|
Long, long ago, there existed large, monolithic machines. They were larger and heavier than we humans, and they could crush us easily - and did so. In fact, bizarrely enough, we humans would pay these machines to crush us: we would insert coin after coin into the machine, and - in a dramatic proof of humanity's masochism - the more the machine punished us, the more we would feed it coins.
These machines were called "arcade games", and one day a way was devised to bring them into your home for play on your own TV. This began a long series of events that led to such games becoming played by larger and larger audiences of players. Developers of these games realized that maybe humanity wasn't completely masochistic after all... in fact, it seemed like the less punishing a game was, the more people would play it. Within a mere couple of decades, this proud lineage of monolithic, human-crushing machines were reduced to sleek pieces of white plastic that any grandmother could pick up, waggle around, and actually have some fun with for a while.
 
"Casual Gameplay", aka "Hardcore Gameplay"
And so the question was asked: why have losing at all? Though games continued to often follow a model of presenting users with challenges and giving them a setback when they failed to conquer a challenge, those setbacks became less and less. The excellent modern indie game "VVVVVV" by Terry Cavanagh is described by many as a game of hardcore difficulty... and indeed you are guaranteed to die a great deal as you face the challenges in the game. Yet in a certain way it's incredibly non-hardcore, because the game is liberally littered with save points (sometimes multiple on one screen) from which you'll respawn instantly when you die, with no other penalty. The penalty becomes "redo the last 10 seconds or so" - a far cry from the "3 lives, and each time you lose one you start the whole level over, and when you lose them you must start the entire game over" of Super Mario Bros (much less the "...oh, and also, pay me another quarter!" of Donkey Kong).
Another interesting example was "Prince of Persia 2008" (a name I recommend we all adopt when referring to this game, to reduce confusion about exactly which of the many PoP games we're actually talking about), which was a fairly controversial game for having what was perceived as incredibly low difficulty, mostly because the player character could never die: when the player failed, the Prince was always seen to be saved from death at the last moment. Of course, this did still set you back to the most recent checkpoint you'd touched a few seconds before. Again, this game was labeled "too casual" by most gamers... even though the gameplay I just described is mechanically the same as VVVVVV, the only difference being that you don't watch your character die before the reset occurs. Apparently the pivotal question deciding whether your game is labelled "super casual" or "super hardcore" is, "do I get to watch a character die?" Which makes me think that maybe what hardcore gamers really want are snuff films.
Know What You're Sacrificing
I'm actually okay with the VVVVVV, "minimal setbacks" model of gameplay. I agree with Raph Koster's "Theory of Fun" - that at the core of fun is the experience of learning and mastering a skill - and I love games that present me with a challenge that I must face over and over until I conquer it (learning with each iteration, of course). In general I feel that the tighter you can make that "loop" - of trying, failing, and trying again - the better.
But I do think we should be aware of what we're sacrificing when we choose to make setbacks minimal. Let's go back to Donkey Kong:
- As you progress through a screen of Donkey Kong, tension continually increases: because the further you've gone from the checkpoint (the beginning of the level), the more you fear dying and losing all your progress. In other words, the farther you travel from your checkpoint, the greater you feel the risk of losing is, because you'll lose so much progress. On the first screen of DK, standing on the top platform a few feet away from the great ape and trying to judge the perfect moment to scamper up the ladder to victory, the player's tension and excitement is palpable. If the game created a checkpoint on each platform so that when you died you were restarting on the start of the same platform (rather than starting the entire screen over), you would lose this excitement and feeling of having a lot to lose.
- The same thing is going on, on a higher level: The more screens I proceed into Donkey Kong, the more tense I become, knowing that I'll lose all that progress once I lose all my lives. (And of course if I'm Billy Mitchell or Steve Wiebe, then I'm very invested indeed in avoiding death, as I don't want to see my score reset.)
In more recent examples, several recent iPhone games that I love - Canabalt, Flight Control, and Doodle Jump - have very arcade-like structures in that they're based around generating a high score, and ramping up the challenge as the game proceeds. Again, the feeling of "I've come so far and have so much to lose" comes very much into the foreground when you've gotten further in these games than you ever have before.
Here are a few "high-setback" (and therefore high-stakes) games -
some of which have a major cult following, partly because of the tension
- rogue and NetHack, the
classic ASCII games which a player can delve into for many hours, and
which allow saving your game... but which automatically delete your
saved game file as soon as you load it. The result being that whenever
you die in the game, there is no way to restore any save, and you must
start over entirely.
