Experienced Points: Difficulty is Hard

By Shamus Posted Friday Nov 19, 2010

Filed under: Column 102 comments

My latest column looks at difficulty in games by examining the various and wildly divergent entries in the Prince of Persia series.

Also: This weekend is Desert Bus for Hope. Never heard of it? I forgive you. Just read the FAQ.

I’ll be appearing (call in) on the show on Sunday at 4pm EST. We’ll do some sort of interview-type thing. I will try very hard not to ask any members of the LRR cast to marry me, but all bets are off if Paul answers the phone. Just warning you.


From The Archives:

102 thoughts on “Experienced Points: Difficulty is Hard

  1. Amnestic says:

    I got Darksiders last week and the difficulty curve for that game was fucked up. It flitted sporadically between “This enemy has cheap attacks which are nigh-impossible to dodge and can’t be blocked” to “If you look at this guy, he’ll die.” The last boss was probably the worst, being perhaps the easiest bosses in the entire game and barely knocking off one of my *ten* life bars. This was on Normal difficulty as well. I’m still sort of balking at it all considering the game proudly touts that it got a 10/10 score on the front cover.

    It’s things like that – among many other things – that makes me wonder if games actually get ‘difficulty tested’ as well as bugtested before release. You’d think having your curve be more fractal would raise a few eyebrows with testers :/

    1. guy says:

      They get difficulty tested about as much as they get bugtested, except without patches to difficulty.

    2. Jarenth says:

      This doesn’t actually much on the Apocalyptic difficulty, I’m afraid.

      Meanwhile, trying to keep up with the Black Hammer is still frustrating beyond help.

  2. Moridin says:

    “If you can pull off the individual jumps with 95% accuracy and if a level has (say) thirty do-or-die jumps, then your chances of making it through the level are an abyssmal 16%. (This is based on a simulation I just ran in PHP. There might be a little numeric slop in there because this is working off of a short run of pseudo-random numbers, but this is still more analysis than some designers seem to put into their game. In any case, you get the idea.)”

    Simulation? You had a perfectly straightforward probability calculation and you tried to solve it by using a simulation? For the record, the probability of bypassing 30 challenges in a row with 95% success rate for each individual challenge is 21%, not 16%. You must suck at math. Either that or you’re too used to solving everything in convoluted ways, as programming often seems to work.

    1. Joe says:

      Wow… there are more constructive ways to reply to something but I guess the internet has its claws inside of you at this point.

    2. Someone says:

      Well, I did it by the formula and the result, indeed, was ~0.21

      I guess not everyone knows, or cares to remember, probability calculation formulae.

      I myself screwed the math up 3 times before I got to that value (taking 0.95 to the power of 31 and 32 respectively).

    3. Shamus says:

      “Simulation” was hyperbole. You know, for humor.

      I may suck at math but you suck in the more general sense of being an arrogant jerk.

      Thank you and piss off.

      1. Jon Ericson says:

        Humor sometimes doesn’t always work over the internet. Nor does sarcasm. Nor hyperbole for that matter since you can always find someone crazy enough to try anything.

        For the record, I got your original joke, but calculated the correct answer anyway. Of course it doesn’t matter since the point is testing ought to reveal this sort of problem. These sorts of levels drive me nuts. They’re almost as bad as timed levels that only give a second or two of slop.

      2. Steve C says:

        Actually if you did a simulation and got 16%, it is correct. (.94 typos aside). Arguably it’s more correct than 21%.

        21.46% is the pure math answer. But a computer doesn’t use pure math. It uses pseudo-random numbers to simulate probability… often poorly. So if the player was subject to a “random” chance in a game, it makes more sense to model it via a simulation rather than math.

        1. Sekundaari says:

          But the 0.95 (0.94?) wasn’t the upper limit for “success” for a computer-generated pseudorandom number, it was the probability of the gamer succeeding in a given jump based on his skills (muscle memory, timing, etc). Or at least that’s how I read it.

