When I review books or films or products, I tend not to award very many five-star ratings, causing a few people to inquire what my criteria are for rating things. Generally, I view the written review itself as explaining the reasoning behind each individual rating. However, I have decided that the meaning of the numerical values I assign could use some explanation, so the following are the basic criteria by which I make my ratings.
<1: and="" br="" flawed="" generally="" is="" not="" product="" recommended.="" seriously="" this="">
For example: A book which receives a rating of 0 might be so poorly written that it can't be understood. A book which receives a rating of 0.5 would be recommended only if it is the only way to acquire some technical information the reader might want.1:>
1: This product is seriously flawed and is not recommended for most audiences, but has sufficient redeeming quality to make it worthwhile for a tiny minority of audiences.
For example: A book which receives a rating of 1 might find itself and the proper level of poor writing that it is (barely) readable, but so bad that it's low quality itself provides entertainment value. Alternatively, a book that is generally bad but has a particularly good chapter or an interesting thesis might warrant this rating.
2: This product is flawed, but may appeal to some audiences.
For example: Non-fiction books which are a pleasure to read but contain significant errors will likely receive a rating of 2. Novels might receive a rating of 2 if they have little literary merit but might still be enjoyable to certain readers. Note: a rating of 2 means a work is somewhat below average to average (bearing in mind the old chestnut that most of everything is crap). They're below average for my personal collection, but about average for the entire marketplace.
3: This product is good, and is recommended.
The products that receive a rating of 3 can be trusted to be mostly accurate, well constructed, and worth purchasing. There may be some flaws, but they do not detract from the overall experience of the product. Note that a rating of 3 does not indicate a work of "average" quality. While I tend to primarily rate better works (making my average ratings higher), I would consider a 3 to be a product well above average.
4: This product is very good and is strongly recommended.
These are considered to be among the best of the best products. Flaws are minimal.
5: This product is near-perfect. Purchase it immediately.
I reserve a rating of 5 for a very small number of products. It is generally safe to consider a rating of 4 to be highest marks because a 5 is reserved for those products one might call transcendent.
Intermediate scores (eg., 3.5, 4.5) should be considered to fall within the broad category of the number preceding the decimal. For instance, a 4.5 is meant to be a product that is a bit better than a 4, but it is still considered to be in the "family" of 4-rated products. It is closer in meaning to a rating of 4 than it is to a rating of 5.
Occasionally, a work might receive two different ratings. While I try to avoid this, it is necessary from time to time because a work has two distinct audiences who will find it to be of different value. When this happens, the ratings follow the same criteria as above, and the distinction of which rating is which will be made clear within the review.