By numbers, we mean the style parameters (OG, FG, ABV, IBUs, Color) listed in the Guidelines. The numbers come from many sources, but all are validated by the Style Committee. If you see what appears to be an error in the style parameters in the Guidelines, please contact us as discussed in I think I found an error. How do I report it?
Style parameters aren’t determined by the BJCP through direct measurement. That is, we don’t collect commercial examples and then test them (although we might observe the color of commercial examples). We determine most numbers for our style guidelines using several indirect methods. First, we try to find a reputable source of primary research, such as a book on styles. If the material exists there, we try to use it. Second, we try to obtain style parameters for good commercial examples of the style. This is often quite difficult, since breweries rarely publish all the data we use in our guidelines. We may have to calculate some statistics (such as the FG, given the OG and ABV). Third, we analyze the data and look for trends. We use our judgment in selecting reasonable parameters given the data.
It isn’t our goal to cover all commercial examples in the style parameters. Some examples may have much more or less of some attribute than most other beers. We try to limit the parameters to a range that makes sense for most examples. If we find good commercial examples that don’t exactly fit our style parameters, we tend to make note of that in our Guidelines.
In some circumstances, we may not be able to find commercial examples or good reference works on the styles. In this case, we turn to historical data and secondary sources. For example, the Classic American Pilsner style doesn’t exist today in commercial examples, so we used essays by subject matter experts. Some styles are based on a small number of examples (e.g., California Common). In this case, we use the best known examples to define the style and allow for a little bit of variation. Some styles are defined by beer tax laws (e.g., German and Belgian beers). In those cases, we limit the beers based on what commercial brewers are legally allowed to be brewed.
No matter which methods we use, we try to avoid excessive overlap in the styles. We don’t want to make it difficult to enter beers in a competition or to judge them. We may set the breakpoints between similar styles artificially because we want to draw a distinction between styles and not have their statistics overlap.
There are always exceptions to our rules. We don’t let our rules and processes keep us from using common sense. When we have finished generating the style parameters, we check the calculations to make sure they are possible to achieve. This means we use the OG and FG ranges to calculate the ABV range. If the numbers don’t work, we adjust them until they do. Jamil Zainasheff did much of the work in validating the style parameters in the 2008 Guidelines while he was working on his Brewing Classic Styles book.