          The gamer’s success in tests of his skills might behave pseudorandomly, of course, but then he’s got scary issues…

    4. Nathon says:

      Wow, vitriol. Read the thing you quoted. The “short run of pseudo-random numbers” bit means small sample size. You must suck at reading. It’s probable that his PHP code was perfectly fine but skewed by a large margin of error. The simulation takes about as long to code up as looking up the probability calculation.

      1. Svick says:

        The point is, such code is completely unnecessary, because the calculation is so easy and what’s more, you can figure it out by yourself, if you’ve ever done any probability calculations. (And I would assume the owner of site called “Twenty Sided” has.)

        When faced with programming (or similar) problem, it’s usually best to first think about it, rather than to code the first thing you can think of.

        Of course, this is all assuming Shamus’ goal was to compute the probability, not write an interesting column.

        1. Brandon says:

          It’s entirely possible that Shamus simply couldn’t think of what the probability calculation was off the top of his head (I know I wouldn’t be able to) and decided it would be a fun exercise to write a program to do it in a simulation.

          I doubt it would have taken him long to code it up, and he is allowed to spend his free time doing anything he wants. Regardless of whether or not writing the code was unnecessary, no one really has any reason to be an asshole about him doing it that way.

          If you absolutely had to say something, it might be more appropriate to say “Hey Shamus, I just did a quick probability calculation using the formula, and got a different number. You may want to look into that.” or something to that effect.

          1. Zak McKracken says:

            Just for the record:
            If you must succeed n times and have a probability of p of managing it one time, then the chance of making it is
            P = p^n
            so 21.46% (.2146) for p=.95
            15.62% (.1562) for p=0.94

            reason: you’ll scale the first obstacle with 0.95 probability, then out of the .95 times you managed that, you’ll make the second with a .95 chance, so .95 times .95 (or 95% of 95%) which is .95^2 and so on …

            To add something to the argument: Here’s the rest of the curve with increasing ability:
            p : P
            96% : 30%
            97% : 40%
            98% : 55%
            99% : 74%
            100%: 100%

            So it’s obvious: a small change in the player’s ability (although of course it’s more difficult to go from 98% to 99% than from 50% to 51%, but still) will cause a large difference in the advance he will be able to make in the game. Therefore, if the game designer guessed just a bit wrong, the consequences for the player are dire. There seems to be no easy solution, since the player _expects_ some challenge, and most also like to see the exaggerated effect of improving your skills (i.e. first time you barely made level 1, a week later you’ve mastered the whole game!)
            I wouldn’t fault anyone for not knowing that by heart, but post it here, so next time Shamus can save some time. His point is still valid, nonetheless.

            1. Zak McKracken says:

              gaah, my edit messed the post up … the last paragraph was meant to come behind the statistics explanation, not at the end :(

        2. Veloxyll says:

          The important point is that simulations sound cool while probability makes ones eyes bleed :)

        3. Newbie says:

          Probability to an A-Level standard in England considers this formula to be too difficult for students… I dispair…

    5. Adam says:

      Wow, you’re a jerk!

      I mean, I had the same question. Why use a simulation? The math is trivial. I’m sure a lot of other readers had the same thought.
      But I’ve got basic reasoning abilities, which tell me that if he knows enough to write a probability simulation for it, he probably knows the formula to calculate it. So I typed the equation into google, got the right answer, saw it was close enough that it didn’t disturb his main point, and moved on with life.
      But let’s say he didn’t know. Let’s say there was some blinding deficiency in his education, and basic probability is something he’s not capable of.
      You still don’t need to be a jerk.
      “But he’s WRONG!” you’re no doubt saying to yourself. In polite company, it’s considered rude to call out your host on mistakes he makes, especially with a comment about how dumb he is.
      Did you see what happen when you wrote your post? Everyone’s angry at you. Nobody’s impressed. There was never a chance anybody would be impressed.

      Take this as a life lesson, and know better for next time, I guess.

    6. albval says:

      I don’t know which makes me more sad: one person making an insulting comment or all these other people insulting him back. Internet and emotions just don’t mix – and calling someone a jerk does not make him less a jerk.

      1. acronix says:

        I think it´s quite poetic: he receives what he gave!

