The Altar of Sequence

First, watch this.  It’s a little long, but it’s super cool.

Done?  Okay, now let’s talk about what just happened.

You don’t typically think of Mario as the sort of guy to traipse around maze-like cave systems, his progression periodically halted by obstacles that must be surmounted by the rewards of further exploration.  A fire flower isn’t necessary for clearing out a tunnel of Piranha Plants blocking the way ahead, it’s just for making it a little easier to complete a strictly linear path.  Right?

But the mechanics of the Fire Flower itself don’t dictate much of anything other than “you shoot fireballs.”  The constraints of the item are imposed by the level design.  And the same is true when applied to the genres as a whole.  What this level in Mario Maker has pointed out is that “Metroidvania” is a question of how levels are designed more than anything else.  Even the mechanics of (what would normally be) a railroaded 2D platformer work surprisingly well in the context of exploration-based gameplay.  (This isn’t actually a new idea; I’d argue that Metroid fell down where Super Metroid rose to the occasion based on factors except level design)

The limits, then, are more a consequence of design of the game world.  Even Nintendo franchises whose games once had more open than linear worlds now tend to opt for the straight and narrow path.  Of course, back in the days of Wind Waker and Metroid Prime, these games were still not programmed well enough that the intended linearity mattered; in those titles, it’s possible to obtain major items and complete long-term milestones in a vastly different order simply by knowing the right bugs and even abusing intended mechanics.  In a way, the difference isn’t that games’ structures have changed recently – it’s that programmers have gotten better at rigidly enforcing them.

This ties in somewhat with the increasing emphasis on narrative and puzzles in games, as opposed to the purely action-driven forays on the NES.  After all, the more your game focuses on story or the player following a specific path, the more the experience stands to lose when the sequence is broken.

But I wonder if it doesn’t also have to do with a creativity drought among game designers?  What I see in a lot of games now is analogous to the concept of “teaching to the test” in schools.  Rather than designing challenges as a set of goals to be reached by any of the vast variety of means the game naturally affords the player, developers set up a goal, a starting line, and a definite plan for every step they want the player to take in between.  God forbid that any powerup or item be used for any purpose except the solving of puzzle specifically laid out by developers.  There’s a delicious irony in the player having to “figure out” less than ever in modern games that are often praised for being “smarter.”

This shouldn’t be taken to mean that a “hallway simulator” has no place in the medium.  Sometimes you sacrifice an explorable world and flexible challenges to highlight a narrative and precisely define the interaction you want.  I actually liked Final Fantasy XIII and even Metroid Fusion as a change of pace.  It’s when each and every game is built like this when I worry that developers have forgotten how to do anything else.

For the Challenge, Part II

Difficulty can, of course, be affected through programming as well as or alongside design.  Mario was programmed with hard limits on his jump height and the levels were designed accordingly.  There are some games where programming and design are complementary, and the result for the player is beautiful, lively physics.  The design and the programming tend to be set in stone for games – especially the latter – and that’s a good thing!  You need consistent ground rules in order for your game to be considered as, well, a “game.”  But when some elements are static, like the programming, other elements are often demanded to be dynamic, like the difficulty.

The most classic example of what I mean by “dynamic difficulty” can be found in the presence of multiple difficulty settings/modes in a game.  Easy, medium, hard – or if you’re feeling creative, “Kid Mode,” “Impossible,” “I am Death Incarnate!”, and so on.  They’ve existed for a long time, often as adaptive measures to player testing of a “baseline” difficulty.  The idea is that difficulty can be more flexible by tweaking certain parts of the game one way or the other depending on this setting.  This is more prevalent in games designed around complex stats and numbers – less so in platformers like Mario where player “stats” are as simple as failure upon taking one or two hits.

The problem is these tweaks are usually just that: tweaks.  Developers can easily program things like player/enemy stats to scale with the difficulty setting, but when they take this approach, every other aspect of the design is still operating under “normal mode rules.”  If you design a final dungeon with an arduous gauntlet of tough enemies, then double their stats as a hard mode condition, it’s unlikely that the difficulty of the same dungeon will remain well-balanced and tested.  Instead, it’ll simply take the player at least twice as long to complete a segment that was already balanced for a completely different pace – or else it’ll be unintentionally and incredibly difficult relative to the player character’s capabilities.  I mean, I’m convinced that oftentimes developers don’t even bother to test alternate difficulty modes at all, but that’s a little harder to examine and prove.