- Kevin Yu's Spelunky,
a modern indie game that derives much of its structure from the
"rogue-like" formula, including a lack of savepoints.
- You Only Live Once, a small experimental Flash game
that can literally only be played once!
Flow vs. Self-Consciousness
This tension I've described is actually pretty obvious and noted by many people. One thing that's a bit more subtle, however, is something that emerges from this tension: a form of "performance anxiety", and particularly what I'll describe as "the battle of self-consciousness." It's common to any activity that involves getting into a state of "Flow" (or "in the zone" as some would say): the problem is that almost as soon as you enter this state, and especially as you get closer to achieving a difficult-to-achieve goal, you recognize that you have entered this state; and this recognition itself tends to create self-consciousness that, ironically, pulls you out of that state.
In other words, as the stakes increase, and as the player gets closer to achieving a difficult goal, the more tense they'll become... and therefore the more likely they'll be to become self-conscious and thereby possibly ruin their progress towards the very goal they're approaching. These are inevitable results of humans engaging in an activity that requires high skill, with high difficulty, and high stakes... and while it's not for everyone, this is sort of experience is exactly why some people play games at all.
Psychologists sometimes call this "Analysis Paralysis"; and in sports the phenomenon of "choking" is thought to be a result of this. Of course all this sounds super-hardcore, and it probably is - your grandmother is probably not looking for this experience when she picks up a Wii controller with her grandkids - but ask yourself: if the average person wants to have this unique experience, where are they going to get it? Although a number of hobbies may provide this, the easiest and most direct way to find a "high-performance, high-challenge" experience - and the unique psychological states of that experience - is by picking up a challenging video game and delving into it. It's a part of our lives that games can fill uniquely well, and we should think carefully about taking away a game's capability to give the user these unique experiences.
Conclusion
To sum up: if there's no risk of a large setback, you lose a great deal of tension, and a great deal of interesting player psychology that they experience largely thanks to that tension. Whether you want that sort of tension is your choice for your game; but be aware that you're giving it up when you reduce the magnitude of setbacks in your game, and be aware that may be diluting one of the core experiences that your player can only get from a game (as opposed to a book or movie): a personal challenge, a chance to master a skill, and the chance to test how they perform under pressure.
|
The thing is, this repetition to achieve perfection is attractive to some and repulsive to others. Do you remember the hoverboard-race level in Conker's Bad Fur Day? It has to be in the top 10 of controller-throwing frustration. The reason I kept playing it? I wanted to see what gag was next, what clever parody was in store for me. I had a goal that demanded I continue. I don't find that same motivation to continue in other games.
The main problem with the "hardcore" games is that their challenge/frustration level often outstrips the interest/exploration drive, leading players to abandon the game. Some people have never seen past the first Donkey Kong level, except in the attract screen. However, as games get larger and have more content, there is an incentive by the designer to keep the player moving along, to not throw up insurmountable obstacles (except in sidequests, maybe - Ruby Weapon anyone?), so that the player can enjoy the majority of game content.
This was obviously the approach taken in PoP 2008, where the exploration of the world and the tale takes primacy over a threat of death. Elika "saving" the player is simply an automation of the Dagger of Time from the previous installments - if you had infinite time-reversals and the reverse animation was not played, you would get the same effective result as PoP 2008's approach to death.
There needs to be a distinction between "cheap" deaths and "deserved" deaths. Cheap deaths involve things like falling into pits without warning, getting gunned down because you walked into the wrong place, one-hit-one-kill traps, etc. - anything where death/failure is required in order to learn what to do by learning what NOT to do. A deserved death is one where there are enough signs/warnings/trainings to alert the player of some danger - a noise, a cutscene, or something similar that teaches the player the consequences before presenting the danger. Half-Life is chock-full of these moments - the slippery floor by the elevator shaft, soldiers battling new alien enemies, the scientists running into the sentry gun. Unfortunately, many of the "high-performance" games have a high degree of cheapness in their deaths.
Ideally, we want to give players the confidence and training to see new problems and approach them confidently - in this kind of setting, choking virtually goes away. Prince of Persia: Sands of Time, Shadow of the Colossus, and Assassin's Creed are good examples of doing this. As a player learns the move sets and the limitations of the characters (often introduced in a risk-limited or no-risk way), the player is building a tool-set of how to approach new problems. When the player can chain these tools and techniques together, they are simultaneously achieving "flow" _and_ avoiding choking. Thus, when they arrive at timed-event, high-alert, or otherwise "high-tension" section of the game, they know what needs to be done- while there might be tension, there is also understanding.
