Simulated Skill

January 12, 2008 at 9:15 pm (Game Design, Game Mechanics) (, )

One of the indelible properties of games is that people who play them get better at them; players develop skills that allow them to play more consistently and effectively.

A more arbitrary notion is the concept of the player’s avatar becoming more “skillful,” through a convention commonly known as “experience points” (XP) or “levels.”  The idea is that, within the secondary reality of the game, the character (or other entity) controlled by the player is also practicing, learning, and growing in power, presenting the player with new or enhanced options within the game.  While most closely associated with role-playing games (RPGs), this mechanic is now pervading a wide variety of games, from shooters to puzzle games, rewarding the player’s accomplishments with additional powers, and sometimes greater options for customizing his avatar.

While initially this may seem like a nice secondary feature, easily attachable to a wide variety of games, the addition of this mechanic often substantially alters the way a game is played and balanced–not always for the better.

Difficulty Curve 

The first and most obvious effect of making the player more powerful is that she can successfully confront more difficult challenges.  This usually isn’t a problem, since most games expect the player to confront successively greater challenges as the game progresses anyway.

However, the difficulty increase needs to be calibrated to the rate of the player’s increase in power (plus whatever actual difficulty increase you want).  This tends to be more difficult than adjusting an ordinary difficulty progression, because the normal difficulty is self-correcting:  a player for whom the tasks are easy will quickly solve them and progress to harder tasks, while a player for whom the tasks are too difficult will struggle with them and remain at one point in the game until she becomes skilled enough to handle them (I mentioned this in my discussion of why Difficulty is Hard).

The increase in player ability from XP tends not to be so helpful to the designer.  The players who are best at the game–that is, those who already have the most skill–tend to be the best at manipulating the system to get the most XP, too.  If you present multiple options for character advancement, more skilled players are more likely to identify the most effective or efficient ones.  In short, the players for whom the game is already too easy are the ones that tend to advance in “experience” the most quickly.

Additionally, if a player easily overcomes a challenge and moves on quickly, the player gets little practice–but tends to get as much XP (or more) as someone who struggles through it, because XP is offered as a reward for accomplishments.  By overcoming tasks more quickly or more thoroughly, the player with more skill–for whom the game is most likely to be too easy–builds up XP the fastest once again.

You can give the player less XP for winning quickly or easily, but this tends to be frustrating to the skilled players:  they feel they are being punished for playing well.  Additionally, if not implemented carefully, the skilled players will deliberately allow themselves to come close to losing for the extra reward–and, again, they can do this more effectively and reliably than the players with less skill.

One mechanism that is effective is to give players XP even for failed attempts at a level:  this way a player who fails the level several times before finishing it successfully will have more XP (at a given point in the game) than the player who breezes through, but the successful player does not receive less XP on any given attempt.  However, this only affects players who are so far behind the skill curve that they are actually forced to retry parts of the game–an outcome which we would otherwise probably want to make rare to prevent players from being frustrated.

It’s not all bad here; if you provide unlimited side-challenges (or repeats of required challenges, as mentioned above) that award XP, then a player who finds one segment of the game too difficult can always build up XP until they find it manageable, and then proceed.  But the increased instability in the player’s ability to overcome challenges still makes calibrating the difficulty progression difficult.

Snowball Effect

This is closely related to the difficulty curve problem.  If XP is basically given as an award to the player for overcoming challenges, and having more XP makes the player more able to overcome challenges, there is a danger that you’ll get a runaway accumulation–a higher-level player can accumulate XP faster, increasing his level, allowing him to gain XP faster still, and so on.

The traditional approach for combating this is to have an exponential decay in the effect of XP; the more XP you already have, the more you need to accomplish the same relative effect, and so the fact that you are gaining XP faster is canceled out.  Coupled with exponential increase in the XP awared for progressively more difficult challenges, this is fairly effective at keeping the player at the expected level for a given part of the game (based on the rate at which XP can be earned there compared to the XP required to advance).  However, if the player’s XP depends on little other than his progress through the game, one is forced to ask–why not grant the player abilities based directly on his progress and do away with XP entirely?

While this effect can complicate players’ progression through a single-player or cooperative game, it particularly becomes a problem in a competitive setting–if having more XP gives you an advantage in how quickly you can earn XP, then it’s probable the first player to get ahead will keep going and never look back, especially if players can directly hinder each other.

Games where players accumulate XP by fighting each other are particularly bad.  When the player who gets an early advantage is almost certain to win, players must labor under a foregone conclusion for much of the game, which removes much of the excitement and quickly becomes frustrating.  The high-level player has no suitable target for his abilities, and the low-level opponent has no way to compete.

WarCraft 3 minimizes this problem by allowing players to accumulate XP without directly fighting each other (through neutral “creep” fights) and by preventing most player units from acquiring XP at all, but I think it escapes serious difficulties mostly due to the fact that traditional RTS games have such a slippery slope that the game usually ends quickly after one side gains a clear advantage.  Most MMOs put strict level divisions on PvP and grant little or no XP for player kills.  In general, XP doesn’t seem to make a good primary mechanic in competitive games.

Undermining Player Tactics

Perhaps the most sinister problem accidentally introduced by blind application of the XP mechanic is that the goal of acquiring XP often runs counter to regular game goals, thereby forcing players to use unintended and counter-intuitive tactics to obtain XP, or forcing players to sacrifice potential XP in order to use effective tactics.  This is a Power Now, Pay Later trade-off, but one that can easily confuse the player, and these trade-offs tend to result in problems in most of the games that want to use XP.