All joking aside, sometimes it’s painfully obvious that a game’s multiple difficulty levels are implemented more as an algorithm than a series of design changes or even tweaks.  Look at Fallout 3 or Skyrim.  When you slide the difficulty up or down, the only alteration you see is how much damage you deal and take.  Enemy configurations and dungeon puzzles remain exactly the same throughout.  The same as what they were when balanced for a “medium” difficulty (Fallout: New Vegas at least gets credit for introducing a “hardcore” mode that actually adds extra features and overhauls the mechanics).

I find this to be a slightly lazy solution to the difficulty “problem.”  What it really does at its ugly core is change the pace of a game, rather than the difficulty; you take twice as many swings at enemies, have to spend twice as long dodging them, and probably die a lot more.  Essentially, longer and more numerous iterations.  Like other fake difficulty, what this comes down to is a waste of the player’s time rather than a test of their skill or tactics.  I could seriously come up with some kind of fancy axiom… “the amount of enjoyment derived from a video game is inversely proportionate to the amount of control the player has over its pacing.”

A smarter way of adapting difficulty higher or lower might be changing enemy AI to not fall for the same tricks: a prime example would be the fighting game genre, where CPU opponents of higher levels will imitate correspondingly more complex tactics and reflexes.  Take level 9 CPUs in Smash Bros. games, which can destroy beginner players in much the same way a more experienced human opponent would: through better coordination of offense and defense.  Sure, you do often spend more time playing against a high-level AI than a low-level one, but the pacing is ultimately dictated by the skill of the player rather than the amount of bullets the enemy can absorb before going down.  If you practice a little more at Smash Bros., you’ll eventually be able to conquer an advanced AI (or human player) faster and faster as you adapt to the level of play.  If you practice a little more at Fallout 3, that irradiated mutant animal is going to sponge the same damage at the same rate no matter how much you wish it would die quickly.

But even stat changes can make for a more interesting challenge if balanced thoughtfully.  Look at Critical Mode in Kingdom Hearts II: Final Mix.  Sure, enemies are programmed to deal out more damage, and the player’s max HP and MP gains are cut in half, but at the same time the player does more base damage and starts out with more abilities and ability points.  This results in a real dynamic difficulty where ultimately the player can take their time with cautious play to offset increased risk, but also apply skill and knowledge of the mechanics to actually complete the game faster.  In other words, the game rewards a higher difficulty with a potentially more streamlined pace.

But you’d better believe it took more thought to implement Critical Mode in Final Mix than “Very Hard” mode in Fallout 3.  There simply isn’t a shortcut to well-designed difficulty.  It bears repeating: simply making gameplay take longer does not make it more (legitimately) difficulty.  The whole experience has to be fun for the player in exchange for (and in proportion to) the extra effort.  Even if that requires a little extra effort.

When Ya Got Options

I’ve been wanting to write about optional content in games for a while, but couldn’t quite formulate anything.  Having just finished 100%ing all three of the original Ratchet & Clank installments in their HD Collection, the ideas are a bit clearer in my head, which will hopefully carry over in writing.

Optional content is, ironically, one of the primary things I seem to seek when I’m looking to get into a new game.  It adds longevity to the experience, fleshes out game mechanics to a degree not seen in the “main” parts, and just all-around deepens a game.  Even in the most basic sense, optional content denotes that there’s more of the game there.  Of course, in a modern game industry that sees increasing genre flexibility and open world gameplay, what exactly constitutes “optional” as compared to any other content can get nebulous.  This is almost a whole post in and of itself, but the games I use as lenses through which to view my ideas here provide a fairly clear distinction, so I won’t get too far into it right now.

As it pertains to Ratchet & Clank, optional elements have always been around in the series, but the first two sequels enhanced their presence and offered more incentive to pursue them.  They expanded on the original by implementing rudimentary RPG elements like experience-based upgradeable health and weapons as well as escalating enemy stats to match.  De facto “level ups.”  Also added were “arena” levels for pure combat challenges, sprawling “wasteland” environments with collectible treasures to sell to NPCs, and even more sidepaths offering helpful secondary gadgets as a reward for completion.  All in all, these were pretty well-implemented.  The arenas were so much fun that they’ve since become a staple of every main game in the franchise, and the RPG elements have arguably seen even more emphasis with time.

