informa
Features

Learning to Love Handicaps in Competitive Games

Rubberbanding and injected randomness can level the playing field in competitive games -- letting players of various skill levels play together. But they also can create other problems, potentially harming your game's design. So why not take a look at handicaps?

I'm a huge fan of the N64 kart-racer Diddy Kong Racing. Actually, the N64 had a lot of great racing games. Many people are most familiar with Mario Kart 64, but I much preferred F-Zero X, WaveRace, and even the rarely praised ExciteBike 64.

All of these games shared a similar story for me, and I'll bet that I'm not entirely alone in my experience. Excited to get into this new game, I'd start playing it the day I picked it up. I'd play it a lot -- as all of these were all very tight and challenging games. Somewhere between a few hours and a few weeks after picking it up, depending on lifestyle, I'd get some friends over and play it multiplayer.

There are certainly those social groups that form around these games, playing routinely every night, or once per week, or some such pattern. For those groups, these titles likely lived long, healthy lives -- a long, head-to-head story of competition between rivals.

For most of us, though, the pattern tends to not come together that way. Often, it looks a bit more like this: you'd bring home the game, excitedly play it for awhile, eventually get a friend over to play...

...and you're squarely better than them, to the point where this otherwise great game effectively sucks now. You're so far ahead that you never even interact. It's not fun for either of you, and you'll very likely not play again.

I refer to this as a problem of "skill deficits" -- a difference in two competitors' levels of ability. Very few games can survive this, if any, and it's arguable that those who do aren't worth playing. It's also, in a practical world, somewhat unavoidable. People are going to achieve different levels of skills in games. It's even arguable that the larger the possible ranges of skill-levels possible, the more credit the game deserves. This is what we generally refer to as "depth" in gameplay.

So, while skill deficits are not avoidable, they are manageable. The question is, how are we managing them? Are developers really aware of this problem, and if so, why are some developers making design calls that exacerbate it? Worst of all -- could we be making destructive design calls that damage our games, winning a pyrrhic victory on the war against skill deficits?

Analyzing the Problem

Skill deficits are a problem, but they aren't necessarily a problem of any particular game. In other words, I don't hold the skill deficit problem against F-Zero. In fact, it's important that players of vastly greater skill are able to win by large margins. This expresses the dynamism and depth of the system which most of us agree are important qualities for games to have. Games, it's largely agreed, are systems heavily dependent on learning. If learning is possible, then it naturally follows that bad players turn into good, leaving those who have played less behind.

So skill deficits are a problem that we can't necessarily blame the game itself for. I'll get into who we can blame later, but for now, it's worth having a solid understanding of why skill deficits are actually a problem.

The situation I described above, wherein players end up not playing a great game due to skill deficits is not the problem, but rather a symptom of the problem. The real problem is that the contest element of games -- the "let's see who is going to win" aspect -- is not present.

This contest element is crucial, because it presents the players with a non-pre-determined outcome in which their choices (input) have influence. Surely we can all agree that player input should matter in a game, but if the outcome is predetermined -- as is essentially the case in a vastly outmatched contest -- then your input essentially does not matter.

On paper, that seems like it should be problematic. In reality, it's also known to be problematic. Who wants to play against someone that they truly know they won't win against? For that matter, who really wants to play against someone that they truly know they won't lose against? 


In Different Types of Games

As we'll see in greater depth, different games have different ways of dealing with skill deficits. However, I want to quickly address how some types of systems naturally -- that is to say, without any anti-skill-deficit-problem actions taken on the part of the designer -- exacerbate or diminish the skill deficit problems. First, some that exacerbate the problem:

Racing games. Racing games, by their nature, tend to exacerbate the skill deficit problem. This is due to the naturally low amount of interaction, combined with large in-game space (usually a "track"). This means that players could be on completely different points of a large track, not even seeing each other for much of a match.