        1. Zak McKracken says:

          That kind of poetry reminds me of kindergarten and the old “he hit me fihiiirst!” game.

          1. Jarenth says:

            I’m rubber, you’re glue.

            Bounce off me and stick to you!

      2. Kaeltik says:

        Shamus has done a difficult thing here: he has created a community. If the community is an organism, this is its immune response. It appears TwentySiders are allergic to unthinking incivility. allergic reactions aren’t pleasant, and can even e harmful, but they serve a purpose.

  3. Andy_Panthro says:

    “It’s not harder, it’s just more time-consuming”

    This sums up my feelings about Oblivion’s combat (and to a lesser extent, Fallout 3). But I think enough has been said about that.

    I was half-expecting a comment about COD:BLOPS and the video which has been making the rounds… that you can do the cuba mission while only firing twice.

    I found it interesting that you mentioned tutorials on Spoiler Warning, and it was also a focus of Extra Credits. In their case, they mention “good tutorials” being the key, whilst for more experienced gamers tutorials of any sort are often seen as a hindrance no matter how well presented.

    I’d be all for separating tutorials from your main game, but then you run into issues of people ignoring the tutorial (just like people ignore manuals, something that would be unheard of in the days of Ultima!).

    Unskippable tutorials do bring us hideous additions like the “temple of trials”, and I was going to counter that with a good one but I can’t think of one.

    Extra Credits did mention Portal, which presents you with each new skill one level at a time, but neglected to mention how often used that is (although perhaps not quite in the same way), alternate examples might be Thief with it’s presenting you with new equipment or Command & Conquer giving you new units or even most FPS games ramping up the firepower. They’re all drip-feeding you new tools to overcome new challenges. Basically, I’m not sure you can compare this to a standard tutorial at all (and some of these games have tutorial levels anyway…).

    1. Aldowyn says:

      One interesting thing to note is how lots of flash games work, especially platformers. They have the tutorial at the beginning, with super easy levels (use a and d to walk to the door) and quickly ramping up the difficulty and adding gameplay elements.

    2. Will says:

      I’ll be honest; if you skip the tutorial and then discover you can’t play the game because you have no idea what you’re doing. Well, who’s fault is that exactly? You can’t blame the designer if the players are being stupid.

      1. Amnestic says:

        Include a “Tutorial” button in the menu screen. Problem solved. True, it might break the flow a bit, but being stuck on something you can’t work out how to beat/get past breaks the flow anyway.

  4. Debaser says:

    Gotta get up at 8’o’bloody’clock in the morning to hear you get interviewed. You’d better have some sparkling dialogue!

  5. Bit says:

    The escapist freelancers have a weird habit of accidentally copying each other. You and The Needles used to frequently write about the same things, and Extra Credits was about this topic this week. Huh.

    1. thebigJ_A says:

      Dude. Go back and read the very paragraph sentence of the column again. Or for the first time, since I don’t think you bothered to read it at all.

  6. Zukhramm says:

    I’m not even sure I want to look at the Escapist comments for this one, difficulty just seems to be a topic always stiring up the rage.

    I do agree that when creating a challanging game, just increasing the punishment is not a good way to go, and that a game can be challanging without being punishing. Not passing the challange can in itself be enough, and making the player redo try not only this specific challange but the three previous ones too can be annoying or downright frustrating.

    There are however two things, postivie things, that I feel punishment can bring. First thing is having something at stake. If there something I stand to lose, passing a challange will be more exciting. The second is calming the player down, having to pass two small pits and just one goomba before trying to make that tricky jump again gives some time to relax between trying the challange.

    These two seem to be not go well together. If the part replayed is easy the only thing at stake is wasted time, and if they are hard there’s not any calm time between trying the challange.