Obviously, the game design, camera angles, and controls must be consciously designed to buttress this goal of the player attaining an "aware-flow" state. When everything is done right, it _is_ possible for games to promote a simultaneously small-setback/low-tension and yet high-performance/high-challenge gaming experience.
Anyway, back to your comment. I agree with everything you say, especially about "cheap" deaths (a topic I want to do a separate post about) and your ideas on how to teach the player dangers. You're also correct that games can create a state of Flow without having huge setbacks - VVVVVV's "quick iteration" model probably creates a more continuous state of Flow than, say, Mega Man (a similar, challenging game with bigger setbacks on death).
I do still think that there's a very different experienced created when high setbacks are in play - a feeling of something at stake. But I'll also admit that this is probably only appealing to the most hardcore players. However, maybe designers should think about including a "hardcore mode" that is simply the same game, but with larger setbacks - for instance, VVVVVV could have had a mode that was the same game, but with every other checkpoint removed. Maybe few people would play it, but it would be easy to do and might make the game more appealing to the hardcore contingent looking for an extreme, high-tension challenge. (Blizzard's entire strategy is to satisfy the extreme hardcore AND the broad casual markets, and it seems to work well for them.)
Regarding "content gating"... I've always found it hard to sympathize with this argument, mostly because I don't think that stories and cutscenes are very important to the core of a game experience. Not that they're automatically a bad thing to have; giving the player a neat cutscene can be a fine reward to give the player for completing a challenge. But when people start complaining that they have to play the game to see the story, it should be obvious that this means that your game has serious problems. If you're making your game for the sake of telling your story, well, you've picked the wrong medium. I compare it to making a novel where the reader must complete a Rubik's cube before turning the page each time.
Braid had a good gentle method of content-gating - you always had a choice of many puzzles to complete and weren't forced to complete them in order if you didn't want - and VVVVVV itself did a good job of making it's hardest challenges optional (in other words, it did have a "hard mode", built naturally into the game). When you think of your "content" as gameplay/challenges (and that should be the primary content of your game) the problem of gating content becomes mostly a problem of managing a difficulty curve.
Only today I happened to run across another very interesting article on this subject on a "critical game theory" blog I happened to run across: http://deserthat.wordpress.com/2009/11/26/game-design-the-miyamoto-way-flow-and-
difficulty/
The most interesting part of this article to me was its reference to Flow theory, and its three division between a player being frustrated (too hard!), bored (too easy!), or engaged (just right!)... and the statement that once a "student" (or player) fails and becomes frustrated, then "in order to get the student back to that engagement phase, the teacher has to first bring the student back to the boredom level: you can’t simply drop back into a flow state if the player is frustrated. You have to make them bored first by giving them something very easy. This is exactly why Miyamoto has built level restarts into his games, even if he doesn’t express the reason in those terms."
This is an interesting perspective, and seems to present another argument to say "go ahead and set-back the player back far enough that they have to redo a few challenges."
I would apply it by saying, when designing e.g. a Mega Man level, you have certain auto-save points, which the player will continue the level from when they die.
If you have a particularly hard challenge in the level (say, a miniboss battle) that the player is likely to find difficult and frustrating, you should NOT put the save point just before that battle. Yes the player would be able to quickly attempt the battle over and over... but if it remains too hard for them - which it will if they don't learn anything and increase in skill - then the battle will simply remain frustrating. (A major part of the theory of why "too difficult" is "frustrating" is that the player isn't LEARNING from their failures - they're just failing.)
Instead, a better sequence would be: save point; one or two easy challenges; possibly an "engaging" challenge; and then the "frustrating" challenge (miniboss battle). This accomplishes a few things:
- The player has greater variety in the challenges they face. Instead of running up against a "level 50" challenge 20 times in a row, I'm facing two level 10's and a level 20 first.
- Facing that level 20 challenge (which is "just right" and engages me) is EDUCATIONAL: each time I face it, I'm tempering my skills and learning something - something I can apply to the level 50 challenge (which might be so hard that I'm learning little or nothing from failing against it). Eventually, "grinding" on the level-20 enemy will give me enough experience to "level up" and be able to face and defeat the level-50 enemy. (Weird how this Mega Man example is starting to sound like an RPG - maybe RPGs do a better job of modelling experience/skills than I realized.)
