Especially if you're being taught by someone else and haven't personally absorbed the rulebook, Terra Mystica is a rather complex game – more complex than Tigris & Euprates, I think, the game that in my mind sets the bar for about as complicated a game as I want to have to explain from scratch. Terra Mystica has a lot of moving parts – cult tracks, magic bowls, buildings, towns, fortresses, favors, priests, terraforming, and so on. If someone is sitting there explaining the game to you, it is not immediately apparent why this sucker needs to be as complicated as it is. Plus, then, everyone gets their own race with their own properties and customized player mat, and at this point your head possibly explodes. Good luck if your game explainer is not particularly deft.
On an absolute scale, of course, Terra Mystica isn't that bad. I play ASL, after all. But Terra Mystica is an abstract eurogame. When you look at it from the perspective of sitting down to play one session, as a design it seems to really only have two thematic elements: managing a diverse economy (the 5 different building types do different things and have different building/upgrade costs and produce different combinations of resources), and managing the cooperative/competitive tension of wanting to have neighbors (because it helps you generate magic, useful for a wide variety of purposes) and yet not be constrained or cut off by them (because you are managing an expanding settlement). Really, that's about it. There are of course intricate details to all this, but in most cases they look like VP-optimizing puzzles rather than expressive game systems.
Anyway, that was my impression on first playing it, and the net effect wasn't particularly positive. It just felt overwrought. However, my opinion of the game improved when I understood Terra Mystica isn't really a game best judged on one playing. As I mentioned, Terra Mystica has many – 14 – different playable factions. These are not slightly different player positions. They are very different, more divergent even than the alien species in Eclipse, and this is why the game is as complex as it it is. Without that level of inherent system complexity, it's hard to imagine how you could cleanly support such a wide array of different factions. In fact, Terra Mystica accomplishes its impressive diversity generally by efficiently parameterizing the game's various systems, not through special rules. As a way to go, this is a pretty good one.
For me, this is where Terra Mystica succeeds: in providing a rich exploration experience. Every time you sit down to play a new faction, it's a different game and a different set of challenges. The Swarm and the Witches and the Engineers all play very differently and exploring these different points of view can be powerfully engaging. Eclipse and Terra Mystica are similar designs in many ways (even if the end effects of those design techniques are quite different), and this is one area in which I think Terra Mystica does better.
However, embracing this extremely high degree of asymmetry implies trade-offs. Terra Mystica tries for replayability and variety solely through the different factions and their interactions. Otherwise, there is no luck too the game, no hidden information, nothing that is not on the table before turn 1. Even your faction is not assigned randomly, but chosen in player order (although random allocation house rules seem not uncommon). While some will see this lack of any variability or uncertainty as a feature, it can make it very hard for a game to retain interest in even the medium-term as gameplay can very quickly stereotype absent countervailing forces. As great a game as it is, 1830 is dead to me now because the game space has been mined out.
One question then becomes, how much player interaction is there in Terra Mystica, really? Can the very different factions produce variability through their complex interactions? Unfortunately, I think the answer is: not to the degree it needs to. The board is a field of hexagons in 7 different colors, each corresponding to two factions, only one of which can be in play. As players' empires expand on the board, they are limited to developing on hexes of their color. Developing on other color hexes requires a process of terraforming, initially quite expensive although probably getting cheaper as the game goes on. So during the vital first half of the game, there really isn't much competition for space. Competition could theoretically get tighter as the game goes on, but in practice factions seem to develop enough tools to go around and real resource or space competition seems fairly infrequent.
So player interaction seems fairly light (and if you think about it, that makes sense – to properly ensure some sort of balance between 14 very different factions and all their potential interactions might require a vast investment in development). So you're left with a faction with a specific set of parameters set in an environment locked down before turn 1 and limited player interaction. That means there is an ideal way to play that faction, more or less. You just have to figure out what it is. In a game lacking any randomness and not inordinately complex, at least the broad outlines of that perfect plan should not be too elusive.
This is not necessarily a problem in the short term, when finding those plans amongst the intricacy of the game systems can be engaging, but at the end of the day it means that Terra Mystica can only be a game of learning the right general techniques for each faction and then squeezing out fairly small efficiencies in the margins. It reminds me of the things I didn't like about War of the Ring or Through the Ages: for various different reasons, there is really only one viable way to approach both games, and you win or lose not on strategy or tactics or evaluation, but on ruthlessly going after every small advantage you can find on the way to that strategy. Fortunately for Terra Mystica, instead of one way to go, there are 14 different ones, which will take a while to figure out and significantly extend the period of discovery.
It should also be mentioned that because learning the game's tricks is so important, and because it's pretty complicated, Terra Mystica is extremely punishing of experiences differences. People who have played only a few times will have no chance against more experienced players, to an unfortunate degree. Race for the Galaxy and 1830 are other examples of this sort of game, but my feeling is Terra Mystica is much more punishing and less fun for new players to play with veterans even than those games.
People who have played Terra Mystica will note that I've glossed over a few things in this analysis which might appear to be mitigating. For example, on each of the 6 turns, there are point bonuses available for different game actions (building dwellings, trading posts, terraforming, founding towns, and so on). These are randomly assigned before play, making the game's initial state somewhat variable, and so could theoretically encourage different game rhythms. If the bonus for building fortresses is on turn 2, you might want to change your plan to put off building it until then and build your dwellings on turn 1. In practice, it seems different factions have different imperatives. The Giants, for example, are in a hard spot until they build their fortress and they probably need to slap it down as quickly as possible regardless. So rather than giving the game variability, the different bonuses seem instead just to give bonuses or penalties to different factions, which complicates the evaluation of which faction to pick. Once the play gets started, the factions have to do what they have to do and having to bend to accommodate different turn-to-turn bonuses just makes their job harder.
All this may sound like I don't like Terra Mystica, but that's not true. I think it's more accurate to say I do enjoy it for what it does well, but even now, after only a handful of plays, the obvious limitations of the design are closing in. I enjoy the game when sitting down to play a new faction that I haven't played before, and building the right economic base and evolving it as the game goes on is an engaging little challenge. In the short term, while the experience of the game is biased towards system exploration, there is a lot for me to like. As the balance tips away from exploration towards rote execution, I know it's going to be far less appealing. I'm still a ways away from the point where the game becomes tedious, but I can see it pretty clearly from where I am.