    1. Newbie says:

      The best difficulty curve is y= lnx x>1

      The difficulty starts low, then goes up quick to impliment the new mechanics you will need, this brings you into the game and then the difficulty carries on rising at a pace that becomes less and less but never becomes lets do that same mission again…

  7. Nathon says:

    Because I’m a huge nerd:

    from random import random
    def attempt(chance):
        return True if random() < chance else False
    def run(chance, times):
        return not False in (attempt(chance) for i in xrange(times))
    SAMPLES = 1000 * 1000.
    test = (1 if run(.95, 30) else 0 for time in xrange(SAMPLES))
    print sum(test) / SAMPLES

    gives me 0.21448999999999999. So I see more like 21% success with Python's prng and a million samples. (in 8.5 seconds!) Not too shabby, given that the actual chance is about 21.464%.

    1. Shamus says:

      Not sure where I went wrong. I actually tried a bunch of different numbers while writing the column. The point was to show that a high success rate could be destroyed by stingy checkpoints. 99% didn’t yield interesting enough results (you still had a good chance of finishing the level) while 90% ran the risk of being dismissed. “I can do way better than 90%!” Maybe 16% was the result I got for .94. I can’t remember now.

      1. Zukhramm says:

        But why do it with random numbers at all?

        1. Shamus says:

          Because it was amusing to do a “simulation” in PHP? It’s a bit like someone saying he’d solved the Kennedy assassination by building a replica of the area in legos.

          And yeah, .94^30 is 16%. So I used the result for 94% instead of 95%. Simple editing mistake.

          1. Zukhramm says:

            The difference is that I know building things in lego is fun, while I have never done anything in PHP, which is why I’d ask why for on of them but not the other.

            I just never thought that solving mathematical problems through programing could be done for fun, not because it could not be fun but because every time I have done it has been very much not so.

            1. Shamus says:

              Sorry for the phrasing that made it sound like I thought you should have gotten it. The question mark was supposed to imply “I tried, you know?” Not, “DUH!”. Yeah. The joke was perhaps a little self-indulgent. I mean, I was amused at the idea when I was doing it. I suppose it wouldn’t have been a problem if not for the 94% error.

              1. Caffiene says:

                The funnier part is the number of readers, myself included, who didnt get the joke because running a simulation just for the sake of it seemed like a perfectly reasonable thing to do. :D

                1. Pickly says:

                  Or like me, who though his “simulation” was just putting the numbers into a calculator. (Which is kind of a “simulation” in the “simplified model” sense, though partially I was also in the flow of the article and didn’t think about it too much.)

      2. Someone says:

        Yeah, 0.94 returns ~0.16

      3. Moridin up there was needlessly rude, but he did have a point. This is a very simple mathematical problem and using a simulation with random numbers like this overcomplicates the issue and gets a less accurate result than just doing the maths.

        0.95^30 = 0.21463…

      4. Jarenth says:

        95% à­s often taken as the baseline for scientific significance.

        That could, you know, be useful if you were going to write a research paper about this. Or something.

  8. Aldowyn says:

    Extra Credits is awesome.

    It’s kind of like I would like on my blog, in addition to the specific bits on individual games.

    Anyway, on to the topic. One thing no one seems to have mentioned is difficulty levels, and I just figured I’d add my two cents.

    You need to be really careful, and you need to have enough to make it a challenge for veterans, but still accessible to newcomers, with everything in between. That’s harder than it sounds, due to differences in mechanics, genres, and audiences, but it’s massively important.

  9. Tizzy says:

    The big problem with the original prince of Persia was precisely the lack of a good save feature… that and the bloody ticking clock! So that a game that was by its nature reasonably challenging became an exercise in frustration. It was too addictive not to play, but too frustrating to ever want to replay.

  10. Brandon says:

    I’m glad you talked about this Shamus. Difficulty is a really nasty issue in games, and a lot of my friends (Who are longtime gamers) spend a hell of a lot of their time complaining that games are too easy now.

    On the other hand, I’m the kind of guy who doesn’t mind if a game is a little bit “too easy” as long as it’s fun and enjoyable. I would rather have a game be generous and fun 100% of the time than 90% fun and 10% hair-pullingly frustratingly difficult. If you play a game for 20 hours and 10% of it makes you want to strangle someone, that’s a full 2 hours that your entertainment has spend making you frustrated.

    This is something I don’t see as a problem in other forms of entertainment. People might get bored of a movie and turn it off, or something.. I don’t think anyone has ever rage-quit a movie though, or at least it’s much less frequent than people quitting a game that is too challenging for them.