All these things add to the game, it’s true, but what if the additions aren’t good?

“But wait a minute, you dirty hypocrite,” I hear a heckler saying.  “You just said they were well-implemented!”  Well, yes.  The optional content in R&C is, as a rule, much better-implemented compared to a lot of comparable stuff in other games.  However, whenever you create sidequests, and especially whenever you install a system of experience-based player progression, you run the risk of also introducing poorly-tuned time and XP curves.

Take the second R&C game, for example.  Each weapon can be leveled up to increase its damage output and give it extra effects.  However, the required experience for each one to do so is rather off-balance.  Where the Pulse Rifle mercifully upgrades in a very short time considering its small ammo capacity and limited usage, another weapon with comparable properties like the Synthenoid takes an obscene amount of time.  Worse, “revisits” to a planet cause every enemy to give a fraction of the experience they originally did, making repeat runs for level-grinding even more tedious.

A similar phenomenon can seen in the grind for Bolts, the in-game currency, of which a few million are needed to buy everything in the game.  But enemy bolt drops and bolt containers are also heavily reduced on level revisits.  This and the curtailed XP gain are both probably measures put in place to prevent farming, but the big design problem here is that the XP curves and in-game economy necessitate farming to achieve 100% completion.  In reducing its effectiveness, you don’t cause players not to do it, you just cause them to waste time doing it for longer.  All you have done is make full completion more tedious.

The third game was actually a little better about the amount of grinding needed to achieve 100%.  Even though there were twice as many upgrades for each weapon and even more escalation in their selling prices, XP and Bolt gain was faster, more standardized between weapons, and less nerfed during backtracks.  You needed to farm enemies for much less time to obtain and fully upgrade every weapon.  This became apparent during my recent playthrough, where the amount of time spent actually playing through all the content was longer, and the time spent farming was shorter.  My congratulations to Insomniac for realizing somewhat that artificially inflating the amount of time it takes to reach 100% is bad design, and not to be ungrateful, but there was still an awful lot of grinding.

Grinding in video games is essentially – that is, at its core – the repetition of a simple action to increase the player’s stats or provide additional in-game resources.  This should never be used as a way to add length to or substitute for challenge of the optional content of a game.  For me at least, it’s just a little too close to real-life work for a paycheck or a school assignment (they’re both known as a “grind” for a reason).  In fact, part of what makes grinding so undesirable in a game is that I’m trying to escape exactly that in real life.  Rewards in games should be earned, true, but it should always be fun rather than actual work.

“Chillax bro,” the heckler might retort, “it’s all optional stuff you’re talking about anyway.  It’s not like anyone is making you do it as part of the main game.”  Well, sure.  Optional content is by definition part of a game that will go unseen by many players.  Because so much less attention is paid to it by the player, purely practical prioritization in designing the game means that much less attention should be paid to it by the developers as well, including things like tuning and iterations.

However, my contention nonetheless is that, as part of a game, it still has to be fun.  I used the example of grinding to bring out my point because it’s just about the quintessential enemy of fun in my book, but make no mistake; the principle applies to poorly balanced difficulty, unpolished level design, and unstable programming just as much as it does to level grinding. The content’s status as optional should not mean that it’s equally optional for it to be enjoyable.  Make it engaging.  Build on concepts already established.

Design the game with the philosophy that “play” is the relevant verb.

By the Numbers

Let’s talk about the black sheep of the Kingdom Hearts franchise: Chain of Memories.  This one still sticks out like a sore thumb next to the other installments: it had sprite-based graphics, strange (but oddly charming) writing, and especially card combat.  The card combat part has proven quite polarizing among the fanbase, which is kind of understandable.  The first game, which was the only other one out when CoM was released, had quickly and easily accessible gameplay that didn’t run roughshod too much all over the wonderful aesthetics and sweeping crossover ambitions.

CoM is almost a direct antithesis, recycling the same worlds wholesale and slapping on excuse plots for those (leaving the freshest material for the non-Disney segments) while its gameplay mechanics demanded the player’s full attention.  Every action you take is determined by playing a card, and almost every card exists as part of a numbers-oriented system of “greater than, less than.”  If you play a 4, an enemy can play a 6 and “break” your card, for example (“0” cards have the distinction of breaking any card but also being broken by any card depending on when they’re played).  Your ability from the first game to Guard attacks directly is gone, and Dodge Roll is much less useful, so card breaks are an essential mechanic to self-defense.  For added fun, you can combine up to three cards (sacrificing the first card chosen for the rest of the battle) in a “sleight” for a special combo attack – as well as higher numeric strength, given that the three card values are added together for the attack.

