What I mean by this is the use of ships with special mechanics that are used as a substitute for an interface element that achieves the same functionality. Primary, though quite different, examples of this that the player interacts with are rally posts and the new control nodes.
Though the rally post is in some ways more versatile than the interface request it was originally born from, it was at first a ‘fall-back’ solution, because the issue it was created to address could not be resolved without significant changes to the game engine. Did this make it the inferior choice?
The new control node present in the prerelease is an interesting specimen. It’s a constructible unit that affects a fairly important game mechanic, and yet the unit is not intended to interact with the other units in the game in any manner whatsoever. I don’t think that this is completely counter-intuitive, because the settings menu doesn’t contain any options that affect the game mechanics themselves, so it isn’t as though similar options are being split up in an inconsistent manner. Still, should this option be controlled by a unit, or should it be on a menu?
Looking at both of the above, it’s really a conceptual issue. In the case of the rally post, it’s a question of whether you see the functionality it provides as an order you should really be able to achieve by issuing standard orders yourself, or whether you accept it’s something that requires some special hardware. In my view, I don’t see a rally point as something that should require a unit to set, so the unit feels a bit odd to me. Alternatively, others might see it the other way and so the rally post makes perfect sense.
The control node is a bit more complicated, and the reasons behind that particular design choice aren’t immediately clear to me, but I imagine that it’s one of the reasons discussed below.
So, when it comes to issuing unit orders, the question is how do you distinguish between what should be an interface element, and what should be a unit ability?
Is it the complexity of the order? Is it right that if an order would automate behaviour to too great a degree then it should require a resource expenditure in game?
Is it down to limitations of the game engine? Is it taking a quicker route when the perceived benefit of the feature does not justify the alteration to the game engine?
Is it because a unit gives the feature additional context and makes it conceptually easier to understand and manage?
It’s a combination of the three, but an issue arises when it becomes difficult to discern whether a player should be looking at the shortcut keys or the unit tree when looking for a feature. The answer to me then is that neither approach is inherently wrong, but the approach taken must be consistent.
When considering the correct approach, I think the most important factor is the level on which you see yourself, as the player, interacting with the game, and also the perceived intelligence of the units themselves. To take some contrasting examples, I’ll consider Lemmings and AI War;
Lemmings are ‘dumb’. They do not react to their environment in any manner whatsoever. The important point is that while I can give a Lemming basic orders, it itself has no concept of its environment, and is completely unaware of the context of its task. No one has ever asked if there’s a way of ordering Lemmings to automatically place their own stoppers, because not only is that a behaviour that’s far more complex than anything else Lemmings can do, in no other areas can they interpret their environment as would be necessary to do this.
Units in AI War are not ‘dumb’. They do not simply carry out a task they are given, as Lemmings do, they will decide how to achieve that goal. You don’t tell a unit to fire, you tell it to engage a target - if that target is too far away, it moves closer, if the target is immune, it won’t fire. The ship has a decision making process it goes through when it’s issued a task, and hence is seen as having an intelligence.
Is the ship itself is doing the decision making? Or is the interface is breaking down your orders into smaller chunks based on a few rules, and issuing those to the ship? As that would make the ship itself effectively ‘dumb’. This is an interesting question, but I don’t think that it’s relevant in this context, because either way, the player would expect orders of a general complexity as defined by the existing orders to be possible in either case. In the thematic sense of a spaceship, I interpret it as the former anyway.
When a special auxiliary unit is required to achieve a behaviour that seems comparatively simple to other behaviours a unit can be issued with, and the unit has already demonstrated the ability to interpret the necessary information to carry out the behaviour in other contexts, it strikes me as counter-intuitive.