Some fighting games. Fighting games with a lot of hit-stun or long memorizable combo attacks tend to be extremely unforgiving for a player of lesser skill. Sometimes, a game like this can completely take agency away from the other player, to the point where it would make no difference if their controller wasn't even plugged in.

And now a couple of examples of systems that naturally diminish the skill-deficit problem.

Most European board games. Many Eurogames tend to not have direct-conflict or player-elimination. A number of popular ones have even been derisively referred to as "multiplayer solitaire," because of their "you do your thing, I'll do mine, and at the end we'll compare scores" play patterns. Further, scores are often not fully calculated until the end. This means that even if a player ends up destroying another player score-wise, both players can have a good time exploring the system. In fact, I've heard players go "Aww!" at the end of a Dominion game despite having won, because they had some crazy combo they were excited about executing.

Games with naturally high levels of randomness. Card games like Poker have a considerable amount of randomness to them, which tends to bring all players' win percentages closer to 50. This means that even a terrible player has a chance to win every now and again -- if even by absolute fluke. Technically, this partially alleviates the problem.

The Modern Response

The two most common modern responses to the problem of skill deficits are:

1. Randomness. Either in the form of natural randomness as in the case of card games like Poker or Dominion, or in the form of injected randomness as in the case of Mario Kart. 

2. Rubberbanding. A system of rewarding poor play and punishing good play, creating an effect of bringing all players to the center. 

First, randomness. Above, I used the phrase "naturally high levels of randomness" to distinguish something like Poker, which has randomness fundamentally baked into the gameplay, and something like Mario Kart, which has "injected" randomness to an existing system -- racing -- that could otherwise exist as a perfectly healthy game without it -- save for the problem, of course, of skill deficits.

Indeed, an increased level of randomness is the primary go-to measure for modern game designers in addressing this problem. Just as a naturally random game mitigates the effect that skill deficits have in destroying a contest, you can get a similar effect by adding randomness into an existing design. Team Fortress 2, while having the normal "bullet-spread" randomness found in nearly all entries in the genre, added a new random critical-hit feature. According to Robin Walker, this wasn't added to level the playing field, but it's undeniable that it has that effect, to some degree -- whether they intended it to or not. 


Natural randomness is found in great quantities in many popular boardgames. Arguably the three most popular Eurogames: Carcassonne, Ticket to Ride, and Settlers of Catan all have a large amount of randomness to their play.

These games often get called "family games" partially for this reason. The level of randomness gets even more extreme with games like Munchkin or Saboteur, which are usually classified as "party games."

The second approach is rubberbanding. Mario Kart is probably the best example of this I've ever seen, in that it exhibits some of the most dramatic rubberbanding effects. First of all, as most people have experienced, the items in Mario Kart are not completely random.

If you're in first place, you tend to get only the more benign items such as banana peels (which drop behind you, creating a road hazard) or mushrooms (giving you a speed boost). If you're behind, however, you're treated to a fantastic arsenal of easy to use, powerful items.

One such item, the infamous Blue Shell, travels the entire course hunting after the first player, hitting many players on the way. Even better (or worse?) yet is the more recent addition of the Bullet Bill item, which gives your vehicle a tremendous, prolonged speed boost and the ability to wipe out anyone you touch on the way. One could
be forgiven for feeling that when drawing a Bullet Bill item, they've just effectively been handed first place. Therefore, in this way, Mario Kart's items work to combat skill deficits in both ways: randomness and rubberbanding.

The rubberbanding in racing games is not limited to items, however. Many popular racers -- the Ridge Racer and Dirt series, for example, have rubberbanding baked into the actual racing mechanisms themselves. So, if you're in the lead, your max speed might be lowered. If you're behind, it might be increased. Some of them have even gone as far as to give players who are behind a speed boost in the area right near the finish line, to "synthesize" those "caught up at the last second" moments.

These days, the discussion of skill deficits can mostly be heard in the form of talk about accessibility. The arguments for something like randomness or rubberbanding is generally framed as a measure to allow a larger amount of people to enjoy a particular game. And it may indeed be the case that a game that has a sizable amount of randomness or rubberbanding helps accomplish that task -- but at what cost? 