- Defeating the easy level 10 challenges is therapeutic! When that giant-dog-spewing-flames miniboss just kicked my butt again, it's very cathartic to go back and kick a couple of the non-threatening-puppy-enemies around.
In other words: a good reason to set the player back a good ways when they fail, is the fact that that failure may mean they're facing a challenge that they're just not ready yet. Better to set them back to redo several challenges they've already done, from which they'll learn, until they're ready to conquer the challenge that defeated them.
re: content-gating: Narrative/story in games has always been a point of contention. Some examples of narrative in games done right:
- Prince of Persia Sands of Time (yes, I'm mentioning it again) - Much of the exposition was done by Farrah and the Prince as conversations as they are traversing the palace and beyond. A very natural way of exposing story.
- Beyond Good and Evil - Characterization was excellent, which led to a genuine interest in the story. Also, Jade's camera pictures provided story background and sometimes even attack strategy.
- Finally, there is this indie game called Trauma - http://www.gamedesignreviews.com/trauma/ - go watch the video! No, seriously. Do it. I'll wait. . . . Did you catch it too, the way the protagonist's voice seamlessly narrates and changes as the player moves about the environment? It gives such a richer perspective of what's going on, it's just gripping.
What all these have in common is keeping players "in the moment" - the story just works better when controls and camera aren't torn away from the player at the designer's whim - the re-orientations and context shifts kill any feeling of flow.
But anyway, back to the point at hand - Difficulty. Team Meat, the guys developing Super Meat Boy, have some cool things to say about this very topic. Check it out: http://supermeatboy.com/13/Why_am_I_so____hard_/ One of the more interesting features is the ability to see every playthrough in a level replayed at the same time.
I can't tell if this feature is available at any point, or only when you finish a level, but if you _can_ watch the replays while still working through the level, it might be another teaching tool. Come to think of it, I would love to see a Frankenstein'd NES emulator do the same thing with Super Mario Bros. Recording and replaying ghosts of speedrunners or players that perfected the "infinite lives" trick on level 3-1 would be fantastic on numerous levels.
When playing Mass Effect I had one of the worst game experiences of all time: a difficult fight against a Krogan - near the end of my first planetary mission, so perhaps I was "under-levelled" for the encounter - in which I died at least 12 times... and each time I died, the following happened:
1. A Loading screen came up, and stayed up for at least 20 seconds.
2. A cutscene played which I had to skip.
3. A dialog played - the Krogan talking - which I couldn't skip.
4. I was presented with a totally meaningless decision with three options, and had to choose one before I could proceed.
5. The battle finally began, and (on at least two occasions) I was literally killed within 3 seconds.
6. Goto #1. And yes it will again take 20+ seconds to reload the game state, even though I was on the exact same level with the exact same content loaded... what?
This, also, is indefensible. I was learning nothing, I was building no skills, it simply increased my "iteration time" needlessly.
I'm particularly annoyed by the Loading Screen part. I know that this requirement would probably mean a huge change to the game's serialization system, but it seems unforgivable that I should go through a reload that long each and every time that I die.
Quick digression - I can't believe I forgot to talk about Bioshock's vita-chambers in the article above. Can I say real fast: They're terrible, terrible, and terrible? I haven't played Bioshock 2 yet but I hope to God that they fixed the fact that you can simply come back and find your enemies at the same health you left them at, allowing you to slowly "kill a Big Daddy with a wrench." Yes I know I'm not the first person to point this out. But it disturbs me how many game developers dismiss this criticism as if it's not important at all. There is an incredibly obvious degenerate path through your core combat gameplay! If you don't regard this as a bad thing, I have to ask 1) whether you actually know what games are, and 2) why you're calling what you're developing a "good game." (Read my previous blog on "cheese" to see more thoughts on why it IS our fault, as game designers, when we leave a degenerate strategy in our games and it results in players engaging in un-fun activities.)
But although the Vita-Chambers are a broken method of "setback", I bring them up as an excellent counter-example to my Mass Effect complaint... because they involved ZERO LOADING. Zero interruption to the gameplay, and I could run right back to the challenge that I was facing and tackle it again. That is absolutely wonderful, and more games should follow that example... it's fine to just respawn me in the same world, but please reset the challenges I was facing (i.e. reset the enemies' health) so I can face them again. I know this is not an easy problem to solve, but I encourage game designers and programmers to take it seriously and give it a lot of thought.
Also, I think I've now written twice as much about this subject in the comments to the article than I did in the article itself. :)