Reflections on Catan
I have been playing Catan for at least twelve years (possibly more) and it remains one of the most remarkable games I’ve ever played. It has been one of the most influential games, to both the industry and myself, professionally and personally.
I doubt that it is ever going to change.
Fellow blogger and game enthusiast Sarah Trager recently declared Catan was still her favorite game after all this time, an act which one again forced me to consider my own journey with the game over the last decade or more.
Professionally, working on Catan was one of the hardest and most rewarding aspects of my career. I began just before Catan first became a phenomenon. My professors were unfamiliar with the game but loved the idea – crafting AI agents that mimic human emotions when playing a board game, especially one not previously studied – and this meant that I was on my own with little backup. Catan’s intricacies were left for me to discover and conquer alone; I remember struggling with the Longest Road algorithm for three months because there was seemingly no way to implement it such that it could be used as a factor in every possible move evaluation. The calculation simply took too long to handle every case. Then I had a dream about a volcano erupting on a Catan board, got my inspiration for an optimization and solved it the next day. This was not even part of the thesis, mind you, yet it was essential for creating competitive AI. I created five different personalities, including a vindicative player who kept track of which player hurt them the most and specifically targeted them in non-game threatening situations. I had to think of every crazy scenario and program for all possibilities to make my AI as powerful and competitive as possible. As a result, I got really good at the game and yet my AI still beat me routinely. Statistically, most agents failed to have a runaway leader when looking at average overall scores, and I was quite proud. My advisors felt that they needed to create a “With Distinction” category to separate my work from everyone else, the first in the University’s history. It would have been my dissertation had life lead me differently.
Even if my Board Game Bandit career is a colossal failure, I will always have my Catan game – a reminder that I did at least one good thing for the gaming industry.
Personally, I became a new person entirely. For the first time in my career, I felt like I had done something meaningful and I wanted more. I wanted to become part of the gaming world, so I dove in headfirst. Catan gave me a new hobby, lead me to make new friends, gave me confidence to apply for a new programming job and has nothing short of transformed me into the person I am today.
But the last year or so has not been kind for Catan and I. I teach the game far more than I play it and I worry that my experience has been irrevocably altered. When I chose to study the game for my Thesis, my gaming life was very different from what it is now – practically non-existent. I had played only a handful of modern games; I was years away from having any semblance of a gaming group and knew very few people that I saw with any regularity. I had no idea that there would be an entire lifestyle awaiting me, what studying the game would do to my experience playing it or the person I would become as a result.
But because of that project, today, more than ever, more than any other game, everyone wants to beat me in Catan.
For the longest time my insider knowledge created an advantage over other players that would allow me to play competitively and win consistently despite everyone teaming up to keep me down. In Catan, it is fairly easy to ensure one person loses – don’t trade with them and consistently block them with the robber. I used to be able to overcome this by maximizing my efficiency with every move. One of my favorite games of Catan was using the five-six player extension (which I actually hate, topic for another day). I was invited to a Catan party … with twelve people. This meant that everyone was divided into teams of two for a game where having a second set of eyes isn’t really an advantage. In fact, for the few times people did trade with me, I had to convince two people rather than one, which made my task harder. My own “teammate” was sympathetic to their cause and didn’t want to be there and actually tried to sabotage my efforts. All five teams targeted me almost exclusively with the robber, and yet I still won, and it wasn’t even close.
Those days seem long gone.
Admittedly, I don’t play the game much anymore because I’m inviting frustration upon myself. I’ve worked hard to train myself so that I can consistently play almost any game at a high level and as a result I will always be viewed as a threat in every game, regardless of my score or situation. I accept this. But with Catan, once someone knows my history with the game, the desire to beat me inevitably means I am playing with a severe handicap. By not playing often, I’m not as sharp in Catan as I was in my prime and, as the years went by, I forgot minor things about my thesis, whittling away at any advantage I had. But this doesn’t mean that players will alleviate their attack on me.
I tried to salvage the game by playing only with players who did not know me personally, but in the few games I played my secret concerns about the game surfaced time and again. The flaws of Catan, in this small sample size, could not be overcome with even the best of strategies. I’ve always believed that Catan had three major flaws with Catan, some of which have no solution, but until now I had managed to overcome them.