Costs of the Modern Approach

It can't be argued that a naturally random game has taken any kind of "damage" from its method of dealing with skill deficits, since its method is something fundamental to its character anyway. In the case of games with injected randomness, however, there is a serious cost.

Taking a look at Mario Kart, we have a generally coherent and deterministic gameplay loop as the basis: accelerating forward, turning, avoiding obstacles, trying to maintain your speed. Players engage with this, making decisions about when to start turning for this curve, which fork in the road to take, whether to dodge that obstacle on its left or its right, etc, and get consistent, reliable, deterministic feedback for it.

Then, there's the items game on top of this, which not only is random, but is extremely powerful. In fact, rarely do you get similar positive feedback from playing well as you do from drawing random items well. So this can have a very frustrating, confusing effect for players. It diminishes the feeling of accomplishment from a win, and diminishes the feeling of having learned anything from a match.

Further, while many players might shield their ego by saying something along the lines of, "I only lost because he drew a Bullet Bill right at the end there," it also signifies a disconnect between that player and the coherent, natural system that he is trying to understand. Again, a game is essentially a conversation between a player and the system. Losses in particular are a place where the game speaks volumes to the player about what he can learn, but in the "lost due to randomness" situation, the conversation is shut down. So, while you may get a short-term benefit from this effect, the long-term interest in the game is likely to be diminished. 


Worse yet, in their attempts to use random items to create a rubberbanding effect, they have actually clouded what constitutes "good play." In fact, it may be the case that staying out of first place is optimal, until the very end. Perhaps staying in second place is best, to avoid the terrible draws of items? Creating this kind of confusion about something so fundamental to the game is almost certainly an unwanted side-effect of its items system.

One defense of the modern random-items approach would be something like, "well, if you don't like the effect that items have on a game, you can just turn them off!" This is true in some cases. In Super Smash Bros., tournament play almost always has items set to off, and the game works great. In fact, you might even say that the game is a bit strange with the items on, for the same reason that the items are strange in Mario Kart: they are great exceptions to the flow of the game.

However, there's at least one example I can cite of a situation where, even if you could turn items off, the damage was already done. Super Mario Strikers, a soccer game for the Gamecube, is a generally solid, deterministic game based loosely on the rules of real life soccer. It has random items, but they can be turned off, which I was pretty quick to do. Now, one huge difference between Strikers and real soccer is that you can freely shoulder-ram other players to take them out of the game for a few seconds.

It's an extremely powerful move that's 100 percent effective and you don't even have to aim the attack; it automatically targets whoever's closest to you when you push the button. It sounds unfair, but there's a counter-balance: if you're shoulder-rammed, you draw a random item. So, by turning the random items off, I had actually unwittingly turned the shoulder ram into a clear, optimal strategy. What ensued was some of the most violent (and broken) games of soccer ever witnessed.

The lesson is that it can actually be difficult to have items be well incorporated into a design, yet also be an option that can be turned on or off. Either the game needs them, or it doesn't. If something doesn't need to be there, it's pretty easy to argue that it shouldn't be there. And generally, things that are "injected" into a system don't need to be there.

I should mention that whenever you point out a problem of randomness or any sort of unfairness in a game system, defenders will often make the following counter-argument: "What are you talking about? This game totally has skill -- the better player wins most of the time." This is a straw man argument. It is of course true that Mario Kart, even on its most random/rubberbandy day, still involves a lot of skill and the better player will win perhaps even most of the time. However, that doesn't refute the arguments being made here: that the win percentage is being pulled towards 50 percent due to randomness and rubberbanding.

Ignorance of the Problem?

Strangely, there exist examples of competitive games that actually create a positive feedback loop for those who exhibit better play. Online FPS games such as Call of Duty have a pattern of heaping rewards on the players who play the best, increasing the already-existing skill deficits. Why do they do this? Why on Earth would they want to make the problem of skill deficits even worse?

