The other day, a friend and I were discussing the introduction of a respec option in patch 1.13 for Diablo 2. It’s a familiar point of contention; my friend’s reaction towards this was decidedly unenthused. As for me, well, this was the only change I requested on Blizzard’s forum post for player suggestions for 1.13. The feeling, of course, is that such flexibility cheapens investment. But mechanistically speaking, it’s really just a trade-off between “the two sides of progression” as Mr. Bycer discussed in his post last week—investment in time versus investment in tactical skill.1 
The New Respecialization Feature in Diablo 2
Similarly, this same friend is not a fan of Guild Wars, though I find the ability to tailor one’s build to meet a shifting challenge to be extremely satisfactory. My understanding is that permitting greater agency in the player’s capacity to alter his strategy allows the designers to achieve more tactical depth as the game design is freed from the need to allow a single unchangeable build to be viable for the whole game (a flattening in which one size needs to fit all). Such a situation existed in Diablo 2 prior to 1.10. The problem was that the difficulty bar for any given area was set by the least effective build for that area, which meant the challenge was significantly reduced for the most effective build for said area. These builds, then, essentially became the only builds you made to do those areas. But Blizzard’s response to this in 1.10 was to increase the overall difficulty and change it so that, for the most part, no single build was universally effective. No longer could an Ice Orb sorc solo Bloody Foothills in an 8 player Hell game, for example. This made sense at the time as Diablo 2 was always a time investment over skill investment type of a game.2 But without respec, the effect was actually an exacerbation of the problem in that it made builds even more singular in their usage by locking their efficacy into very specific (immunity based) areas of the game (read: everyone plays hammerdins and smiters in pve). As well, certain areas in Hell became completely inaccessible to certain builds—a problem magnified by Act 3 mercs being rendered more or less useless. At any rate, I mention all this because the problem is equally relevant to narrational navigation: you can’t respec a narrative progression3, and the answer to solving the problem of having only a single “correct” solution to a ludic challenge does not necessarily lie in restricting player flexibility/agency even further. Moral Predictability as Player AgencyCritics often decry the moral predictability of games, and not without good cause. But moral predictability serves a very specific ludic role: it gives the player agency over an unpredictable narrative. Since narrative progression cannot be respecialized, in order for the player to have (or at least feel he has) agency, either the narrative events or the moral consequences need to be predictable. If agency is to be maintained, the alternative is to make the game narrative redundant or completely linear and fully delegate agency to the ludic aspects of the game. To clarify, it should be noted that moral predictability does not have to preclude narrative complexity and unpredictability, as can be seen in the example of Revan in Knights of the Old Republic. As well, there is considerably more leeway for moral vagueness in narrative events that are not a result of player choices as such details are outside of the player’s agency to begin with. Even further, moral predictability does not have to mean that moral decisions are rendered trivial. For example, most difficult moral choices in Dragon Age are actually straight forward: you know something bad is probably going to happen down the line. What makes them difficult is that you are asked to make a ludic sacrifice if you want to keep the “bad stuff” from happening. So for instance, if you’ve been relying on Morrigan as off heal and dps or crowd control, losing her before the final battle can be a pretty big deal. Thus in the context of invested interaction, a morally predictable choice can still be a complex and involved one. Moral Predictability and Narrative RangeIt is frequently noted that games usually provide only a single “correct” moralistic solution, with the effect being that the designer’s values are imposed upon the player. But there is equal danger for the imposition of the designer’s will when the player is not allowed to make an informed choice.4 This is a sleight of hand akin to the mechanical device of the cheating AI. Either situation can leave the player feeling like he is restricted to rock or scissors in a game of RPS. Moral predictability in player choices is like scouting through the fog of war that is an unpredictable narrative progression. If you eliminate the player’s scouting ability, you are eliminating the player’s capacity to make an informed decision, therefore flattening of the range of unpredictable narrative outcomes which can occur without alienating the player. In other words, all possible results from blind decisions need to be satisfactory, or risk losing the player’s investment. They don’t have to be, but railroading the player with unsatisfactory consequences is a dangerous proposition. The last thing we want is for the player to be held responsible indefinitely for something he had no control over. That is to say, if a designer makes both narrative outcomes and moral consequences opaque, it must be recognized that the designer is asking the player to cede his agency in both axes of player expression. This in itself is fine, but pursuing such a design is a risky venture of balancing the player’s inherent trust that the player’s actions are not entirely in vain against the imperatives of the narrative. The risk can be taken, and sometimes to great effect. But the designer should be careful that the player’s trust is not stretched too far (at least, not during the course of the game), or the player will leave the game.5 And given that the threshold of fairness differs from player to player, the delicacy of this calculation should not be underestimated. Agency and InvestmentOne could say to the above, well, what about a game like Vampire: The Masquerade–Bloodlines? Morality is, for the most part, taken out of the picture there, and yet the narrative can still be unpredictable, right? Not exactly. While the game narrative is quite opaque at times, in all of the narrative events which permanently affect the player’s investment, the player remains fully in control. For example, the true nature of Tourette isn’t exactly obvious, but the real decision making doesn’t happen until the revelation is made. Likewise, all of the possible choices in the finale are quite blatant in their consequences. The challenge for the designer, then, is to give the player all the necessary information without eliminating the difficulty posed by moral decisions or destroying the revelatory nature of storytelling. This is, to be sure, a tall order; separating game world morality from player choice morality, or narrative predictability from moral predictability, is a hard task indeed. In the end, the level of player agency in a game is ultimately a stylistic preference. A cheating AI does not necessarily result in automatic player disenchantment/disenfranchisement, for example. And actually, it is the fight to gain greater agency which creates investment to begin with—agency itself is often (if not always) the reward of gameplay. When there is no more agency to be won, a game ceases to be fun. So limited player agency is a good thing: too much agency and there is nothing more to be gained, which destroys player investment (the question, “what’s the point of trying?” cuts both ways with either too much or too little agency).Finally, game narrative agency and ludic agency are not always one and the same, though such cases usually dictate either that the game narrative exists only as an excuse for the game mechanics, or vice versa.
Footnotes
1 In many ways, it actually is a cheapening of the player’s time investment. This is always a difficulty with games that reward time investment more than skill investment. Once a task is allowed to take less time to accomplish than it used to previously, veteran players can feel cheated (thus some of the vocal complaints about changes in WoW’s leveling time, which in this case is directly linked to actual money). However, the two “paths of progression”, of course, do not occur in isolated realities; there is skill involved in figuring out how to make best use of one’s time, and likewise it takes time to develop skill.
2 Which explains the changes in the current patch to Fire Enchanted monster explosion, WilloWisps damage, and Iron Maiden. These are changes that make Hardcore more palatable as something of a counter to respec.
3 This is not always true. For example, the player can take an Open Palm path for the whole of Jade Empire, but then take the presumably Closed Fist ending at the very final point of the game. However, such a choice was interpreted by some players as flippant and cheapening all of the preceding moral choices in the same way that introducing respecialization affects time investment.
4 On a related note, there’s a certain perception pertaining to moral complexity which I believe needs mentioning. The division of a game world into purely good and evil is obviously simplistic. But perhaps what is not so obvious is the reactionary (that is, opposite but equal) simplification that all actions are inherently selfish and nothing is ever actually good. Both are simplified abstractions, and the designer who chooses to impose the latter should not be mistaken in thinking that such a design is any less an imposition than adherence to the former. I mention this not to discuss the nature of morality (whether the latter is actually a simplification is admittedly debatable), but to note that strictness in either worldview can be experienced as an artificial limitation of the player’s agency.
5 cf. Issue 230 of The Escapist
