Get the latest Education e-news
 
  • Adaptive Difficulty

    [12.04.09]
    - David McClure
  •  Adaptive difficulty is a term broadly applied to systems that alter game states in direct reaction to a player's perceived level of competence so as to moderate the degree of difficulty that the games challenges pose. The normal intention of adaptive difficulty systems is to ensure that a game remains constantly challenging and interesting without unduly hindering a player's progress, thus keeping the outcome of the game uncertain and the level of tension constant.

    This is apparently the most enjoyable state possible and therefore, according to advocates of adaptive difficulty, the balance that should be aimed for. (Koster, 2005) When adaptive difficulty is discussed there are commonly references to the concept of 'flow', devised by the psychologist Csíkszentmihályi. This theory states that the optimum experience in any activity will be one that is perfectly balanced so as to be neither too difficult nor too easy; further, when an experience involves a higher challenge than the subject has skill for the theory states that they will be frustrated and this will have a much more negative impact than when they encounter a challenge that they out-skill and are bored by. (Csíkszentmihályi, 2002) Juul's research offers some support to this theory in relation to games. (Juul, 2009)

    Falstein proposes that the level of difficulty should vary in a non-uniform manner to create a more interesting and textured game play experience. (Falstein, 2005) This brings to mind an interesting parallel with the structure of multithreaded stories, which could provide a possible source of inspiration for game structure; a sensible expansion of Falstein's concept would be to consider the various types of challenge a game involves and then to plan alterations in their level of a difficulty and frequency of occurrence in a manner similar to the way the choice of focus in multithreaded stories can be described and decided using 2D graphs. (Figures 1, 2 and 3, below)




    While the idea of flow may seem a truism and much of Csíkszentmihályi's work seems to revolve around anecdote, the fact that Juul has proved it has worth in relation to games and that Falstein has come steps from relating it to concepts that apply to other media is a point that must be stressed. Some theorists have argued that play is the primary form of learning and thus the optimum game design and the optimum learning experience will share the same structure, one that encourages flow by increasing the difficulty of a task in relation to a subject's skill, while others have argued that a structural similarity between what makes games enjoyable and learning effective is proof that games originate in learning.(Gee 2003, Koster 2005)

    Although we have seen above that this may well be true with regards to the optimum structure of both activities, the notion that games and learning are intrinsically linked has a few salient flaws: if flow is the optimum point in all activities then it should be expected that similarities between the optimum state in all fields of activity exist and that a crossover is not proof of a link between two activities; if we take it that successful games are fun and that fun activities are the most optimal for learning the fact that research in to the effect games have in terms of learned behaviour offers minor and inconclusive results is incongruous (Byron, 2008); the very notion that play is a form of learning is also very possibly mistaken, a point given in no uncertain terms by Huizinga (Huizinga, 1944).

    If our intention is to make enjoyable games and if, as Huizinga states, play is an irreducible element of being then perhaps we would be better to think about games in relation to fun and play as opposed to constructing what are potentially ennobling fictions that seek to provide an unnecessary base in the earnest. (Huizinga, 1944) Irregardless of whether flow is a universal concept, or indeed one that should be given much weight in general, the results of research do tend towards it being desirable, even necessary, for games to include adaptable difficulty if they are to be as enjoyable as possible. (Juul, 2009) If so, why does adaptable difficulty remain controversial?

    In practice adaptive difficulty is not the universal panacea of game balance that it at first appears to be, perhaps in part due to a lack of available time in the tight schedules that game developers currently work to for the careful adjustments it requires to work properly. Apart from increasing the length of time available to developers, what can be done to improve the use of adaptive difficulty and what pitfalls should designers be aware of?

    First of all, it should be noted that adaptive difficulty is generally a massive problem when applied to multiplayer games that involve competition between players as it can both result in an unjust reward for a player who deliberately plays badly and prevent players from approaching the game in the manner that leads to the most enjoyment. (Saltsman, 2009)

    Beyond making the system used more granular, including an element of chance that makes the system less predictable and decreasing the general power of the adaptive difficulty system to enact changes there is little that can be done to remove this negative effect. In fact, in the case of multiplayer games adaptive difficulty can easily cause a conflict between the necessity of tension within a game and the necessity of following rules within a game. (Huizinga, 1944)

    It also illustrates the reductivism behind the often stated demand that all multiplayer games are balanced by forcing us to consider the question of what precisely a balanced multiplayer game is. Is a balanced game one that can be won by any player or team regardless of the skill with which they play, is it one where a variety of strategies are able to be used with equal efficiency or is it merely any game where each player or team begins with the same resources and in a comparable position?

    Perhaps the best hope for enjoyable and adaptive multiplayer games is to be found in online ranking systems that could either be used to handicap players or decide the composition of teams on basis of ability. In general, adaptive difficulty that strays from such a method is a risky proposition to include in a multiplayer game, albeit one that could extend the life cycle of the typical multiplayer focussed title by allowing it to remain accessible to new players long after its initial release.

Comments

comments powered by Disqus