The fact could be that they aren't thinking about the problem of skill deficits in this way. Instead, it's likely that they're thinking along the lines of a single-player RPG or other non-competitive, completion-based software. In these types of apps - say, something like Fallout 3 or Final Fantasy VII -- it makes a sense to allow the player's power to exponentially grow in a positive feedback loop. These are systems built around completion, after all, not competition.

However, taking that same mindset to the world of competitive games can be very destructive. While the Call of Duty example is pretty extreme, more subtle examples can be found all over the world of competitive games. Deathmatch in the original Quake, for instance, tended to revolve around hoarding the good armor items and the rocket launcher weapon, both of which existed in a fixed spot on a given level. When one player would kill the other player, he would then establish a position right near these items to prevent the other player from getting them, thereby exacerbating any skill deficit that existed in the first place.

We need to be consciously aware of this kind of power-snowball at work in our games and make sure that we're not actively working to make the problem worse, before we can figure out how to make it better. 


Learning to Love the Handicap

Game-players these days tend to have a knee-jerk groaning instinct towards the idea of handicaps. To suggest that a match should perhaps be played with a handicap is considered an insult, even if it's well known that one player is vastly outmatched. While it's certainly not the case that handicaps are the only way that skill deficits can be tackled, for competitive games, we are doing ourselves a gargantuan disservice by not embracing them.

Handicaps have been a part of games throughout their history. From Sumo to Squash, from Go to Chess, most of the world's most famous and beloved games have had a long history involving some form of handicap.

Obviously, the handicap has a downside, which we're all intuitively aware of: they aren't "fair." The sense of victory or defeat associated with winning or losing, respectively, is diminished. You didn't "really" win if you won with a handicap, and your opponent didn't "really" lose.

I argue that while this is a legitimate sentiment, it's in our interests to move beyond this way of thinking. It helps to remember that when you enter a match of a game, it is based on "agreement on a set of rules." People house-rule and alter those sets of rules all the time, and handicaps can be thought of as a house-ruling. If we both agreed to a game of Go where I start with three stones down, then that's what we've agreed to, and the outcome of that match is completely legitimate.

It also helps to remember that in many cases, the game would simply be boring -- arguably broken -- without the existence of handicaps. If you are playing against someone who's vastly better than you, you simply are not going to win (a possible exception if the game is heavily random. If that's the case, though, are you really earning that victory anyway?) If you are not going to win, then the contest is not functioning and your decisions (input) do not matter.

Essentially, the argument against handicaps can be phrased as "handicaps cause games to do a poor job of reflecting who is better at the true, unmodified game". Sure -- granted! But I'm not arguing that you should never play with handicaps off. You've always got the option to play a vanilla game if accurate true-game measurement becomes more important than playing a good match.

Handicaps as Metagame

In most games that have handicaps, you'll play a number of un-handicapped "placement matches." This is how the system determines what kind of a handicap you will get. What's interesting about this is that your handicap level acts as a sort of RPG-like metagame ladder for you to climb -- often, "rank." Rank provides a very clear illustration of your improvement in games like Go.

Richard Garfield has a nice article in defense of handicaps, wherein he writes about this effect:

"It became an appealing metagame to improve my handicap, and gave me a much more visceral appreciation for my improvement than seeing my win percentage go from 15 percent to 25 percent, which I might not even notice."

This aspect of handicaps is something that digital gamers should love. It's a meaningful metagame that doesn't distract from the core gameplay experience, as some metagame has been known to do; instead, it pushes you further and further on your path of improvement.

Super Smash Bros. for the Nintendo 64 has an auto-handicap feature that I quite like, and often use. Players all start at "5" handicap, which is balanced. As you win matches, that might start moving downward, which means that you're knocked back further from attacks, and your attacks are less powerful.

