Slashdot Mirror
Dealing With Fairness and Balance In Video Games
MarkN writes "Video games are subject to a number of balance issues from which traditional games have largely stayed free. It can be hard finding players of comparable skill-level to create even match-ups, diverse gameplay options can quickly become irrelevant if someone finds a broken feature that beats everything else, and some online games make your ability to play competitively a question of how much time and money you've invested in a game, rather than the skill you possess. In this article, I talk about some of the issues relating to fairness and balance in games, in terms of the factors and strategies under the player's control, the game's role in potentially handicapping players, and the role a community of gamers plays in setting standards for how games are to be played. What are your thoughts on managing a 'fair and balanced' gaming experience?"
You are exactly right that tight 'game balance' tends to lead to blander options.
I have seen this in miniature wargaming as well as in computer gaming. But what is really interesting is that even when races aren't perfectly balanced they can still balance well in a competitive environment. In StarCraft the Terrans were seen as being the weakest race, Lim Yo-Hwan then built a reputation as arguably the best SC player while playing as Terrans.
This is what players refer to when they talk of meta-gaming, which is players gaming the game. If I know that Snipers are the best weapon in Halo 3 and that players will go for them and practice with them more than other weapons then it makes sense for me to learn anti-sniper tactics. Very quickly Sniping will become balanced (or even disadvantaged) because you are playing a strat that everyone has trained to beat.
Okay, I'm sorry, but people throw around the word "skill" as though it is some special property that people either have or do not.
In fact, skill is the totality of what you need to be able to do to compete. In Darwinist terms, it is synonymous with fitness. Fitness, as a term, does not specify exactly what makes one fit; this is specific to each individual case, and is derived directly and in all cases from the conditions.
There are several methods by which one develops "skill", and they might or might not apply to a give game.
1. Skill is representative of reflexes. This is rare for most games; it usually is only true when the game is new and no-one really knows anything about it. In these cases, reflexes appear to be the same as skill because they represent an advantage that only one player possesses. Over time, other advantages usually develop to the point where reflexes alone do not represent (much of) an advantage by itself (but is a considerable edge for those who possess other advantages).
2. Skill is representative of time. This is the most common, because almost all humans are capable of both learning and adapting, and so in most cases practice results in elevated mastery. In almost every game, time spend playing is the single biggest advantage that one can have.
3. Skill is representative of money. This mostly used to apply in the days of coin-op arcade games, and is more of a derived skill representation as it enhances the advantage of reflexes by limiting the possibility of time. Those with money could get time to develop their reflexes; those without money had no time to practice and could get no advantage over reflexes alone. Money as a representation of skill is however making a comeback, with paid-for exclusive downloadable content starting to represent real advantages such as better weapons or earlier access to new maps.
4. Skill is representative of knowledge/intelligence. This is in many ways derived from time, but also somewhat independent of it insofar as an individual's natural pattern-matching and information processing and memory potential is concerned. This allows one to discover and or apply complex strategies or unusual rules to a competition, in their own favor. This is in many ways diluted over time, as information becomes disseminated.
Most games represent several values; Fighters and shooters represent skill mostly as Reflexes in the early days, but mostly Time at later stages of it's lifetime (some fighters can also significantly represent skill as knowledge/intelligence where long combo chains and complex moves are particularly important, or detailed knowledge of spawn times and map layouts). MMOs mostly represent skill as Time, but also Money (to varying degrees depending on many factors, and usually mostly in the early days). RPGs usually represent skill as Time and also in many cases significantly represent skill as knowledge/intelligence; on the other hand, Reflexes typically mean nothing at all. And so on. Other classes of advantages representative of skill exists, but are minor and mostly derived from the above.
Sorry to all of the elitists who decry modern games as overflowing with noobs who can't press buttons fast enough/aim their mouse precise enough, but "skill" doesn't just mean that. Almost all advantages come back mostly to time, and all are - in principle - subservient to it. If you say that you're skilled at a game, but don't really mean that you've simply played it a lot, then what you're really saying is that your advantage is likely to be temporary, unless you put time in.
You are right in that it's about whether the fastest rate of gain can be achieved through fighting low or high level opponents, but just because the numbers you picked don't work doesn't mean the system doesn't work.
The idea is that the gain/loss from playing each match is related to the probability of winning. In your example, killing noobs gets you 24 points a minute, while the PvP session gets you 2.4 and thus your example is unbalanced. People will grief as the reward per minute is higher.
In reality the score would be set up so that your expected gains from fighting the low levels is no higher than the expected gains from high levels, and should be less.
Thus say in 10 minutes you could find and kill 10 noobs each with a 99% success rate. If each kill gets you 1 point and each loss loses 50 then you're looking at an expected increase of 0.49 points per noob (0.99*1+0.01*-100), or 4.9 points per 10 minute session.
Now in 10 minutes, maybe you could find and attack 2 people of a more comparable level. Your chance of success with these guys is only 70%, but now you're getting 10 points for a kill, and losing 15 for a death (still losing more than gaining since chance of winning is greater than 50%). You now expect 2.5 points per person (0.7*10+0.3*-15), or 5 points per 10 minutes.
The system is then balanced and you should be neutral between griefing noobs or fighting more challenging opponents. If we then tweak the rewards slightly we can create a level slope whereby the expected gains per minute from noob griefing is less than the gains per minute of doing comparable level fights. You could even tweak the rewards such that the expected return from fighting noobs is negative, even though each win is an increase. If above it was a 100 point loss on losing to a noob, your expeced increase would be -0.01 (0.99*1+0.01*-100).
Now the system means that while you can get points from griefing, you won't get more points than someone who is fighting higher level opponents. Thus if you want to be #1 you need to be gaining points faster than the current #1 and so you won't grief.
I had the chance to play Bobby Fischer at chess once. He kicked my ass around the block.
Then he called me a fucking dirty Jew.
Which was weird, because I'm not Jewish.
Serious or not, that's the best metaphor for gaming online I've every heard.