To the player who enjoys more “pure” action games this probably sounds like torture, but it actually works very well, and here’s why in a word: visibility.  Most JRPGs, for all their reliance on numbers, tend to make those numbers highly opaque.  If you want an example, just look at the EV mechanic in Pokemon or the way character stats affect their offense/defense in Final Fantasy.  A complicated series of behind-the-scenes equations drives most of the character progression and leaves the player with just a vague sense that a higher STR stat means harder hits.

With CoM, all cards are on the table.  So to speak.  If you play a 7 and your enemy plays a 9, your card gets cancelled and you immediately know why.  If you want to access a certain sleight, the game tells you exactly which cards of what numeric values to combine to use it.  If you want to engineer your entire deck around a set of boss-annihilating DPS sleights, or only load cards with higher numbers than what you know enemies can play, that’s entirely doable.  To wit, the mechanics are conveyed well enough that you can master them early on by using strategy and basic arithmetic, rather than simply follow them with varying degrees of awareness throughout the game.

And the rules of the game are consistent for enemies too.  Mobs have access only to pre-determined card values, thankfully never 0s, that are also shown on screen – such that you can learn them after a couple battles and revise your deck to counter if you want.  Bosses have the same options as you in terms of playing 0s and sleights, but also the same weaknesses.  Their 0s and sleights can be broken, they can be attacked while reloading, and they still have finite cards in their deck to play before doing so.  If you take the time to optimize your deck, you can trivialize battles.  And no enemy in the game is exempt from all these rules.  It’s all-in-all more fun to learn the combat system here because it’s always expressed clearly.

A “what you see is what you get” approach serves RPGs better than you’d know from examining a lot of entries in the genre, since the defining trait of their gameplay is that it’s based on strategy rather than timing or execution.  How they tend to play out, though, is that having stats is the key to victory, not employing them.  The reason many of them can’t have clearer expressions of how their mechanics work is that it would reveal how little the battles are influenced by the player compared to Random Number Generators and the raw stats themselves.  The near-automatic battles in Final Fantasy XIII get a lot of flak for their nature and exclusion of the player from the whole process, but it’s really just more honest about the player’s level of intellectual involvement than other installments in the series.  These strange and large numbers floating above the characters’ heads don’t have to get in the way of the graphics and the narrative that way.  And that’s fine, if that’s the kind of game you want to make.

Just, if you want to make one where you claim player involvement is supposed to be central, then make sure you communicate the rules of your game fairly enough that they can be involved and have it matter.

Patchwork Solutions

Recently I made a post on the Terraria subreddit regarding the most recent boss added to the game, the Moon Lord.

For the uninitiated, the Moon Lord is far and away the toughest enemy in the game ever to be added.  It deals and tanks massive damage and drops the current best equipment and resources in the game.  It has been referred to by the developers as the “final boss” and been subject to no small amount of hype.  Come the 1.3 release, the only turn of events faster than players’ finding ways to “cheese” (come up with methods and game mechanic exploits to trivialize the battle) Moon Lord was their reaching the point in the game where they fought him in the first place.

Then came the patches.  1.3.0.2-1.3.0.7, all of which contained both blatant and subtle buffs to the Moon Lord and/or nerfs to the players’ preferred cheese methods.  Where before the boss was vulnerable to HP draining attacks, he now inflicts a debuff that blocks them.  Where before the player could use teleportation back to home base and quickly heal while he flew to your location from a distance, he now teleports along with you instantly.  Where before players would exploit the damage from spike traps and weak enemies to trigger invincibility periods against the Moon Lord himself, the Moon Lord’s attacks now operate on a separate timer than every other damage source in the game.

It seems the developers have been systematically targeting and amputating unintended methods of defeating this boss, and this is where my post on Reddit comes into play.  In it, I argued that these patches aren’t such a great thing for the game experience overall.  I won’t reiterate the entire post here, but it definitely spawned some… er, controversy as to the nature of developer intent, and what the player “should” be allowed to do in a game.  A lot of commenters made some interesting points, on both sides of the issue.