There are a surprising number of subtle mechanics that penalize players for being efficient.  In some games, you need to fight every enemy (instead of cleverly avoiding them or prudently retreating) in order to earn XP for them.  In others, quickly dispatching a target that is generating opponents, while tactically sound, denies you the XP for the opponents it might have generated.  In still others, you gain XP with every attack, and so an enemy yields more if you kill it inefficiently (with weak attacks and defensive fighting) than if you fight it effectively.

This can theoretically be solved by ensuring that you always reward the player for effective and sensible play, but that’s often harder to define than it sounds.

This problem is particularly pernicious when the player is separately awarded XP in different categories for different achievements.  For example, when the player is given multiple characters with separate XP totals, or several styles or abilities that improve as they are used.

There are several problems that arise here.  The first is that, if you cannot benefit from all of these at once, then you can invariably increase your peak power by focusing on one or a few, and using them only.  This allows the player to increase in power more quickly at the cost of variety and versatility, which makes the game less interesting (fewer available options) and simultaneously easier than intended–hardly a desirable outcome.

This problem perversely reverses itself if you hit an XP ceiling–suddenly your most effective tool can no longer be improved, and so you forfeit all XP when you choose to use it, and must use an inferior option in order to continue accruing XP.  Again, the player is asked to choose between effectiveness and development.

It’s even worse if the player is given any tools designed to be used infrequently, or only in special situations.  If the player actually uses them infrequently, they level slowly, and prove to be weaker than other tools when they are needed, punishing the player for his earlier good judgment.  If they level so quickly as to be of appropriate power despite being used infrequently, then the player, by using them frequently, can quickly reach inappropriate power levels in the other direction.  The player is pushed, in the first event, towards ignoring his special tools in favor of leveling the general ones, and in the other, towards using the specialized tool so often that it becomes outrageously powerful and can be used more widely.  In either case, the specialized tool loses its intended function and gameplay is disrupted.

It may be argued that a game can be designed so that the point of the game is the accumulation of XP, and considerations of short-term effectiveness simply an obstacle to overcome–thereby making it perfectly acceptable if the concerns for XP utterly disrupt what would otherwise be appropriate gameplay.  But if you do this, you have something very unlike a normal game:  you are asking the player to focus entirely on the reward, instead of the gameplay.  This is the sort of design that turns games into an endless, tedious grind towards some dangled carrot and results in people paying other people to play the game for them.  Perhaps that is what some people want, but I would argue you are no longer designing a game, in the traditional sense, at that point; small wonder if the same analysis does not apply.

The Effective Use of XP

By this point, the reader may believe that I dislike the general idea of XP, but nothing could be further from the truth.  I revel in the advancement and customization of my gaming avatars, just as many of you probably do.  There are a great many games that benefit enormously from the inclusion of an experience system, which is why it is such a popular mechanic.  Used carefully, it makes playing the game more rewarding, adds depth and personality, and can even mitigate some problems with games’ difficulty curves, as already discussed.

But, like all mechanics, it should not be used in every game, nor should it be used carelessly or indiscriminately.  As I discussed in Burning Away Impurities, it’s easy to fall into the trap of thinking that you can make a great game by accumulation, simply throwing every good idea you’ve ever had into a pot and stirring vigorously, but it just doesn’t work like that.  Making a great game is as much about cutting out the bad and incompatible bits as it is about putting good ideas into the game; two things that are good individually are not necessarily good together, and two things that are good together are not necessarily good separately.

Game mechanics like experience are like spices.  By using them where appropriate, and in the amount appropriate, you can make great games.  But pile too much on, and you kill the flavor.

Advertisements

2 Comments

  1. Jordan (JJ10DMAN) said,

    Excellent post! The failure to realize the elements in this post is the single reason why the mod “Art of Ascension” (http://aoa.planethalflife.gamespy.com/) is a hopeless failure; it simply gives too much strength to high level players focusing on two or three stats. When I quit, the developer was ferociously defending the shotgun (the only short-range firearm in the game) being able to kill from clear across the map in one shot, given the proper statistics; levels went so far to promote leveling a specific skill and eliminate all player skill as to turn a shooter with a dozen weapons into an instagib mod.

    Call of Duty 4, on the other hand, lets you achieve maximum level in a few dozen hours even if you’re terrible. In Battlefield 2142, the unlocked equipment and weapons are actually more difficult to use effectively (for example having more damage but a smaller magazine, front-loading damage but resulting in less total damage per reload when a player is a bad shot).

    So I know all about the dangers and benefits of levels, and I agree completely.

  2. spellman23 said,

    Very nice.

    I can name a game hat did this wrong. Battlefront 2. They have “Achievements” which gave the player boosted stats (damage, ammo, stamina) if they did something special during one lifetime. If you got it enough times, it would become easier, and eventually you would be awarded the effects immediately on spawn. Big problem, causing a seriously discernible edge for older players. Luckily, it’s also a game where most multiplayer can also be accompanied by bots you can kill, and you can gain the achievements offline and carry them over to online.

    I also find it an odd issue of “sidequest” material in some games. Often times they are purely optional, but also give some reward that makes your life easier, be it money, shiny items, new abilities, or just pure XP. These are more or less there to make the extra effort worth it, but sometimes they detract as well, making later portions much too easy. Finding balance in how much to give to make it “worth it” and yet not make you overpowered is tough. Some games have tried to mitigate this by giving very little gameplay altering things, but instead exploring more of the background of the world or characters, which I personally enjoy.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: