## sábado, junio 14, 2014

### Are we in a boom of the gameboard industry?

And if so, how long is going to last?

If you are into boardgames you might have the impression (like I do) that in the last couple of years there has been a boom in the boardgame industry. My wife, who is not very much interested in boardgames, has noticed that as well, which makes me think that this impression might not be just me. Just compare the number of boardgames available in the 80's or 90's and the ones available nowadays. From all published games since 1970, 50% were published in the last 10 years.

There might be a certain bias in my perception because in the 80's, as a kid, I was not paying the same attention to boardgames as I am doing right now. However, consider the available myriad of resources out there that seem to increase in number (more details in the reference section at the end of this post):
• Books on game board design
• Podcasts and youtube channels devoted to boardgames
• Crowdfunding campaigns (especially Kickstarter)
• Board games review services
• Board game publication services like Print-on-demand (POD)
• Cafes with a large assortment of boardgames to play
All this explosion of the boardgame ecosystem doesn't make you think of an industry boom? If we are indeed in the middle of a boom, the next question that comes to my mind is, how long is it going to last?

Doing data processing for a living, the natural way that I have to answer this question is by using data and mathematical models that explain the underlying physics (or economics in this case) of this data. In short, the model (adjusted using BoardGameGeek data) indicates that the number of published board games will reach a peak around 2023 and then will start to decrease (and the boardgame industry would start to saturate).

This is a forecast driven only by data, with no input from experts of the boardgame industry such as designers or publishers, therefore this prediction might be accurate to a limited extent. In case you are curious on how these results were obtained, you will find the nitty-grity details on my hypothesis and assumptions below.

What do you think about these results? Do you think we are close to a peak in the boardgame industry? I'd love to know your opinion.

### Data analysis and fitting model

The main working hypothesis that I applied in this analysis is that the boardgame market behaves similarly as other products (such as technological gadgets) or ideas. It that is the case, then the well-stablished theory of diffusion of innovation can be used. This theory states that initially a small group of innovators will adopt the product and "spread the word" to a wider audience. This causes that the consumer base increases until a peak is reached and then starts to decay as the number of adopters exhaust, reaching a saturation of the market.

For an accurate analysis, the yearly sales of the boardgame sector are needed. However, I had no access to such data and, therefore I used as a proxy the number of published boardgames per year reported in BoardGameGeek instead, being the main assumption that both parameters should be highly dependent on each other.

The market share evolution from start to saturation is then modeled using a sigmoid function according to the Theory of Diffusion of innovation stated above. This function is monotonically increasing and goes from 0 to 100% (or the market size) with the following shape:

 source: Wikipedia
To control the size of the market, growth of rate and origin (i.e. market peak) the sigmoid function has been modified as follows:

$$F(t) = \frac{M}{1+e^{-p \cdot (t-a)}}$$

where:
• $$M$$ indicates the market size (in our case, is the total number of published board games at the time when the market saturates)
• $$a$$ is the time at which the peak of the market is reached (in this case the maximum number of boardgames published in a single year)
• $$p$$ is the rate of market growth rate
Note however that the data used for this analysis is not the size of the market up to a given time (that is what the sigmoid function represents), but the number of boardgames published each year. Therefore, instead of $$F(t)$$ we actually need its first derivate ($$f(t)$$), that indicates the number of new adopters in each year (or, in our case, the number of new boardgames published each year). The expression for the derivative is:

$$f(t) = \frac{dF(t)}{dt} = \frac{M \cdot p \cdot e^{-p \cdot (t-a)}}{\left ( 1+e^{-p \cdot (t-a)} \right) ^2}$$

That is the function that has been used to fit the data and the one represented in green in the plot above. An additional assumptions that has been made is that the data after 1990 has been given a higher (5 times larger) weight than the data before. The rationale behind this data weighting is that BoardGameGeek was up and running from the year 2000 and the uncertainty of the data before that point might be larger than more recent data.

### References

This section includes some references to the resources mentioned earlier in the post. This is not intended to be an exhaustive list of resources on board games. The entries are arranged alphabetically by categories, not by any sort of relevance ranking:
• Web pages on boar games
• Books on game board design
• Podcasts on game boards (included in
• Youtube channels on
• Platform supporting board game development and publication
• Boardgame cafes