« Suppose Death is a Woman, What Then?¹ | Main | SF Link »

February 09, 2009

Managing Difficulty

Lately I've been playing Buffy the Vampire Slayer on last gen's Xbox, and it has stirred up a few thoughts I have about difficulty, mostly because it gets it so horribly wrong. I've been gaming a long time, and have come up with a long list of must-haves for most games, particularly games which target the mainstream audience. In my career at LucasArts I helped steer difficulty in some specific directions, some bulleted below, and I actually got a game credit in the "hey, thanks" list for a late but timely suggestion to the project's design director when he used it whole cloth.

The other thing that I ran across in the last few weeks was a little video project by a blogger where he discussed what he felt was the most innovative game of last year -- Prince of Persia, which in a way dropped difficulty altogether, by making the Prince more or less invincible. The Prince was accompanied by a companion who would rescue him when he misjudged, bringing something we saw the beginnings of in Prince of Persia: The Sands of Time to fruition: a less punishing form of death. Now, I'm fairly certain that I would prefer the latter to the former, but I understand the impetus to applaud the designers. After all, they took a thorny problem and tried something different -- they eliminated difficulty altogether.

Now, bear in mind that I'm targeting mainstream games -- these bullet points are not for games like Ninja Gaiden, which use their difficulty to club gamers into submission. That is more or less its design goal -- to provide an extreme level of challenge, and managing difficulty for them is and should be to make the game as difficult as possible. Similarly, performance games like Guitar Hero, which have difficulty levels but for which the practicing and not the "getting through the narrative bits" is most of the fun, are exempted -- they should adopt and have adopted some of these, but ultimately, it's not what they're about. Here I'm basically talking about mainstream-targeting games with a narrative through-line, primarily action-adventure titles and shooters.

(Program note: I started this post a week ago and a Firefox crash erased it for good; hopefully, I'll recover all the points I had in mind. Crashes always throw off my creative rhythm.)

  • Make it easy to switch difficulty whenever the player wants. This may have been somewhat more difficult last-gen, but not appreciably so, so I'm not prepared to give Buffy a pass for this. I'm several levels through this game, and I've decided that the difficulty level is distracting from my enjoyment of the game. I came in looking for some basking in the Buffy-sphere, and picked the "Normal" difficulty, thinking that I'd take it easy on myself, as I used to play games like this on "Hard".¹ However, here I am, maybe a third of the way through the game, perhaps half, and I'd like to dial it back and coast awhile, probably to the end, get a little extra Buffy fix. But changing the difficulty in this case means... starting over. Wow. What. Were. They. Thinking. This is rule #1. This one can't be broken.

  • Name your difficulty settings well; describe the user experience for each. We have enough space on the screen to say, "Use this setting if you are unaccustomed to first-person shooters; you can always make it more difficult!" or "You will die. Many times. Most of them unpleasantly. Regardless of your experience level." It's okay to say Easy, Medium/Normal, or Hard... but we have to know what that means to the designer. I thought "Normal" for Buffy meant, "Normal for the sort of person who would watch Buffy" but apparently it actually meant, "Normal for a game designer, who has played more hours of games this week than you do all year."

    Note: there may be a temptation to name this stuff from your fiction, but there's a fine line there. If Buffy named its Hard mode "Slayer", I'd want to play just because, hey, I want to be the Slayer. Isn't that why I'm playing this game? Mainstream players may not understand that you're being cute, and may be turned off when you call your easy level of difficulty "Puppy mode".

  • Adjust to the player. I'm not talking some extreme form of dynamic difficulty adjustment, that fabled Shangri-La of difficulty design which somehow magically keeps the player in the sweet spot of perfect level of challenge (and which we will never reach). Sucker Punch did an amazing job with this in Sly Cooper; I don't recall it making a return in the sequels, but it was in the original game and was inspired. After dying a few times on a level, the game would grant you a "lucky silver horseshoe" when you returned; this would prevent your death, returning you to full health once over the course of the level. If you died several more times, it'd give you a gold, which was worth two deaths. It was a simple little crutch, accommodated different levels of ability and the fact that the developers may have been unable to judge the difficulty of their levels. I recommended a variant of this to my friend years ago, and that's what they implemented.

