In a recent post, a person on these forums complained that he did not like CivIV because it was “simplistic” – to which Brad replied that this person would probably dislike Elemental as well.
I am going to try to argue that simplicity or complexity are not the (only key primary) elements which determine a TBS’s strategic depth and fun. On the contrary.
POSTULATION: Axiom A
I would like to postulate an axiom on TBS fun, let us call it Axiom A. I will claim that one (primary, key) element that no TBS may lack without becoming boring is this: A (consistent) stream of strategically relevant and ambivalent (or better: polyvalent) choices. By ambivalent I mean a choice about whose outcomes there should be room for discussion; a choice which allows for various normative interpretations. The diametric opposite of an ambivalent choice would be a no-brainer. By consistent flow I mean that there should not be many turns in which the player need not rethink a strategy.
By way of demonstration, let me return to Civilization IV mentioned above. I did not find CivIV simplistic; I found it unexciting after the first three playthroughs because I did not feel strategically challenged once midgame started. I did not have a plethora of strategically relevant choices; I was “merely” optimizing strategic advantages by means of tactics, always coming closer to my goal. I was rarely surprised. There were many turns during which I found nothing interesting or challenging to do. In addition to the fact that many cities seemed impersonal (because in any city you may only have one type of improvement, e.g. only one temple, forge, etc.) and hence interchangeable, I found most turns to be interchangeable as well.
This is one of the reasons I have de-installed GalCiv2. I love the idea of the game, but disliked its slow plodding. While I did not have to micromanage, and was pleased by this fact, I found myself nevertheless hitting “return” for turn after turn; this was neither challenging nor fun.
Consider chess for a moment. Chess has proven itself over the centuries as a benchmark of classic TBS success. (Please note that I will not be arguing that Elemental should resemble chess in many ways.) One of the reasons chess is considered a “good” game, in my opinion, is that it fulfils what I called Axiom A above: There is almost no turn in a well-balanced chess match (that is, when two players of halfway similar skills are paired) by which a player is not challenged to think about the repercussions of moves many turns in the future. Players are required to think about moves and their counters. This is because each move in chess is polyvalent. Some moves, are of course, no-brainers in the sense that these would be “stupid” moves; these moves are not polyvalent. The fact that chess is a good game is also manifested by the fact that some moves can be called “good moves” – a novel, surprising, challenging move.
Yet chess has simple rules. Chess is not bogged down in minutae, yet still challenging.
The fact that chess has simple rules is also a huge factor in making it relatively easy to program AI for it. Simple chess programs can beat the vast majority of chess players, and top-of-the-line chess programs cannot be beat by anyone but the very best human players in the world.
But it is simple to learn (and cheap to make or create a chess board and pieces). It is simple to learn, but very very difficult to master.
RULE R: EASY TO LEARN, HARD TO MASTER
Of course, chess is not for everyone. Not for me, for example; I play about only 10 games a year. I prefer more complicated computer TBS’s, like what I hope Elemental will become. But I think we can learn a lesson from chess, and from other very exciting computer TBS games which have borrowed something of chess’s rule of success: Easy to learn, hard to master. (Another game that fits this category is a niche product called Star Chamber, whose bafflingly simple rules have created an almost unthinkingly complex ground for consistent strategic challenge: there is no turn in SC by which one really need not (re-)think strategically. Star Chamber is, however, hampered by two huge shortcomings: The first, it is a TCG (trading card game), an unsavoury genre; And second, it has no single-player mode, it is multiplayer only. Additionally, it is not epic; it can only uphold its strategic tension for 30 turns.)
I think CivIV failed here; CivIV is not terribly easy to learn; but once you have learned it, it appears to me that mastering it is not far off. It does not offer a consistent strategic challenge. I also felt that GalCiv2 is in the same boat.
For this reason, I appeal to us all to forsake the flashy appearance of “more is better”. Having “more” might not necessarily increase our options for a steady flow of true polyvalent strategic choices. Remember GalCiv2: There were limitless amounts of ships to be made, but in reality, we all ended up making slightly different flavors of the same thing. That was indeed fun for the first 5 playthroughs, but it lacked depth despite complexity.
So what does that mean?
More counters. I am wary that the path that unit construction and combat might go is not undercomplex, but rather without counters. If the main differences between units will be primarily arithmetic or superficial – that is, one unit can be “better” than another unit because it can do more damage, have better defense, or be faster – then units will be easily quantifiably better, but not different. This will be the exact opposite of chess. CivIV, as do most other TBS games, worked around this problem by having built-in counters coming from a system of “classes” and “types”, e.g. a “pikeman” was both a melee unit (class) and a spear unit (anti-horse unit type); spears are more efficient (are counters) to horse; maces (or whatever) are counters to spears; archers (or whatever) are counters to maces. Other TBSs follow suit. Chess is very, very similar in this respect: Each type of unit is as different from the other as it can get given the simplistic ruleset.
I see the danger of Elemental’s “open” unit-building system and clear-cut combat system (from what has been hinted at so far) to run counter to this. If – simplified – units can be better, but not different, then our strategic choices will be vastly limited, because we will be, in essence, merely optimizing our arithmetic advantages. Our strategic military choices will be limited to finding ways of getting “more”. There will be fewer military surprises, and fewer options to make a “good move” militarily.
Hence: Employ types. If not types of damage (e.g. piercing, slashing, bludgeoning) then types of attacks (which I have called “maneuvers” elsewhere) or classes of units (e.g. flag a unit as “is-a-shield-using-unit” with rules for shields or “is-a-riposte-using-melee-unit” with rules for this particular offensive or defensive maneuver). Employ type- or class-like differences to enable strategic counters. Various weapon, armor, or even unit types should be "good" in situations, not (primarily) "good overall".
I am suggesting this now because Brad, in his recent post on “Spells and Magic”, claimed that certain things, if decided early enough, are easier to change before vast quantities of resources have been devoted to it.
It doesn’t have to be complex, and certainly not complex for complexity’s sake. But it should increase the possibility of a steady and continuous stream (over and during as many turns as possible, well into the end game) of strategically polyvalent choices.
Your input and feedback are welcome. Thank you in advance.