Wargames – increasing complexity?

A caricatured view of the history of wargames suggests that there was a Golden Age when there were very few wargames around, and they were relatively simple to find out about and get to grips with. Then, in response to the demands of the wargames community, in particular the “grognard” community, many more wargames were published, and they became more and more complex. So, at some point, wargaming had declined (is still declining?) from the halcyon days when the average wargamer would potentially know about all the offerings available. Potential new wargamers are put off by their complexity, and by the more knowledgeable (and older) grognards. Now, we have a shrinking community, especially in historical wargaming, despite vast and increasing numbers of titles.

A couple of examples leaning into this caricature will, I hope, suffice. From an article The Re-popularization of Commercial Wargames by Maurice Suckling on the Ludogogy website https://www.ludogogy.co.uk/article/the-re-popularization-of-commercial-wargames/ 8 November 2020:

“In this period [the article mentions 1990s to early 2000s, but the 2 examples quoted are from the 1980s – ed] we also see games becoming more complex and ‘denser’, with behemoth rulebooks, becoming longer to understand and play, with play times extending to 20 hours, and, more realistically, far beyond. Games in question are such as The Civil War: 1861-1865 (1983) and The Korean War: June 1950-May 1951 (1986). From BoardGameGeek (BGG) scores it seems these games were well enough received at the time, but they did a poor job of introducing new players to the hobby. As designer Jason Matthews has described it: “War-gaming destroyed itself.”

The article is an excellent summary of the rise, decline, and resurgence of board wargaming from the 1950s through to now, and I recommend that you read it.

And secondly, from a recent interview of Philip Sabin by Harold Bucanan and Volko Ruhnke at the San Diego Historical Games Convention (12 November 2021): https://youtu.be/luHPZUbF4fY,

Professor Sabin said: “In some ways I think that the changes that have occurred since then [Jim Dunnigan’s 1971 Napoleon at Waterloo – ed] have not always been for the better. Particularly the way in which complexity has spiralled. This is well-known, and it was happening early on, it was happening in the ‘70s let alone recently, with the grognards wanting more and more complex games and thereby excluding newcomers.”

Again, I strongly recommend that you listen to the interview, as all three participants, especially Professor Sabin – he was the guest after all! – offer fantastic insights into the phenomenon of wargaming.

But...

How much of this complexity problem is real?

My attempt to address this problem starts with the complexity rating used by the popular board game website boardgamegeek.com (BGG), which rates games from 1 (light) to 5 (heavy), using the terms “weight” and “complexity” more or less interchangeably. The database calculates each game’s weight or complexity rating by multiplying the number of votes in each of the 5 categories by its integer rating (1 to 5), then dividing the total by the number of voters. For the games mentioned above, Napoleon at Waterloo (published in 1971) has a BGG complexity rating of 1.90. The Civil War (1983), in contrast, has a complexity rating of 3.81, The Korean War (1986) 3.68. It’s the playing time in both of these last two games that seems excessive (120-1200 minutes, and 180-960 minutes respectively). The famously “most complex board wargame of all time” is Richard Berg’s The Campaign for North Africa, coming in at a massive 4.70, and a playing time of 60,000 minutes; it was published in 1979.

I don’t want to get too far down the definitional rabbit-hole, though I note the Oxford Dictionary definitions of complexity: 1. Consisting of many different and connected parts. 2. Not easy to understand; complicated. Those seem to sum it up pretty well. In any case, the complexity ratings quoted here are from BoardGameGeek and are entirely from within the minds of the users, who are asked to interpret “Weight / ‘Complexity’ Rating” defined as “Community rating for how difficult a game is to understand. Lower rating (lighter weight) means easier.” So, BGG has opted for the second definition, but users will interpret it in whatever way they wish (some will not have read the definition on BGG, of course!).