My stance, in short, is that game exploits, if they require the player to go out of his way to use, do no harm to casual and “normal” players, and enhance the experience for those who enjoy using them.  Why is it, then, that developers in the age of DLC and patches seem so hell-bent on enforcing their own “version” of how to play the game?  Given the medium’s fundamentally interactive nature, the user experience is probably as important as it ever could be in a work of art.

I find that the games I most frequently return to are the ones that give me some element of control over the pacing, difficulty tuning, or other aspects of play.  Minecraft.  The Elder Scrolls.  Terraria.  All highly dependent, even essentially so, upon the player to decide how the game mechanics come together to make the experience.  Even franchises like Metroid provide examples of the same principle (albeit to a lesser degree).

I think the best relationship between developer and player is one where the developer wants the player to succeed, gives him the tools to do so, and lets him decide the rest.  In the age of patches and balance fixes, the relationship is in some danger of becoming player vs. developer, rather than player vs. game (please note that I’m not addressing the implications of the use of exploits in multiplayer scenarios, where I’m a lot more likely to be against it).  What I mean by this is that the developer begins to see patches as ways to curtail “naughty” player behavior, rather than ensure that the game is stable and enjoyable for as many people as possible.  If you remove sequence breaks from a special edition re-release of your game, but leave in game-breaking camera bugs or softlocks, your priorities might be a little skewed.

Developers are always in danger of it, but they should never forget that just because their job is to place obstacles in front of the player doesn’t mean they’re on opposites sides.  The more “open” the game, the more their preconceptions of how the game should go take a back seat.

Power to the people, stick it to the man, etc.

“Just” a Game

The introduction of Metroid Prime does much to bring its series into the world of modern gaming after skipping a console generation, making the 3D leap and boasting surprisingly slick controls and fantastic presentation (every bit as good today, frankly).  It also serves as an “excuse” to strip Samus of her signature abilities granted by upgrades to her power suit.  To explain why the hunt for power ups was necessary in a sequel after collecting them all in the previous installment was uncommon.  Usually the audience just accepted that the game would be boring if they began with access to every item from the get-go.  It’s about the journey, after all, especially so in the Metroid series.  But nope, says Prime, we’re going to explain why the planet crawl and item progression is necessary.  And so a quick cutscene was devised wherein an explosion threw caused Samus to take critical armor damage and lose most of the upgrades it had.  Now players could feel completely comfortable with the need to re-collect all of them, without the inconvenient narrative hangups.

The concept of “Gameplay and Story Segregation” is a pretty familiar one if you’ve played video games for basically any length of time.  It’s almost omnipresent in the medium, and probably one of the few conventions to exist solely as a result of games’ own inherent nature: interactivity (as opposed to their borrowing plots, physics, presentation, or other elements from film or TV).  Because it tends to highlight the tension between player actions and writing decisions, one sort of tends to get the impression that it’s a bad thing.

Even if we grant this (which I’m not sure we should), though, sometimes the cure is worse than the disease.  The bigger impact the player has on the progression of the game, the more egregious it tends to become when writers try to maneuver around the resulting contradictions.  This can lead to some clumsy-looking/sounding ambiguity in dialogue and other text or voice acting; and it can  raise issues with continuity when sequels roll around.

Back to Metroid Prime, whose intro explosion set the stage in the simplest possible way.  It’s more than its distant ancestor, Super Metroid, did to explain why Samus was only armed with the basic beam back in 1994 (i.e., nothing at all).  This fit overall with Prime, which was ever so slightly more focused on the narrative than Super.  In a way, that one simple explanation foreshadowed the equally simple but elegant presence of the lore in the rest of the game world, which, thanks to the scanning mechanic, was only there to the degree that the player desired.  Still, as an isolated incident, it wasn’t exactly incredible – “boom, your power ups are gone” in a literal sense doesn’t integrate much better with the rest of the story than it does in a figurative one.

The sequel, Echoes, on the other hand, took every aspect of the narrative one step further.  There were more mandatory cutscenes, more mandatory lore, and more integration of the loss of power ups into the gameplay.  Prime gave you one vague objective (“track the Space Pirates and stop them”) that snowballed as you opened up the map, and whatever you found along the way to help was fair game.  Echoes spelled out the main conflict explicitly from the start, but it also set up the recurring theme of getting your own abilities back from invading the monsters that stole them – much more implicit, but still central to the game.  Showing Samus lose her items to the enemies in Echoes arguably worked even better than the one-off cutscene explosion in Prime because it gave the player (not just Samus) targets to hunt.  This was much better plot integration, a point of real development in a sequel.

