À la mode

Why do so few videogames provide a sufficient diversity of game modes? Given that the choice of game mode sets in place the high level structuring of play – and critically controls the duration of a single play session – shouldn’t game designers make more of an effort to afford players the chance of finding the mode that suits them? 

One of the unfortunate consequence of the ‘games as art’ argument is that it occasionally misleads commercial game developers into believing that they have every right to assert their own desires into the game design they are working on, irrespective of the needs of the audience. But commercial game development is about making games for an audience – the player is the client. Is this not the social contract behind all game development that is neither indie nor free artistic expression? (That is, the social contract behind every game on sale in your local game retailer). Commercial game designers who forget this can be justifiably chastised by their employers, and the people who buy their games.

Why is it necessary to provide diversity of game modes? Because there is a diversity of players. Even if our models of the gaming audience are still somewhat nascent, we can be in little doubt that different play styles exist, and this creates a certain pressure on commercial game designers to account for these differences.

Before proceeding to the centre of this issue, we must briefly take into account a contrary argument. Games that are targeting a mass market audience must necessarily be simple to understand. In this regard, having too many game mode options can present a confusing picture to new players. However, while this is a point to bear in mind, it is trivially avoided. Just as advanced functionality can be hidden in the infrastructure of a design for those players who desire it, so can additional modes by discreetly positioned in the menu space of a game. In the worst case, a top level menu item entitled ‘Additional Modes’ or ‘Advanced Modes’ as appropriate can always group together potentially confusing additional functionality. 

Let us now consider some key ways in which game modes can vary. 

Play Session Duration

The first issue to be addressed is the length of time it takes a player to complete a single outing of a game. For certain games, this is a function of save point location (a cogent argument can be advanced that the player should always be able to save and quit, but this is a side point), but for others the play session length varies according to how the game modes are structured. This is particular relevant when the atomic segments of play are comparatively short, especially those build around mini-games or upon frequent repeating segments of play. 

Let us for the sake of discussion consider a play session to be bounded by choices made from the game’s front end menus. When the player or players sit down to play, they may wish to play for ten minutes, or an hour, or to just play until they want to stop. Is it not desirable for the player to control this element to the maximum degree appropriate for the game’s core concept?

Consider the most recent Micro Machines game (V4). Although I am absolutely delighted that our friends at Supersonic have regained this license, as they are the undisputed masters of this style of frenetic multiplayer racing game, it is a source of some confusion that there are no tournament modes in the new game.

In earlier editions of the franchise, the players could either elect to play single races, lasting perhaps 2-5 minutes a piece (although epic duels of up to 30 minutes are far from unheard of!), or they could choose tournaments that contained approximately half a dozen races, lasting therefore 10-30 minutes on average. This provided a tangible choice of play session duration to players. 

The removal of this mode hurts the game in a tangible manner: it forces the players to choose a new track between every race, breaking them out of the flow of the game space and putting a burden of choice on the players, whether they wish it or not. The inclusion of a tournament mode would not have significantly added to the development costs of the game, but it could have made a substantial difference to the enjoyment of many of its players.

It is not at all clear why this mode is absent. Either Supersonic do not play their game in this manner, and therefore didn’t feel it was necessary for the audience either, or they received poor instructions from their publisher. Either way, it is a warning for any other developer working on a game whose atomic segments of play are only a few minutes long. 

Random Choices

This brings us to the next issue: whether the player chooses the next action, or whether the game allows for a random choice. While it is well established that some players demand absolute control over their game space, it is often overlooked that the reverse is true: many players (in fact, orders of magnitude more players than those requiring complete control) would like the game to take the trivial decisions away from them so they can just play. 

The aforementioned Micro Machines V4 demonstrates this problem: there is no random button to select a random course. Another trivially inexpensive design element, yet it is conspicuously absent. Perhaps this reflect the tendency for game designers to overlook the importance of alea to play. 

We can also see this in Nintendo’s Big Brain Academy. After each round in the multiplayer mode, lasting a few short minutes, the losing player must select the next game. There is no option for a multiplayer game in which the games occur at random, nor for individual players to choose for the next game to be random (which would be functionally equivalent). There is no convincing cost argument for omitting this option, and its absence is both confusing and disappointing.

