One of the biggest issues that I see with games’ inherent designs is that many of them are unintentionally annoying, and it really ruins the value of the game. Sometimes these games are trying for a Dark Souls type approach and just fail to pull off the work to make the consequences part of engaging gameplay, but sometimes there’s just poor design to blame.
You’re running along, collecting all the coins, and you fall in the pit. You get to restart the level, getting a second chance to do it right. This doesn’t sound annoying, and that’s because it, for the most part, is not actually that wrong; the inability to meaningfully fail is a common convention of video games. Fifteen times after falling into that pit, especially if you know the formula to pass it, you will quickly become annoyed as the game punishes you for being an inch off on your execution. It seems pretty obvious, but many game designers include this for a variety of reasons.
Difficulty is frustrating to those players who are barred from winning. That’s the core reason why so many modern games include a difficulty selector, allowing players to customize their experience. However, some games rely upon difficulty to augment their core mechanics, whether to tell a story, create a jump scare, or encourage certain forms of play. Dark Souls is a great example of a game that will kill the player and make them restart over and over, but is generally fair about deaths and as such does not become annoying. However, I recently played Mirror’s Edge, and despite the fact that I beat it in three hours, got stuck several times in certain areas as I frantically tried to execute moves that I knew but the inputs weren’t hitting right. So what’s the difference?
Difficulty becomes frustrating when the player feels ganged up on by the game. In Mirror’s Edge, I was getting fed up because Faith would do a plain jump instead of wallrunning, or would get shot off of a swing-bar while I readjusted her position, plummeting to her death on both occasions. What this really boils down to, on a conceptual level, however, was the fact that I was losing my progress. A couple times throughout the game I’d be falling in the same places over and over because of the first-person platforming being a little gimmicky and not wanting to choose between climbing over a ledge and jumping off to the side of it, and then I had to restart from a checkpoint thirty or so seconds back and go onward. Mirror’s Edge, however, is a game about speed, and having to redo segments (and particularly slow climbing heavy segments, no less) was killing that, since at its core there’s no real challenge to most of the platforming-you see the goal or you don’t and the paths are either really obvious or linear or flexible enough that they’re capable of being guessed, so the fact that the control scheme (mouse and keyboard) was failing to execute my commands quickly became frustrating in ways that, say, Brink never became, because Brink, for all its failings, did what I told it to, and didn’t set me back to the same set piece every time I tripped in front of a baddie with a machine gun or bounced off a ledge I meant to cling to.
One way to avoid this difficulty based frustration is simply to make the game more easy. Had Mirror’s Edge cut out the four or five sections I found particularly annoying, I wouldn’t be using it as the bad example for this piece because I wouldn’t have given it a second thought. Putting checkpoints a little closer to where I was would have also worked, and I probably wouldn’t have noticed the difference. However, this isn’t really the best solution, because a complete lack of risk for failing actually makes the game more annoying; it can create the conception that you are coddling your player. The Burnout series does this well, by making the somewhat lengthy penalties for crashing in the middle of the race a showcase of how your car is getting wrecked, which is somewhat of a moot point when your opponents go from last place to beating you but still graphically impressive and a great awesome factor for a really painful event (it certainly beats driving into a barrier with an indestructible car in Need for Speed).
Another thing you can do is to counteract this is to create a “lives” pool, so that players can only fail a certain number of times, but these tend to actually be annoying or a source of anxiety in and of themselves depending on how they work-skipping the player past an unimportant section of a game they’ve failed in a narrative based game may be okay, but if the gameplay is your core focus it’s going to cause issues as players who are “bad” get less of the game (or otherwise fail to learn proper methods and get skipped through areas nonstop through the harder portions of the game). In arcade games, this was used to tell the player to put in more money, but in modern games it’s generally just used to tell players that they need to get better, which can be insulting.
A second source of frustration is ambiguity (alternatively referred to as artificial difficulty). As I mentioned above, much of my frustration in Mirror’s Edge came from the ambiguity; either the game was unclear as to what I was supposed to be doing and the solution involved going down a level but the markers for where I was supposed to be were all above me. Quick-time events are a major cause of this as well, as players are forced to mash buttons rapidly instead of playing the game normally, and if they’re too brutal they just feel like a “you fail!” moment. Controls are a major cause of this; it’s impossible to get a total virtual reality setup where players control their avatars with their minds right now, but it is possible to make the controls intuitive or clear about what a character will do. One of the things that Mirror’s Edge would have greatly benefited from is a smarter control scheme that assumes that if the player just looked to a ledge they want to jump to it instead of climbing over their current one, for instance, and while the game includes an indicator (in the form of Faith extending an arm) for most of the times the player is going to jump while clinging to a ledge, being a degree off can mean the difference between the two controls.
To avoid causing ambiguity, make sure that your game has easily understood mechanics; I personally am a fan of using as few inputs as possible but leaving everything on a single key-set; this means that any action is available at any time, and is always the same. The military simulator series Arma falls victim to this all the time, with its cumbersome interaction menu that can often interpret player actions incorrectly to assume that they wanted to change a weapon or reload when they wanted to open a door or get in a vehicle. All of this could have been solved with a differentiated keyset; making weapon selection based on hotkeys, differentiating external object use and character gear use buttons and selectors, and so on. Sometimes simplicity is less important than comprehension; if a player has to push a separate button to reload their weapon and use items, they may need to have more controller buttons but they’ll have a much better time when trying to pick up that briefcase in the middle of a heated firefight. Context-sensitivity can be pulled off very well, but it’s as a general rule something that should only be done with controls that cannot have ambiguous interpretations; it’s okay to have a key that uses everything in front of the player in a sometimes ambiguous manner, but if it’ll cause the player to dive on the ground when there’s nothing contextually selected you’ll find the people playing your game having some unintentionally frustrating (or hilarious) experiences. Another side-note is that this frustration can also come from having too few options; if I have to walk around a huge walkway in Dragon Age because a foot-tall obstruction is in the way, I’ll wish that the game had either the option to clamber over such obstacles or had just built in a quicker path-this can be in part a level design issue, but designing a game so that every possible scenario supported by it can be easily completed by players who know what they’re doing is the best bet for making a game that players won’t get angry at.
Finally, one of the causes of frustration is user preference. There’s no real way to fix this one; I hate playing certain turn based strategy games because I have to wait for the other players to finish their turns. There’s no way around this, at least not in most games, though you can do certain things to get past this. Dungeons and Dragons Online, for instance, has a third-person shooter styled mouse-driven control set and a more traditional MMORPG control set; by switching between the two I can play a session without any real angst, relying on the third-person set for when I’m fighting with a ranged weapon or in close quarters, and switching to traditional MMORPG designated targeting for when I want to fling some mojo or heal an ally in the thick of a fray. As a designer, this is something that you can only consider in fringe cases; not every game is going to allow players to customize their experiences and still succeed, typically such games rely either on alternating player roles to allow for some differentiation in controls, but doing so means that certain players may wind up locked in certain means of play; if I want to fly a huge freighter in Freelancer and my buddy wants to escort me in a fighter, we can still make things work, but I’m not going to be breaking off and joining dogfights and perform as well as a fighter would at dogfighting.
So basically, when designing a game you must consider the player’s frustration, which can be caused when the game’s difficulty gets in the way of their desired experience, if errors in the interface cause issues with player’s execution of intended commands, or when the player’s preferences to create a game.