Here’s the thing, though: it created a bit of an issue for future installments.  Lo and behold, Metroid Prime 3: Corruption comes around.  The developers eschewed the ability loss this time around, and it might seem puzzling at first.  After all, the Prime series had always been praised for its marriage of the franchise’s signature gameplay with a more modern narrative sensibility.  Why lose that?  Personally, I think the developers realized that it might be getting a little tiresome to keep coming up with new ways to justify the treasure hunt from the standpoint of story.  And for all of Corruption’s departures from the previous games that didn’t turn out so well, I think this one was spot on.  After all, the only thing more important in good game design than pushing boundaries, inter-series as well as internal to individual games, is quitting while you’re ahead.  The more you disguise the same tropes, the more the disguise is going to look like a pair of Groucho glasses (I’m looking at you, Pokémon).

It became tragically ironic when Other M eschewed the eschewing a couple years later, perfectly displaying the danger of trying too hard to justify gameplay with narrative.  Instead of Samus losing her abilities and finding them all over the course of the game, we had Samus lacking military authorization to use them and progressively activating them only as instructed by the general in charge – ostensibly because of their sheer firepower and the danger they would pose to other soldiers (who, to be fair, are usually absent or already dead when Samus arrives on the scene).  But this explanation fell apart in the worst possible way when Samus couldn’t or wouldn’t activate a purely defensive function of her suit in a life-threatening situation.

Ultimately this was one of the most damning incidents in Other M’s plot, and it definitely didn’t do the rest of the story any favors when fans leveled criticisms against it for portraying an abusive relationship between Samus and her “father figure” commanding officer.  It’s far from the only culprit involved in the game’s overall mixed-at-best reception, but it is the best example I can imagine of pushing too far to explain things that don’t need to be explained.

I think there’s this assumption among game writers that the old tropes and tricks, such as starting your protagonist from scratch in the sequel without any previous power ups, are just too juvenile to work in a game meant to be taken seriously.  But titles like Other M make me wish developers would accept more often that sometimes it is enough to say “here it is” and leave the rest to be explored.  I would contend that if gamers wanted everything in the experience to be told to them rather than left to be determined, they would be reading novels or going to movie theaters.

There’s nothing wrong with pre-determined narrative experiences, of course, but video games have the unique potential to create that tension between what the developer presents and how the player handles it.  Perhaps it’s not always a bad thing when the gameplay doesn’t quite align with the story, and perhaps it’s not universally desirable for them to be mangled to fit in with each other?

When you as a player make something happen in a game that maybe wasn’t quite intended, that contrasts with what is pre-determined by cutscenes, it’s a reminder that this is an experience where you too have a degree of control.  I think that’s an important aspect to the appeal of video games as a whole, and that’s powerful.

For the Challenge

I think a lot about the concept of difficulty in games.  Why can I be totally immersed in one difficult game when another just as totally turns me off immediately because of that difficulty?  What about that whole concept of “legitimate” difficulty?  When is it fitting for a game to be difficult, and when does it just ruin the experience?  It’s a huge aspect of what makes or breaks a game, and there are a lot of different angles to explore.  Here’s one of them.

It seems like a good chunk of the market these days has become geared toward “challenge” games, or maybe more accurately, “exercise in frustration” games.  Is it about the masochism inherent in the video game experience?  Exploiting a sense of nostalgia born from a perception of older “classic” games’ being less forgiving?  Simple appeal to every gamer’s desire for bragging rights and a sense of accomplishment?  I can definitely sympathize with these things, since I do love experiencing them myself.

The problem is that not nearly as many games as claim to hold this appeal do it well.

Take Dark Souls, for instance (because why not start off writing about video games by bashing a nigh-universally adored favorite).  The entire selling point of the game is how difficult it is.  Merciless and unforgiving.  Feed your inner masochist.  Relive the days when video games didn’t hold your hand.  Feel the ego boost when you finally defeat that boss.  The problem is that it tends to conflate “difficult” with “miserable.”

