw (infection, making it a brewery issue, or oxidation etc., making it an issue out of the hands of the brewery) will be noted.
Finally, as this site is one of the very earliest beer review sites on the web, with many reviews dating back to 1994, consider the older reviews to be anthropological artifacts, of limited reliability when describing a given beer in 2002 or 2003. Indeed, many of the breweries reviewed on this site no longer exist, let alone the beers themselves. As I revise the presentation of older reviews, I will highlight the original publication date in red as a warning if the review is older than four or five years.