I have a reputation for being vastly better at the game than any of my friends, to the point where they all refused to play it with me. With auto-handicap on, however, that has changed. Over the course of numerous matches, my friends will compete with each other to see which of them can get the lowest handicap -- and best of all: Instead of winning 100 percent of the matches by a massive landslide, I now win somewhere around 50-60 percent of them. Matches are exciting again! 


Unforeseen Consequences

Not all handicaps are equal. Some handicaps fundamentally change the nature of the game, often in ways that make the game less interesting. While Go is a great example of a game that can be handicapped wonderfully by adding a number of starting stones for one player, Chess is more difficult to handicap. Removing a piece or adding a capture restriction in the tight game-space of Chess can dramatically change or even break the game.

Since digital gamers tend to not care for handicaps so much, digital game developers tend to not think too deeply about their handicap systems, if they even include them at all. For example, while my Smash Bros. experience was vastly improved with the handicap on, there were some unforeseen consequences of the handicap.

As most are aware, Smash Bros. involves knocking opponents around, building up their damage percentage, until they're sufficiently damaged and can be knocked out of the stage. However, if my handicap is low (meaning my attacks' knockback power is reduced), it actually makes me able to string together God-tier juggling combos wherein I hit the other player some 20 times before they're able to do anything at all. I generally mitigate this by simply vowing not to take advantage of this new power, but it's clear that Smash Bros. with handicaps is nowhere near tournament-ready.

Embracing Handicaps

If we're going to embrace handicaps, which I think we should, we're going to have to take the endeavor seriously. Coming up with handicaps is a serious game design job, so it's best if you design official handicaps rather than forcing your community to come up with some. It's often not enough to tweak an HP or damage knob and call it a day.

There are a few questions you should ask before you go forward with creating your handicapping system:

  • Does your game need handicaps? Again, games with high levels of randomness don't need handicaps. Light family/Eurogames with a lot of toy-like building and not a lot of focus on direct interaction can work fine without any handicaps. Racing games, fighting games, or other kinds of serious strategy games, on the other hand, should probably have some kind of handicapping system
  • How "potent" should your game's handicaps be? It's generally best to have handicaps exist in an understated way. It's often better to take a "one player gets a head start" approach, rather than a "permanent difference in power" approach - it's much easier to balance that way and doesn't feel as unfair 
  • If your game has an online ranking system, consider incorporating your handicaps system into the ranking system itself. For a great example of this, I recommend looking up KGS Go, a great online portal for playing Go online. It uses the classic Go ranking system, which you can read about here.

In short, our avoidance of handicaps is silly, and we're allowing our competitive games to be damaged unnecessarily because of this silliness. Instead of using handicaps, we're injecting randomness, or worse, modifying the rules of our games to make them fundamentally "looser" so that skill matters less. Currently, developers feel that they have to make a choice between being more accessible (more friendly to players of low skill), or being more competitive (allowing for vast skill deficits). Handicaps allow a single game to accomplish both.

Skill deficits are a real problem, but if they're a problem, handicaps are the only sensible way to deal with them. The negatives to using them are vastly outweighed by the positives, and I think if everyone makes an effort to give them a chance, we can start moving towards a more handicap-friendly -- and therefore, a more competitive-game friendly -- world.

Latest Jobs

Infinity Ward

Woodland Hills, California
11.3.21
Sr. Multiplayer Design Scripter/Programmer

Disbelief

Cambridge, Massachusetts
11.3.21
Jr. Programmer

XSEED

Torrance, California
11.3.21
Head of Marketing
More Jobs   

CONNECT WITH US

Register for a
Subscribe to
Follow us

Game Developer Account

Game Developer Newsletter

@gamedevdotcom

Register for a

Game Developer Account

Gain full access to resources (events, white paper, webinars, reports, etc)
Single sign-on to all Informa products

Register
Subscribe to

Game Developer Newsletter

Get daily Game Developer top stories every morning straight into your inbox

Subscribe
Follow us

@gamedevdotcom

Follow us @gamedevdotcom to stay up-to-date with the latest news & insider information about events & more