First, while the game is very strategic, there is no defense against the dice themselves. I realized this early on when I was coding Catan; I began to integrate analytics to look for trends among victories. A sad truth emerged – whenever a number was rolled disproportionally higher than any other, the player who had the most settlements on that number usually won. This makes sense on paper – having more resources, even if they are of a singular type – creates more opportunities. There are at most six opportunities to have access to a number and possibly less if settlements are placed in a way that restricts building on many locations. Players have no rational reason to assume that any one number will be rolled ~25-40 percent of the time, and they certainly could not predict which one it would be should such a phenomenon occur. Yet, this happens (it came up routinely during simulations) and there is nothing we can do to stop it. Except maybe moving the robber – speaking of which.
The robber is a bit too powerful. A seven has statistically the highest chance of being rolled at 17 percent. However, it also means that it there is an 83 percent chance that it will not be rolled on a given turn. Before the end of the game, when you need to block the big, obvious payout, the robber tends to be used incorrectly by newer players by focusing on the number, not the type. If you block someone’s only grain production, it severely hampers them. They lack the grain to build the development card that obtains a Knight to move the robber off their property, while also restricting the settlement purchase to expand into new lands. I’ve played games where I was the unlucky recipient of a robber early in the game and he never moved for most of the game, because no one bought development cards and a seven was not rolled for some time. Or if he did move, he was immediately moved to another location that hurt me and I lacked the means to defend myself. Add in the fact that I have a resource stolen from me and you can see how this adds insult to injury.
My friend Mauricio once tried to introduce me to the “Friendly Robber” rule, in which you could only harm someone with the robber if they had more than two points. My problem with this variant is that there isn’t a clear delineation as to when the robber should be allowed. Someone winning 3-2-2 is hardly “winning” in the beginning of the game and, as mentioned before, the robber might linger there indefinitely. Imposing it too late in the game however would fail to hinder a runaway leader. Perhaps the game should have included a one-time token to force the robber back to the desert to allow players a lifeline, or, alternatively, allow the robber to expire after some time. Under the current rules, however, the robber is too powerful.
The final flaw in Catan is one created by experienced players attempting to play under advanced rules – the modular board isn’t always useful. Of the terrain hexes, only three are brick and, when placed randomly, can produce some undesirable maps where brick’s scarcity increases severely. I have seen the same phenomenon occur with Ore, as it prevents the purchasing of development cards and no settlement can be upgraded to a city. Additionally, the rules suggest that, if you want to, you can place the number circles randomly as well (as long as no red circles touch) but this can further create areas of imbalance across the board, creating a draft advantage for specific players and possibly bifurcating the board into a weak and strong side. When these two phenomena are combined, Catan devolves into a terrible game of luck; the one player who has access to brick has little incentive to trade it and any attempts to block it with the robber slow the game to a grinding halt. At this point, the game becomes about what development cards are drawn and Catan loses all its appeal. Personally, I believe there should have been one less sheep tile and one more brick, as it would have expedited building and streamlined the game. It does not make sense to me that brick and wood would be equal in terms of building materials yet one is rarer than the other. When players complain that the game takes too long, I believe artificial scarcity to be the source of this, which prolongs trades as players are less likely to give up resources so difficult to acquire.
Despite its flaws, Catan remains an excellent game, even to this day. There is a reason why I still teach it so frequently – my current record is four games simultaneously, because I hate myself apparently – and I often think about why I want to share it with everyone. It is not because the game has universal appeal, because it does not; I’ve taught the game enough to know that there is a certain subset of humanity that will always deem it too complicated. But it comes really close. Most gamers who have evolved beyond it certainly enjoyed it when they first started and it’s OK to move on from a game. Catan solidified a paradigm shift for the entire board game industry, moving away from American-based player elimination and luck-based games, instead focusing on deeply strategic gameplay and forced partnerships. It introduced real, meaningful trading to generations of gamers and revolutionized how dice were used. The rules were streamlined; player aids and a thorough rulebook made the game easily explainable (yet people still get specific rules incorrect all the time). I am not certain if Catan was the first game to use a hexagon tile, but every game that came after was clearly inspired by it. The game’s influence directly created publishing powerhouses like Z-Man Games and Rio Grande Games (which was created by Jay Tummelson, who was pivotal in bringing Catan to the United States). There are entire convention areas dedicated to this single game.
Even if Catan became mechanically irrelevant (which is years away, if it even happens at all), it remains a valuable historical artifact that should be experienced by the masses. If Board Gaming ever becomes large enough to have games for historical preservation, this should be first on the list.
I often talk about my amnesia games – games I’d be willing to suffer catastrophic memory loss to experience again for the first time – and Catan is top of the list. Though my experience with the game may never be the same, I wouldn’t trade the life it created for me.