    1. Mari says:

      I’ve rage-quit a few movies. There are many more where I stayed on to the final, bitter moments and it left me with a deep-seated hatred of everybody involved in the movie from the actors, directors, producers down to the lowly caterers and limo drivers. Actually that’s become part of my review matrix for movie review sites that require numerical reviews. One star is for movies that leave me wishing ill to everybody involved in a film (and occasionally their families as well). There are only a handful of these but enough to make it a criteria for the one-star rating. Anything else automatically gets bumped up to 2 stars just for not making me hate it so much that I want to see the original prints in a studio vault somewhere firebombed during office hours. (For the record, “Nude for Satan” was one of my first 1-star films and sadly the most interesting part of the movie was the title.)

      1. krellen says:

        I totally rage-quit on the movie “Black Sheep” (with David Spade and Chris Farley); I was expecting something much like Tommy Boy and shut it off in disgust after a half hour of completely failing to see it come about.

    2. Jarenth says:

      I would rather have a game be generous and fun 100% of the time than 90% fun and 10% hair-pullingly frustratingly difficult.”

      Your dichotomy seems a bit off, though. Think on this: would you rather have a game be sort-of-fun 100% of the time or really great fun 90% and hair-pullingly frustrating 10%? Because that, in my eyes, is the problem here: making the game ‘too easy’ can avoid the frustration problems, but for plenty of people this will come at the expense of enjoyment.

      1. Brandon says:

        Plenty of people aren’t me. They are free to go play whatever game they like. If they want it to be hair-pullingly difficult, more power to them.

        Don’t be confused. I’m not saying I don’t want the game to challenge me at ALL. Just that if it is so difficult I want to punch something in frustration (Or pull my hair, which is something I often actually do) then I am not enjoying it. You’re right, part of the fun of games is overcoming obstacles. I just prefer that the obstacles actually be reasonably overcome with a few tries.

        At the point where I have replayed the same 5 minutes of a game for 3 hours, I generally turn the game off and never come back to it. There may be people who love that kind of thing, but not me.

        1. Jarenth says:

          Yes, ok, I see your point. That’s the sort of thing difficulty settings should be able to adapt to, really.

        2. Blake says:

          “At the point where I have replayed the same 5 minutes of a game for 3 hours, I generally turn the game off and never come back to it. There may be people who love that kind of thing, but not me”

          This exactly.

          There’s not many games that do it but when they do I quickly lose interest.
          I enjoy a nice challenge in games but I think the problem arises when the only way to progress is through a super hard section.
          Take Super Mario Galaxy 1 for instance, if you got 3/4 of the way through the game when suddenly the only star you were able to get in order to advance was the 100-Purple-Coins-on-the-disappearing-blocks-in-the-shape-of-Mario star, then you’d have most people stop playing at that point. As there were other ways to advance instead it became a challenge for those of us that wanted it, while letting everyone else still enjoy the game.

          I’m also one of those people who’d rather cruise through a game then have to repeat a level more than twice. I quite enjoyed Bioshock on easy, not because I couldn’t play it on a harder difficulty, but because I’d rather look at the scenery than my ammo and health.

          Shamus once posted on here something along the lines of ‘some gamers will die on a level 3 times and say ‘aaargh that level is so hard! I died three whole times before I beat it!’, while others will say ‘pfft that level was way too easy, I only died three times before I beat it!’.
          I think it perfectly sums up how even for gamers of the exact same skill level, difficulty settings are still extremely important.

          1. Jon Ericson says:

            Super Mario Galaxy has an almost ideal difficulty curve. I cruised through the first 100 or so stars hardly noticing that I was getting better at the game to the point where the early levels are trivial. Not having any true roadblocks for the first half helps. (In other words, each completed level potentially opens up another level which may be completed in any order.) Now there are 14 stars left, but I doubt I’ll finish more than 4 or so of them. Thankfully there is a sequel!

    3. On the other hand, the Wing Commander series had an interesting difficulty sequence, down to the Mission 13 spikes. One of the satisfying things was solving the Mission 13s.

