Game analysis: A Tale In The Desert

I’ve been meaning to try A Tale In The Desert for a while now, since it’s a fairly successful MMORPGs with a truly deep crafting system (the other one being Star Wars: Galaxies, which won’t install for me anymore). I just downloaded and played the starting-island portion over about three hours. Unfortunately the UI is really annoying so I won’t put more time into it.


  • The crafting system is indeed deep. Basically the entire game is about making stuff. You gather resources (modeled nicely and placed throughout the world) and build tools (also nicely modeled) to transform them into ever more complex products. The starting-island tutorial is basic but I read some fascinating things about advanced crafting on the wiki. For instance, you can grow grapes and make wine, but first you can try cross-breeding grape strains to get seeds with better growing properties, and during the growing season you take a series of actions to “tend” the field, which affect the final quality of the wine produced. (“Product quality” seems to be a good way to add depth to crafting – SW:G uses this too – but obviously it complicates the server since it can’t treat all instances of an item type as fungible).
  • No combat, although there is PvP in the form of competing for achievements. ATITD proves that there is actually a good middle ground between purely social “chat rooms” and purely combative PvE/PvP games.


  • It’s a grind.

The resource-gathering system is cumbersome. For example, early in the demo you have to gather slate to make stone blades. Slate is found dotted along shorelines, but there is no indication of exactly where a slate node is located, except for an icon that appears only when you are standing right on top of one. Gathering slate is a process of walking around randomly until the icon flashes up, then backtracking to find exactly where to gather it. There should be a visual display of some kind to show where the slate nodes are.

Another resource you need early on is grass. You can pick grass anywhere the ground is green, but each time you pick up a handful you have to wait a few seconds for the character to animate, then move aside and click again. The gathering animations should be interruptible. Or, if the goal is to limit the rate of gathering, then you should be able to “batch-gather” a large amount of the resource with a single mouse click, like WoW’s “make all” button. (edit: apparently there is a way to automate grass-gathering during times you are offline, but you first have to gather 2500 units by hand!)

Along the same lines, there is no automation for “processing” steps like cutting wood into boards. You sit there clicking, waiting for an animation, clicking, waiting for an animation, etc. Worse, the blade breaks every so often (for no apparent reason), sending you on another shale-finding quest. I can understand exposing players to these mechanics temporarily, but not on a long-term basis. You should reach a skill level that allows some automation or batch-processing, or at least invent a little point-and-click game (with decreasing difficulty as your skill increases) to control the frequency of success and failure.

  • ATITD uses its own widgets and controls for the GUI, but implements them poorly.

Here’s my take on implementing your own GUI widgets, as opposed to using the native widgets in Windows or OSX: it’s like asking for a bout with the heavywweight title-holder: the rewards are good if you succeed, but you’d better have the skills to back up your challenge.

If the native widgets aren’t ideal for your intended use, then go ahead and write your own custom widgets, but be careful and design them well (good examples are the custom GUIs in certain 3D apps like Lightwave and Houdini).

Just keep in mind it’s incredibly tough to correctly handle all of the corner cases and behaviors that users expect from modern widgets. Sure, your homemade button works when I click it, but what about keyboard focus? How does it move when I scale the containing window? Does it work well on a tablet PC with a touch screen? A PDA with a stylus? Does it handle Unicode text rendering? Right-to-left languages?

Widgets that get this stuff wrong feel unpolished. Like, for example, ATITD’s windows, which you can only resize by dragging a pixel-wide line somewhere in the interior, or the super-annoying pop-up notifications, which appear directly under the mouse cursor, preventing you from clicking nearby until you dismiss the pop-up.

It’s a shame the UI is so annoying, I was really looking forward to seeing some more advanced features. For instance, check out the ATITD wiki page about plant genomics!

Game analysis: Galactic Civilizations II

Galactic Civilizations II was developed by a small team at independent studio Stardock. It’s a sci-fi 4X game, essentially “Civilization in space,” or an evolution of the Master of Orion series.


  • The beginning is always peaceful. This is a great feature! One of the most annoying things about Civ games is the tremendously unbalancing effect of early battles, like having your settler taken out by barbarians or conquering an enemy city before their defensive unit moves into place. In GalCiv, you have to get pretty far into the technology tree before you can build units with any offensive capability whatsoever, and the techs that enable you to invade other planets are even deeper. This ensures that all players have some time to expand freely without worrying about military threats.
  • The AI doesn’t cheat (at all but the highest difficulty levels). Stardock put enough effort into the AI to make it competitive without giving it unfair advantages in economy/tech or the ability to peek at your map.
  • Turn-based action is addictive. There is something about quantized jumps of progress that make the action more addictive than a real-time game. This is the same principle behind the prevalence of numerical levels in RPGs (compared to continuous skill improvement). I think it has to do with the “rush” you get from large spikes in activity, which you wouldn’t get from continuous evolution.
  • The first couple of turns are the most exciting. It’s fun to uncover nearby plans and imagine the possibilities for developing them (and how to stop the AI from getting there first!). I think this is because I’m an “explorer”-type player.
  • The technology tree is very well-designed. Aside from the late-starting offensive techs, I also like the rock-paper-scissors aspect to military techs: there are three kinds of weaponry and three kinds of defenses, each specialized against one of the weapon types. The techs are sufficiently resource-intensive that you can usually pursue only one line of development. This mimics real-life arms races where new technologies are developed specifically to counter the opponent. Also, in an expansion pack, Stardock added unique tech trees for each race. This makes game-balancing a lot more difficult, but if you can pull it off (e.g. Starcraft), the variety contributes a lot to replayability,


  • The map view is cluttered. When zoomed in, it’s difficult to see smaller ships and distinguish their nationality. The schematic zoomed-out view shows ships much more clearly, but is cluttered by the large number of trade-route freighters all over the place.
  • They still haven’t solved the micromanagement problem. I always bog down and lose interest in the later stages of 4X games. The ratio of management effort to turn duration grows beyond what I’m willing to tolerate. I make strategic mistakes like not searching the map for sneak attacks while clicking through turns waiting for a tech to complete. The game needs more powerful automation of planet development, and some ways to direct your attention to important spots on the map (like a “proximity alert” when an enemy unit pops up well behind the front lines).