It occurred to me that with the BGG database, we have a mechanism to investigate the issue. There are obviously caveats. BGG includes primarily board games, so some miniatures rules may not be there, though a cursory search suggests that very many are included. I’ve already noted that ratings are subjective, but at least there is an element of consistency here – it’s always up to the users. Also, although the database includes published games, it includes self-published games, and many will have had small circulations. I take this as a very big positive, because it means the database is more representative of the games in circulation; it does not just concentrate on the popular ones.

So, I pulled together some data into a spreadsheet. For ease of data collection – by which I mean using the BGG advanced search over and over, in the absence of a data endpoint for something more sophisticated – I limited myself to searching 5-year periods starting at 1970 and ending at 2019, by complexity range (1 to 1.99999, 2 to 2.99999, and so on up to 5, the maximum) with the board game subdomain filter = “Wargames”. For any techies out there, there’s an XML API on BGG, but it is limited to string searching and extracting user collection information, so is not helpful for this little project. The complexity filter can be used to produce results where the mean complexity level is anywhere from the lower level to the next one higher; for example, from 1 (the minimum) to just below 2. This has the side effect that level 5 includes only those where all users rating the game have indicated the level as 5, “heavy”, which is in most cases a small number. I could have merged those with level 4, but that feels wrong, as the scale is actually 1 – 5, each integer having its own definition (light, medium light, medium, medium heavy, medium).

I counted the number of wargames at each of the 5 complexity ranges (1 = 1 to 1.99999, 2 = 2 to 2.99999, and so on) within each 5-year period, then used the spreadsheet to give me the total number of wargame titles in each 5-year period. From this data, I calculated the mean complexity for each period, and the overall mean complexity for the 50 years from 1970 to 2019, which is, by the way, about 2.46 for the approximately 7,500 wargame titles in the BGG database.

The number of wargames in each 5-year period has grown since the end of the so-called heyday; perhaps the heyday’s end was 1970-1974? That was the period of my own time in senior school; I left for university in 1975, and for the last few years of my schooling, our wargames club had an S&T subscription, as well as playing figure wargames and other board wargames frequently. 1971 saw the publication of Napoleon at Waterloo, mentioned above by Phil Sabin as part of the heyday of the popularity of wargaming.

Here is a graph of the total number of wargames on the BGG database by 5-year period from 1970 to 2019:

This graph shows the growth in the number of titles, but note that this says nothing about sales or popularity. I would guess that the growth from the turn of the millennium reflects the reduction in the cost of entry into the market place from the advent of desktop publishing, then the internet. Note the drop-off after the financial crash of 2008.

Let’s have a look at the complexity ranges. A broad-brush review using simply 1, 2, 3, 4, and 5 shows this pattern for the whole 50-year period:

This table shows the preponderance of games rated in the ranges 2 and 3, with relatively few at 1 and 4 or higher, though it’s worth pointing out that ‘relatively few’ still means 750 games in range 1, and a similar number at 4 or more. And I should mention again that 1 is the range 1 to 1.99999, and so on up the scale. I have assumed that the games are more-or-less evenly spread throughout each range.

A little tabulation of the 5-year period data for the mean complexity rating:

• 1970-1974 2.19

• 1975-1979 2.40

• 1980-1984 2.41

• 1985-1989 2.51

• 1990-1994 2.55

• 1995-1999 2.66

• 2000-2004 2.49

• 2005-2009 2.38

• 2010-2014 2.34

• 2015-2019 2.55

And a graph to help to visualise this:

It’s worth pointing out that the y-axis only goes from 2 to 3, in order to focus in on the differences. There has been an increase in the complexity of wargames in the BGG database over the whole period, but only from 2.19 to 2.66 at its peak, a bit less than 0.5 of a complexity level over a 30-year time span. So, there was an increase in complexity from the early ‘70s into the late ‘70s and beyond, but only a small one, an increase of 0.2 of a BGG rating, then gradually up to nearly 0.5. More interesting perhaps is the variation within the 50 years, with a peak in the late ‘90s and another climb in the last few years. However, it’s quite possible that the variations from the mid-70s to now might be explained by random variation around a mean.