  • Make it clear what dials the difficulty knob turns. This is one we sort of failed on my LucasArts projects; we had a very clear idea of what difficulty was going to be, but ultimately we didn't communicate it to the player. It's a few years back now, but what I recall is that we simply applied a multiplier to the damage enemies did to the player. The thinking was that players would get the same experience, they'd just survive longer and thereby be able to defeat more enemies.

  • Provide the player with more knobs. It's great to say Easy, Medium, and Hard, but it's even better to allow the player to adjust certain aspects of the game themselves. Perhaps a gamer wants harder puzzles but simpler combat or vice versa. If your game supports jumping puzzles, feel free to give the player a knob saying, "OK, you can jump a little further." The best example of this I can recall is System Shock 2, which gave three axes of difficulty via its configuration files.²

  • Do not hide the things that make the game easier. Buffy hides secrets in each level, and tells you on the pause screen how many there are to be found. Unfortunately, in almost all cases, these are things that make the game easier -- health potions that you carry in your inventory, and health and power crystals you give to Willow to power you up in between levels. This is insane. Not only is the game difficult, but I have to seek all over your levels (risking more spawning vampires) to find the things that'll make my life easier? Legend of Zelda has been hiding hearts in stray clumps of grass for years -- don't be stingy! Your mainstream players want to get through the game and feel a sense of accomplishment. Be big-hearted and let them.

  • Test your difficulty settings on real people. Years ago I was playing Dark Forces 2: Jedi Knight, which LucasArts had released shortly before I started working there. I was playing on the hardest difficulty setting and got to the mid-game level where Kyle Katarn has to escape a falling Star Destroyer, or whatever, within a certain time limit. I tried again, and again, and again. Finally I asked a designer friend of mine, and he said that the way they set the time for that level was to take the fastest tester's time to complete it... and to subtract ten seconds. I could have played that for days and not beaten that time. Finally, I just asked for the cheat code to move to the next level and moved on. Particularly when we develop for the mainstream, we are not our audience and we do not share our audience's goals. This is true for me with Buffy: I'm looking for more Zander-Cordelia banter and Willowisms... not another ten nailbiting vampire combats.

    This is true not just of combat. I knew someone who struggled with getting out of the Black Mesa lab -- because it didn't occur to him to break the glass on the elevator door with the crowbar he was carrying.

  • Give players time to get used to new tools before you throw a challenge at them that demands those tools. Buffy has thrown several new kicks and spins and other combat moves at me to absorb into my arsenal of moves. However, because I can only really use these in combat (since they use up a resource that I can't otherwise recharge), I'm kind of stuck. I'd like to be able to practice these before I have to use them in combat, but I don't have any option to do so. Zelda games have historically done this well also -- big challenges appear after you've gotten a new ability, but usually you have an opportunity to use that ability in a safer, less threatening environment, typically in level navigation.

  • Make suggestions. We have the tools to fight player fatigue. If a player spends a long time in an area, we can detect that and give them hints. That can even be one of the knobs, "give me hints when it looks like I'm lost." I know that Perfect Dark Zero got some flak for this particular decision, but honestly, I think it was a good one. Hardcore players should be able to turn it off, and it should never be a crutch to avoid careful level design... but it should be used as a crutch for players who are easily disoriented in virtual spaces.

  • Your easiest setting should basically be "push button, win game". You will think that it can't be made easier, that there are no wall missions. You will be wrong. Make it easier.Give them an out.

I'm sure there are more, and almost certainly I had another one or two in mind last week, but I'm getting tired and thinking of finishing a movie before hitting the sack. I'll add to this if anything from last week occurs to me again, and I encourage comments to throw out ideas I might have missed or forgotten.

Difficulty often breeds frustration, particularly in the narrative-plus-action games that licenses lend themselves too. Give your players a break... and they'll come back.