In multiplayer games, the players are generally most engaged when the play is delivered in a comparatively steady stream. Random choices facilitate this goal at negligible development cost and are always worth considering. 

Difficulty

Perhaps the most staggering absence in terms of game mode is the lack of difficulty settings. Now here it must be acknowledged that multiple difficulties do imply additional costs – in the QA stage of development perhaps more than in the design phases. But we can be no doubt that there are many players for whom the typical difficulty of videogames is considerably too hard for any enjoyment to be gleaned. 

Where are the options that allow players easier versions of games such as Metroid Prime or Shadow of the Colossus? They are wholly absent, and this is typical of the majority of games from the upper market. Even though these particular games are specifically targeting a hardcore demographic at the centre of the gaming audience, why omit an easier mode? The stereotypical hardcore player often considers it failure just to select an easier mode, so the argument that easy modes undercut the challenge of the play seems fallacious.

Furthermore, in terms of the wider gaming audience, there can be no lower limit on difficulty. You simply cannot make a game too easy for the broadest conceivable audience. The player needs to be able to find their own level of challenge, and this may be substantially below where the game developer believes it should lie. Many players towards the casual end of the spectrum struggle simply to control the avatar – for these people, just moving around is an achievement! 

For a very small class of games, specifically sims, a sandbox mode has become somewhat de rigour. Why the option to play in a strictly toyplay manner is so rarely featured in other games is open to debate, but it is most likely a failure to recognise the significantly varying needs of the gaming audience.

The issue of dynamic difficulty adjustment is too large to go into here, but in brief, I do not believe this is a viable approach at this time. Until we take stress measurements from our players by some means (the DS might detect stress in the voice of the user in its microphone for instance!) we are not really in a position to consider taking control of the game difficulty away from the player. 

Neither is this problem constrained to players requiring the game to be easier. There are a smaller number of players requiring harder difficulties, but they are often potential game evangelists, and as such should be taken into consideration.

A cost balance must be made in terms of difficulty settings, but it should always be considered at the very least. 

Having said this, it is worth remembering that the inclusion of different game difficulties creates problems for which we currently have no solutions, as discussed in this earlier post Riddles of Difficulty. The problems discussed in this post remain extant. However, while it remains the case that we have no tenable solutions, we should at least include the option for the player to control the game difficulty by some means - even a poor method such as a difficulty selection choice before the game begins is better than nothing.

Co-operative Modes

For certain games, the absence of co-operative modes is conspicuous. There are a significant number of players for whom co-op play presents a significant draw, especially those who do not enjoy interpersonal conflict (and this could be up to half the potential audience for games!) The cost arguments for co-op play can be even more demanding than with difficulty settings, but the benefits are potentially enormous. It is difficult to estimate to what extent the success of Halo: Combat Evolved was enhanced by its decision to include a complete co-op mode.

It is arguably the case that too many games focus on the multiplayer competitive mode at the expense of a potential co-op mode. While it certainly seems to be the case that the established game audience does enjoy the agon of competitive multiplayer, it seems that a poorly implemented co-op mode can actually be enjoyed more greatly than a poorly implemented competitive mode – which further suggests that co-operative play is a bigger draw than is currently taken into account. 

Another reason why co-op modes might be of importance to the games industry is that they allow new players to get involved in gaming, thus expanding the audience as a whole. I hope to see much more attention paid to co-operative gaming modes in the next few decades, and perhaps we may even see some asymmetric co-operative modes – where the players contribute to overall game progress through entirely different play activities, as well as other co-operative modes such as Katherine Isbister's suggested Mentor Mode.

Other Game Mode Issues 

These few issues are particularly notable aspects of the question of game modes, but there are many more besides. There is the issue of the total game length (which may be varied by mode, as with the Chapter concept for Reluctant Hero), the issue of access to individual segments of play (versus being forced to replay the entire game space), and the issue of choice of avatar gender, for instance.

The bottom line is that variations of game modes can be the most inexpensive addition to a game – but given that each game mode can be practically a completely different game, variety of game modes increases the chance of players finding a way to play the game that they can wholeheartedly enjoy. Puzzle games like Taito’s Bust a Move franchise have taken advantage of this for some time – but other games continue to reject this issue as salient, or add additional modes as rewards for completing games, thus denying access to these modes to the vast majority of players. 

