Hobby Game Trends 2000-2014 Round 3: Game Categories

Categories and Classifications

A very common and expected next step in a market analysis is to classify and categorize the data to make comparisons.

At a Glance

Breaking down the release data by game category seems like a relatively simple process, but it gets sticky quickly.

Let’s start with what seems like a simple question, “What game categories are you interested in analyzing?” This is where I started this whole discovery process. I wanted to analyze a category of games often called “Social Deduction” games; specifically ones that implement the mechanics of “hidden identities/factions” and “hidden objectives.” Neither of those categories exists on BGG. Let’s take a look at some of the categorization challenges and what we have available for categorization (short of categorizing thousands of records myself).

The Data

BGG categorizes games by several criteria that are loosely based on the following four classifications:

• Subdomain: This is presumably a base classification (since there are only a few options), but it overlaps considerably with Category.
  • It is noteworthy that this classification is maintained by BGG users through voting and therefore, not all games are assigned a Subdomain.
• Category: This is a loose classification based on genre, intended audience, source, principal components, etc. and overlaps considerably with Subdomain.
  • All games appear to have at least one Category and usually several.
• Mechanic: Obviously, the mechanics in the game. This is what I consider to be the “purest” category of the four.
  • Unfortunately, not all games have a Mechanic assigned.
• Family: This is another loose classification to indicate a family of games that are associated together through any of a number of criteria including: Intellectual Property, Game Series or Expansions, Mechanics, Genre, etc.
  • Many games do not have a Family assigned.
  • There is no means within the BGG Advanced Search functionality to compile a complete list nor to designate or filter on Family.

Without getting too linguistically technical, one of the main problems in analyzing this data is that the established categories cross classes. Due to the limited number of designators (4), those designations are overloaded. Another difficulty is the ambiguous or unexpected use of the terms within a category.

An Example

An example that may help illustrate the challenge in dissecting this data is described here.

Let’s say we want to exclude “lone-guy crap-games” like ones maybe I’d publish myself (geez, I hope not). A Publisher of “(Self Published)” indicates the obvious and a Category of “Print & Play” should do the trick, right? Not so fast. Each indicates a game that has ever been self-published or ever had a PnP version and both are assigned to the production version record for the game. “(Self Published)” gets you games as trivial as Twilight Struggle and a total of just over 2600 games in 2000-2014. “Print & Play” returns BattleCON: War of Indines and several Kickstarter games that have done quite well (a total of almost 3400 games in 2000-2014).

Combining the two criteria narrows the list to about 600 total games, 516 within the Study Games, and once again Cards Against Humanity is right there near the top. This list has only 34 rated games in it, so it looks like an opportunity to eliminate some obscure games from the list. However, since this list is compiled through a negative test, it would have to be removed from the Study Games manually year by year. What a pain, but it might be worth the effort.

One can guess that with the decreased cost of home printing there would be a trend here worth noting, i.e., there could be value in excluding them.  It could also be wasted effort if the trend for the games on this list matches the trend for all games, then it is just background noise.

Ultimately I have left them in the Study Games and will report on some interesting trends within them in a separate article. Hopefully, this gives you a sense of the challenge.

The Categories

Although there has been much debate on BGG forums about suggested alternate categories, these are what are currently available.

Note: The terms in red are the categories that were excluded from this study.

Conclusions

As mentioned several times, the data available on BGG through Advanced Search is problematic. However, it is also a rich data set that can be used to derive trend information for the game industry as a whole and for specific categories of games. The analyst must be careful in dissecting and assembling this data.