The game balance is tuned such that the player is on guard constantly, with even the lowliest enemies presenting legitimate threats.  Credit where credit is due, when you play Dark Souls you have to observe and adapt a lot.  Where I start to lose it is the point where the game discourages you from doing just that.  Exploration and experimentation are usually unnecessary risks because of checkpoint starvation combined with harsh consequences for dying.  But in a game where the tagline is “prepare to die!” it becomes questionable to punish deaths so severely.  If the expectation is that death is a normal and frequent part of the gameplay, that could be a really interesting opportunity to explore the concept of trial and error as a learning method.  Instead, the player learns early on that even though Dark Souls includes frequent deaths, it’s ultimately not designed around them any more than most other games.

Consider how in-game death is used in a “normal,” (i.e., not Dark Souls) game: as a deterrent, a way of saying “oops, you dun goofed.”  Given this usage, a well-tuned game will give you the knowledge and training to avoid a hazard consistently such that when you do meet it, you’re expected to react to it based on what you’ve already learned.  This is the entire foundation on which difficulty curves in games are traditionally built.  You know how to string together complicated jumps in world 8-3 of Super Mario Bros. because you worked your way up to them building on the skills you learned in world 1-3.  If you should fall down a pit, the game restarts you from the beginning of the level as a way of saying “come on now, bub, you should know better than to do that.”

Dark Souls can’t make up its mind on whether death should be avoided or embraced, though.  Again, the tagline is “prepare to die.”  That is entirely accurate, because right from the start, enemy stats and attack patterns make it clear that you’re in for it.  It’s also entirely inaccurate, because a good chunk of the game is spent with the player nearly paralyzed by the fear of losing all his progress and gained experience/currency as the consequence of that death.  If your game’s main method of teaching the player is also something to be avoided, what you’re essentially telling the player is “avoid learning.”  Put another way: when the entire point of the game is to be difficult, the player must be given the tools to adapt to it, whatever form they might take.

Let’s look one of those forms.  When you play Super Meat Boy, you definitely don’t expect an easy time of it.  Like Dark Souls, it’s gained a reputation for being tough as nails, and for killing its eponymous character liberally.  Unlike Dark Souls, however, those frequent deaths are an integral part of the design, and they work as a teaching tool because they don’t punish the player.  “You died on this insanely difficult passage?  No worries, there was a checkpoint about ten seconds ago.  Keep trying until you learn to do it right.”

It’s easy to think of checkpoints as “easifiers,” and the lack thereof as “challenge enhancing.”  But I would contend that not only do frequent checkpoints not lessen the difficulty of Super Meat Boy, but they encourage the player to keep plugging away at the more difficult portions without worrying about death as deterrent, and they allow the design to throw some nasty challenges at you.

Imagine how a change as small and simple as giving the player a checkpoint upon traversing pre-boss white mist would do wonders for the pacing of Dark Souls.  Bosses would feel so much more justified in the crazy offensive and defensive power they possess because their fights would be friendlier toward learning their patterns and capabilities.  As the game is, though, you don’t get to take another crack at a difficult segment until you recover your progress (and in the case of many bosses, take even more time on top of that to set up buffs, summons, and consumable items that you lost).  What part of that cycle is meant to encourage the player to keep attempting to meet the challenge?

Much of the good design present in Dark Souls is completely overshadowed simply by the omnipresent bad design of punishing the same trial and error it requires to progress.  But let’s be clear, this problem is present in more games than Dark Souls.  The otherwise excellent Terraria requires the player to complete a lengthy endgame quest to fight the absurdly difficult final boss who essentially requires several failed battles before victory can be achieved.  Many JRPGs such as Final Fantasy and even Paper Mario will require the player to traverse a good portion of a dungeon or sit through lengthy in-battle cutscenes following a death due to a boss attack that was literally impossible to anticipate.  The original Metroid respawns Samus with critically low health after death, prompting tedious farming for energy drops from enemies.

All of this amounts to a tedium, wasting players’ time on parts of the game that aren’t challenging rather than letting them spend it conquering the parts that are.  Much of this has to do with padding game length rather than deliberately trying to be punishing, but whatever the reasoning, I don’t think it’s entirely fair to view it as real difficulty.

The player should never experience fatigue while playing simply because he’s tired of the process of continuing from a failure.  If your main appeal comes from the challenge of your game, that doesn’t mean you have license to make everything obscenely difficult and ignore tuning it to the player experience.  The onus is actually even heavier on you, because the game at its core still has to be fun, and it still has to encourage the player to engage with it.

This is harder to do if your game is hard.  If you’re not up to the challenge, then, as the Dark Souls fanbase will universally tell you, “git gud.”