Conclusion

Once upon a time, the social contract between the game designer and the player was something of a laying down of the gauntlet: “I dare you to beat my game!” But this focus on challenge at the exclusion to all other elements of play is no longer appropriate to our modern games market. 

Should we not strive to empower players to find the play they will enjoy? And what better way than to provide, in a carefully considered and thoughtfully constructed manner, a set of meaningful choices as to how the play of a game is delivered?

Note: Edited 4/9/06 to include reference to an earlier post, Riddles of Difficulty, in response to a point raised at Casual Game Design. It should be clear that although I do not think selecting a difficulty mode before play is a particularly good solution to the problems of game difficulty, even a bad solution to a problem is better than no solution at all.

Comments

[Peter Crowther]

Loud applause from this (generally) casual gamer. Difficulty settings and co-op modes, in particular, tend to attract me to games. If the difficulties can be selected per-player this is a further bonus, as it's then possible to have an engaging co-op (or, indeed, competitive) game for everyone. I won't touch most FPSs for this reason, as I come into a competitive environment and die rapidly and repeatedly - an unpleasant experience for me, and not one that makes me anxious to go back to it. Wolfenstein: Enemy Territory is one exception to this, as the variety of roles available and the team play make it possible for someone like me to feel like I've made *some* contribution to the team.

[Patrick]

I think Play With Fire exemplifies the virtues of active difficulty adjustment and active mode selection, i.e. "seelction" of a mode through play rather than discretely through a menu. There's a huge potential in characters to streamline mode adjustment as well. Think about it, people naturally gravitate to characters they like and adjust their behavior likewise. What if that interaction predicated the nature of the gameplay, both in terms of time commitment and difficulty? Co-op would probably have to be a discrete menu thing, maybe, though I bet there's a way to make it fluid.

[Chris]

Peter: upper market developers still tend to make games for themselves and not their audience. Some people actually admire this - I guess in the sense of co-opting the resources of a large corporation to 'do your own thing', it has an "admirable" Robin Hood quality. In commercial game development, I argue that this is an extremely bad practice. But we're a long way from having best practices for game development, for a number of reasons.

Patrick: I agree there is potential in this area - we experimented with a design which involved narrative decisions affecting game play (for example, by giving the player an early choice between a camera and a gun to select between violent and non-violent play) but the game in question never proceeded. The tricky thing is, no matter how cleverly designed, an invisible selection mechanism will always mis-sort a proportion of the audience. The question remains, therefore, of how to detect and correct a mistep.

Best wishes!

[Joao Bispo]

Hello,

difficulty is a tricky subject... personally, I don't like choosing a difficulty setting at the beggining of the game, essentially because I don't know which one would be the best for me! (But this applies to any choices players are forced to do without any information. It feels like a roulette)

One glaring example of this (but I bet this is taken as a marvelous thing by many people) was when I started System Shock and I had a sheer number of parameters from which difficulty could be individually selected.

This break away the "suspension belief" of the game for me, as for one side, one could wonder in the middle of the game "what if I had chosen a higher puzzle level setting?", the worst was the lack of a consistent "game world" idea.

In a sense, this game adopts the "let the player change the world", although, as a player, I prefer the inverse.

[Chris]

Joao: this is a fair complaint you raise here; a game-literate player may well want the ability to tweak the settings within an inch of their life, but even in this situation how does one know what the consequences of those settings might be without having played the game first? And for a more casual player, such choices can be intimidating.

This issue of difficulty settings remains one of the key unsolved game design problems in my opinion. Dynamic difficulty solutions only work for certain games, and even then, they are far from perfect - and the few good ones can be quite expensive to play.

I think the solution may come from one of two approaches.

Firstly, asking the player information about themselves, instead of about the game - as with the mountain climbing/hiking question in Resident Evil (GameCube), but perhaps made less arcane. "Do you want us to challenge you, or entertain you?" or something similar.

Secondly, dynamic difficulty systems that "read" the player's performance and use this to adjust their settings. This side, however, may be incomplete without some biometric feedback system. Perhaps, however, just a speaker would be enough - we could listen for the player cursing. :) Even this will be incomplete: a challenge-seeking player swears blindly, but it doesn't mean they aren't having a good time!

Thanks for sharing your comment! This remains a difficult problem in modern game design.