Designing Failure: Making loss core to the game

This is the 3rd and final post on failure in game design.  This time around, rather than focusing on tacked-on death penalties or mitigating the impact of harsh loss, I instead want to focus on embracing failure so thoroughly in your game design that it just becomes a smooth part of the experience.

Dual Gameplay Modes

Payday 2 is not a game very everyone, but there's a very strong core design behind their gameplay that there's a lot to learn from.  In a way, once you're fully invested in it, I see it as a strong idealization of what stealth gameplay has touched on in many games before.

Many of their levels go like this:

  • Equip yourself with a variety of gear, either tuned towards stealth or towards heavy combat, sometimes with special tools for handling certain situations, like iron bars you'll need to cut through, cameras you'll need to disable, or safes you'll need to blow open.
  • Enter the level in stealth mode with 3 friends.  Examine the situation by identifying cameras, guard patrols, and civilians that might spot you.
  • Engage in stealth play.  Discretely take out guards without being seen.  Break into the security room and turn off the cameras.  Hold up the civilians before they run.  Stop the cashiers before they press the silent alarms.  This gameplay is very deliberate, requires planning and coordination, and is very challenging.
  • If you fail, which you often will at some point, you pull out your guns and try to do the heist the hard way, with SWAT streaming in trying to kill/arrest you as you cover doors and crack the safe at the same time.  Hectic high action.  If you succeed, you try to get your score safely to an escape.

The design I really appreciate here is the idea of a very challenging stealth game that falls back on a more straight-forward action game.  This is something common to stealth games, but accomplished on particularly potent high-level in Payday 2.  Succeed in the hard mode, and you make it out with more loot more safely and more quickly.  Fail, and the action is still a fleshed out very satisfying experience for good reward.

Through this design they achieved very high challenge and very high stakes but removed a good portion of the sting of a Game Over screen.  They made failure an immersive world experience that you roll with as part of the action.  And they gave you an environment where you can continually practice the hard version of the game, safely knowing that you have a fallback that will be fun too.  It lands a great deal of the immersion and tension that usually comes with failure, but with few noticeable downsides.

Player critics will point out some rough online social behavior that results from this depending what difficulties you play at, which I'll acknowledge, but I think there's some very strong potential in these types of designs that's worth taking forward to new places.

Failure Fallbacks

Tom Francis wrote a fantastic post about Metal Gear Solid V's Failure Spectrum that you should absolutely read:

Rather than attempting to recount this in a less effective form, here's the critical exert summarizing how it plays out:

When you can fail at something but still carry on playing, I call the range of states between perfect success and total failure a ‘failure spectrum’.

MGS V has most of the stealth genre’s most generous failsafes, plus an incredibly generous one of its own inserted at the crucial moment – Reflex Mode. The result is something like this:

  • If a guard sees you, you get an ‘awareness’ indicator showing you where they are. If you reduce your visibility, that goes away completely and the guard won’t even investigate.
  • If you stay in sight and/or make yourself more visible, the guard will very, very slowly come over to investigate. Even then, this alerts no-one else and doesn’t count against you in any score or performance metrics, and you don’t even have to move: going prone and using a ‘hide’ button makes you damn nearly invisible – I’ve had a guard stood 2 feet from me shining a torch directly on my body without spotting me in that mode.
  • If they DO definitively see you and recognise you as an intruder, Reflex Mode puts the world in slowmo and you get a huuuuuuuuge amount of time to do something about it. Your view is snapped to the person who saw you, the yelp of recognition they make seems to be inaudible to other guards, and if you shoot them in the head with a quiet weapon (you start with two) in this ample time, no alert is triggered.
  • If you fail to take them out in this time, or someone else sees them die, the surviving guard will yell. Others in earshot will be alerted, but no-one beyond that at this stage. Your default weapon is rapid fire, accurate and silenced, and if you can take out everyone who heard before they have a chance to radio, the alert is contained.
  • Even if you do give them time to radio, it will do nothing if you’ve already taken out their communications equipment.
  • Even if they manage to radio for reinforcements, it’s easy to run away and they won’t give chase.
  • Even if you don’t run away, it’s quite possible to kill everyone without taking a hit.
  • Even if you take a hit, your health regenerates for free.
  • Even if you get hit a LOT – even if you get hit by a mortar – you only go into a ‘wounded’ state that restricts your movement but still gives you a chance to take everyone out.
  • If you fuck that up, yeah, you’re dead.

The core idea being that you create opportunities for failure, and then you fall back on a new series of gameplay options that are interesting in their own right, which you can also fail at to fall back on another, etc.

In my ideal designs, I'd leverage this type of strategy to put more challenge at the top of the spectrum, and fall back on more and more forgiving mechanics the further down the spectrum you go, doing everything I could to payoff and reward the challenging gameplay in particular, but the value stands either way.  It creates opportunity for the player to push themselves against challenge with noticeable results if they don't succeed, but significantly lightens the string of failure by not making it read as failure at all, and instead still offering something meaningful on the other end.

This is perhaps one of the more ideal forms of mixing a type of design that is very approachable with something that still creates opportunities for failure.  If it is too forgiving, it still sacrifices much of the immersion and tension that comes from more meaningful consequences, but it's a fantastic example of striding the full range of design considerations while erring on the side of ease.

.Reflections on Designing a Deeper Game

Across these posts, I see these examples as designs that are recognizing the value of failure and consequences, and that are treating them with respect.  This is something that I think is tragically infrequent with many modern game designs that omit real failure almost entirely, often losing some of the meaning of the game in the process and not finding any effective way to insert it elsewhere.

I've been a gamer for a long time, and I'm prone to appreciate more intense/challenging games than some segments of the audience, but my priorities for designing a meaningful game experience go somewhat like this:

  • Make a game that is deep and lasting.  A game that creates opportunities to learn and improve.  A game that makes you proud to overcome challenges.
  • Appreciate the value of failure in this context, and then look for methods of creating fallbacks that are are still part of the game's core fantasy.  The Second Wind when you're wounded.  The action when the stealth fails.  The exciting extraction or escape mode you get to play out when things go south.  Spend real time on this as part of the intended world you're presenting; don't treat it as a less-fun afterthought.
  • Use your flow and your rewards to encourage players to approach the challenge in a way that's not overly obtrusive.  Make it natural for players to challenge themselves, but don't make it obnoxious when they need to fall back on the secondary options.
  • Plan out the steps of progress that players will be able to take to learn and overcome more and more difficult challenges from an early learning point.  Don't be too heavy-handed, but make sure the path is there for those that are personally trying to improve.

Ultimately, all design is full of trade-offs, and making the decision to make your game more approachable at the expense of depth is a legitimate one that you can logically make.  Particularly true if yours is a game looking to social behavior for its meaning, or if your game is intent on being a relaxing escape above all else.  My goal if anything is to illuminate that there's value on both sides of this design choice, and that the potential of the deeper lasting game may actually carry with it measurable weight for success as well, if not more so in some cases.  And it may provide a more meaningful experience for players in the process, whether it's acknowledged it in surface-level feedback or not.

Or if nothing else, hopefully these posts at least provided some interesting areas to think about along the way.  Thanks for reading!


[My thoughts and opinions are my own.  They are not those of Blizzard Entertainment, and they do not necessarily represent Blizzard design philosophies.]