¹Sad but true, I'm also getting older, but it's not a lack of finger dexterity that gets me in the end, it's the lack of time to play on a more regular basis. I got very close to the end of Metroid Prime 2 some years back and then got quite busy with work. I've never gone back, because attempting to play once your skills start to fall away is no fun at all. (back)

²Normally I'd say putting it in the config files was bogus, but it was definitely a hobbyist game, and it was on the PC, where config files were practically the latest and greatest tech. :) (back)

Posted by Brett Douville at February 9, 2009 11:33 PM


The trouble I have with exposing more difficulty controls is that it may risk turning it into more of a toy than a game. There's something appealing about there being a bar and I get myself over that bar. The more temptation there is to tinker with that bar, the less I feel compelled to achieve. But that might just be me.

Posted by: Fran at February 10, 2009 10:40 AM

That's a good point. I'm also not the sort of person that would probably approach a game that way either. Having the knob doesn't mean that you have to turn it.

In some games, these knobs are there and you're not allowed to turn them. In the game to which I made a difficulty suggestion, it doesn't even tell you it's happening, you just suddenly find yourself getting over the hurdle, and you don't even notice the difference, and you convince yourself you did it. And frankly, you did :)

In cases like these, I probably wouldn't turn the knobs -- I played System Shock 2 on the hardest difficulty level and didn't turn any of those knobs down. But that makes us different from a lot of gamers (i.e. we are looking for the challenge, not just the experience) and the broader audience we have lately been trying to reach, as the cost of game development goes up. My own father is playing games and on occasion finding them too frustrating to continue -- that's both a lost customer and a lost opportunity for us as an industry. Sure, my dad'll keep trying games because I keep sending him ones I'm done with, but many new players might try the same Wii game I've sent him and put down certain types of games forever.

One thing I've remembered that I wanted to mention, and I may amend the post later, is that we shouldn't treat "Easy" as "You suck", either.

In any case, I appreciate the comment, and I know that these sorts of controls can actually work to the detriment of part of our audience. That's why I would tend to limit it to the big mainstream games; you and I can always go play Ninja Gaiden :)

Posted by: Brett Douville at February 10, 2009 02:26 PM

Not to be too flip about it, but I played the new PoP, finished it, and the lack of death never bothered me. Not once.

It's a flawed game in several areas, and there are plenty of things that I would recommend changing, but I don't believe the death mechanic (or lack thereof) is one of its failures.

PoP is a collecting and exploring game first and foremost, no matter how much effort they claim to have put into combat. As such they've built a beautiful and intriguing world that challenges you to unlock its secrets in their totality. It is also, by virtue of the fact that the world appears to have been designed by someone who never once picked up an OSHA standards manual, a very treacherous place indeed. The only way you can fully explore this particular place is to take many, many chances, experiments that will often fail if you are ambitious enough. If you had to die every time, or even if you had to dip into some fixed-reservoir roll-back mechanic (i.e., the dagger in the first PS2 PoP), it would end up being extremely frustrating.

It worked. For me, at least.

Posted by: Chris C. at February 14, 2009 03:40 PM

Hey man, thanks for dropping by. Love to the ladies in your life...

I'm definitely not trying to be down on the new PoP. I can't track down the link at the moment, but that video I linked to showed a game that looked like I'd really enjoy it, and I'm thinking I'll give it a go when eventually I get a 360 or PS3. It's certainly possible that "failure" in this context (with the addition of collectibles? I didn't know about those) is really along a different axis than character death, and allowing you a free pass to take chances with exploration and what-not seems very in keeping with the design in that case. I haven't played it, and wasn't trying to focus on that aspect too much.

It sounds as if they may have made a good stab at giving you the chance of failure, without punishing you too much for that failure, which is a really good approach to difficulty in its own right. I'll definitely give it a look; I've been afraid to return to the series, especially when it got all "gritty" with the second and third titles last-gen. It seemed like the magic of the first one was one of those moments in game development that is just stunningly difficult to repeat or build upon.

Posted by: Brett Douville at February 14, 2009 08:47 PM

Post a comment

Remember Me?