Last week I suggested that game weight, rated on a 5-point scale here at BoardGameGeek, could be best conceptualized as cognitive load. This is the amount of information that we need to hold in working memory while we are performing a task. Simply put, heavy games require more thinking (information processing) than light games. This week, I wanted to examine the association between game weight and the perceived quality or popularity of a game, as operationalized by average user rating. It looks like there is a pretty substantial, linear relationship between these two variables. As a general rule, the heavier the game, the better the rating.
According to an analysis done by Chris Wray and Jeff Lingwall at The Opinionated Gamers, heavy games and newer games are rated higher than light games and older games. A look at their graph below indicates that the very best games (rated 8+) tend to be relatively recent heavyweights. Really, all you have to do is check out the top ranked games and you can see this for yourself. Illustrative examples include Gloomhaven, Through the Ages, and Terraforming Mars.
Another recent analysis done at Ludometrica found a nearly identical scatterplot, though this one is broken down by weight and playing time, rather than weight and publication year. Heavy, longer games are rated higher, and unsurprisingly, there also seems to be an association between weight and playing time. Take a look at the next graph, which looks like it includes the entire BGG database.
Incidentally, if you haven't checked out Ludometrica, do yourself a favor and click on the link. His principle components analysis of Gamer Genotypes dovetails nicely with the work on gamer motivational styles I've discussed here. Briefly summarized, he found that in addition to a large, general factor loading on all types of games, there were distinct preference profiles for thematic vs. strategy gamers, wargamers, and social gamers.
Now back to the topic of game weight and game ratings. I put together my own analysis exploring the association between these two variables for over 300 games I had data on. My scatterplot (below) replicates the previous two, but I also computed the Pearson correlation between the variables, which turned out to be r = .62, p < .001. The correlation between game rating and playing time was quite a bit smaller, r = .32, p < .001. The partial correlation between game rating and weight rating, controlling for playing time, was only slightly reduced, r = .58, p < .001. It seems clear that heavier games receive higher ratings on BGG, and this relationship does not appear to be the result of other variables, like publication year and playing time.
Why would heavy games be better games? Heavy, meaty, "thinky" games might provide a more challenging and engaging experience with more opportunity for flow. Light games are fine for socializing and diversion, but they tend to be fillers. Often, their main purpose is to kill time while people are waiting for the main event to begin. I don't care how much you love Azul, I doubt you would want to spend the whole afternoon playing such a light game. Truly memorable, epic gaming experiences that motivate high ratings are most likely to come from big, heavy games in big, literally heavy boxes.
Two Objections Considered...
But wait a minute. People do love Azul and they rate it highly, putting it in the Top 50 here on BGG. Personally, I love Bananagrams and that is about as light as a game can get. Don't these two games refute the hypothesis? No because we are talking about a general pattern, and until the correlation is 1.00, there are going to be exceptions. Some light games are quite good, and likewise, some heavy games are utter garbage. The third edition of Twilight Imperium, for example.
Well what about sampling? Obviously this result only applies to people who rate games on BoardGameGeek. The vast majority of the population will not necessarily enjoy heavy games more. Therefore, my conclusion, heavy games are better games may only hold true among hobby gamers, that is, people who love to play games. But that is probably the best group to judge how good a game is, don’t you think?
Whew, all this talk about weights and heaviness has been quite a workout. Next week we will relax a bit and talk about a different topic related to games and psychology.
Statistics and Speculations on the Behavioral Science of Board Gaming
30 May 2018
- [+] Dice rolls