I’ve not got enough data to examine statistically the likely influencing variables on this albeit small increase in complexity. A quick linear regression against number of titles and against date doesn’t show up much of a link, as expected. However, I think it’s the growth in the number of titles over the period that is most striking; much more striking in fact than the increase in complexity. Here is a series of graphs for the growth in the number of wargames at each complexity range from 1 to 4. I’ve omitted level 5, because there are very few, and it isn’t a range, just a single level 5 rating. I’ve kept the y-axis scale the same in each graph for ease of comparison.

As we can see from the slope of the trend lines, the most marked increase in numbers lies at complexity range 2, so the lower mid-complexity range rather than the higher. However, all ranges have seen an increase until 2005-9 with a slight reduction since, with the exception of level 3, which has seen more of a resurgence in the last 5 years.

I’ve noted above that this bit of statistical analysis says nothing about sales or popularity. I made a brief attempt to review the analysis on the basis of popularity by carrying out the same data collection and analysis but limiting the data to only those titles with 500 or more votes on the BGG database, where “votes” is the number of people who have rated the game. This is a proxy for popularity, though of course it only covers BGG users, a population skewed towards players in the United States, and skewed towards relatively tech-savvy players who are motivated to record their views in this way. 500 or more votes is a fairly arbitrary figure, but it gives us 400 titles to review out of about 7,500 in total, and the likelihood is that these are the more popular titles; this might approximate to sales, but I’d be very cautious using BGG as a proxy for sales.

The patterns of the total numbers are similar for the 500+ sample and the total titles graphs. The range 3 and 4 complexity figures are slightly different in the 500+ sample, showing a major dip in the early ‘90s and no increase in the number of titles in the last 5 years from the expected dip after 2008. Although the mean complexity for the 500+ sample is 2.49, slightly higher than the mean for the total titles, the mean complexity figure starts at 2.54 in 1970-74 (from a total of only 13 wargames from the 500+ sample for the period), climbs to a peak of 2.82 in 1985-89, 10 years earlier than the all-titles peak, and then fades back to around 2.4 for 2000-2019, slightly lower than in the 1970s.

What can we conclude from this brief summary of ratings on BGG?

It seems that the complexity of wargames did increase from the early 1970s into the 1980s and 1990s, but that this increase was not particularly large and fell back in the 21st century. In addition, there was no shortage of simpler wargames, as the number of wargames at each complexity level increased significantly till 2008-9. It was not the case that the number of complex wargames increased to the detriment of simpler ones. Although the number of titles reduced after the financial crash in 2008, it was sustained at an average of around 200 titles per annum through the first couple of decades of the 21st century, spread throughout the continuum of complexity with more than half of the titles being on the lighter side of what BGG describes as “medium” weight.

The short review of the 500+ sample suggests to me that the more popular games in the ‘80s were slightly more complex than those recorded for both the 1970s and the whole 50-year period, lending a little more credence to the idea that wargames were perceived as “getting more complicated” in the 1980s. But this phenomenon was transitory in the more popular wargames, and by the 2000s the average popular wargames’ complexity may have fallen back to below the 1970s level. However, at no time was there a dearth of relatively light games (levels 1 and 2), compared to relatively heavy games (levels 3 and 4).

Bearing in mind that BGG entries and ratings for recent wargames are likely to be more frequent than for those wargames that go back into the 20th century, there does seem to be an increase in the number of popular titles in the period 2000-2019, compared to the period 1970-1999. This factor, the increased availability of wargames of all types, simple, medium, and heavy, seems to me to outweigh the perception that wargames in general have become more or too complicated.

Perception is influenced by what folks know about. With so many hundreds of wargames out there, it’s pretty much impossible to know anything about all of them, let alone to know all of them well. So, perhaps the difficulty here is that many people, including commentators, pundits, and analysts, myself included, tend to focus on those they know and have heard about. And perhaps these come across as popular and complex. I would say that it’s always a good idea to look across the whole range of boxes that are out there; there’s something for every taste.