      Rather than waiting for the random factors to solve them for me. E.g. in the original game, every so often the four enemy fighters attacking the ship you are trying to save will line up so that only two are attacking it and two are putting their fire into the ship in front of them.

      On the other hand, when you triggered the action zone, you had the following sequence that generated about a 75% success rate. Issue break and attack command to your wing man (you need to start him earlier than he auto attacks). Change your target to the second one that targeting locks on to.

      Fire missile as you close to gun range, volley/missile, repeat on the next fighter, then head to the other pair, your targeting will pick up the proper target, which you can volley then ram/dumbfire missile head on, and the other people on your side kill the final fighter at the same time, mission won.

      WCII, it was easier. Tell your idiot wingman to attack the escort ship for the station. He will do his suicide run on it instead of the station and take it out. Then you strip the fighter cover. Then you kill the station. Much easier than trying to line yourself up so you are in the shadow of the escort ship so that the station can’t do you as much damage on the runs.

      By WCV, the last of the series, mission 13 was no longer as elegant.

      Anyway, good call and good points.

  11. Samkathran says:

    Yeesh, lots of people getting caught up over a silly little math error…

    Difficulty is definitely one of those aspects of a game that is challenging to design properly. The target audience is critical like you said, but I think in a lot of cases, companies want their target audience to be “everybody” so they can make the most “moneys”, but then it ends up appealing to “nobody”. Plenty of games can be made for everybody, don’t get me wrong, but sometimes there’s going to be a specific group you’re appealing to and you better know what they want.

    I remember when I was taking a class on game design and we were making games using GameMaker. I thought I did a pretty good job making mine, but when it came time for the class to try out everybody’s work, I found out that my second level was mercilessly difficult for anybody who didn’t make the game (go figure!). So everybody immediately lost interest in my game and found ones that weren’t stupidly hard.

    Lesson learned! Get playtesters!

  12. People always get probability and chance mixed up. (me included from time to time).

    If the accuracy is 95% then the probability of failure is 5%, but the chance of failure (or success) is 50%.

    The chance of failure or success is ALWAYS 50%, regardless of the probability.

    However a probability of 95% for success makes failure less likely.

    Illusionist Derren Brown proves this very well here,
    with 10 coin tosses in a row, all heads. The probability of that being possible is 1 in a thousand chance.

    Now that you have seen that, take a gander at the following… jump to exactly 6:30 to see how he actually made the ten in a row coin tosses happen, and there was no tricks involved, but instead as he explains, the limitation of our perception.

    This is not just true for coin tosses, but also true randomness, or chance and nature in general.

    Given enough time, or a specific point in time and there is always a chance of something seemingly improbably happening multiple times in a row.

    Programmers (and game developers, and gamers) seem to forget this. You can not have actual randomness driving the game mechanics, such a game will be a pain to play.

    Instead the game mechanics must be biased (aka cheated) in favor of the player, always.

    Derren Brown and the film crew spent 9 hours shooting and repeating the coin tossing until he finally managed to do ten in a row.

    That is the reality of radomness and chance,
    and what you saw in the first video clip with only the 10 coin tosses is what a player should see with say close to 100% accuracy for example.

    1. Zukhramm says:

      I don’t people get them mixed up as much as, to most people, chance is the same thing as probability.

    2. krellen says:

      Movies do the same thing. It took Michael Cera 33 tries to get the “over the shoulder into the trash can” toss in Scott Pilgrim, but we just see the one time it went right.

      1. Here’s another Derren Brown, just because how hilarious it is.
        Sometimes it isn’t just a limited perception that gets people,
        sometimes it’s the absence of perception:

    3. Tizzy says:

      The chance of failure or success is ALWAYS 50%, regardless of the probability.

      Sample dialogue:
      A – Will you go out on a date with me?
      B – No chance!
      A – Oh, so 50% then…

      I am not sure what you mean by chance, but this does not appear to make much sense. If your chance, whatever you mean by that, is a number which is always equal to 0.5, it’s not exactly a useful measure of anything, unlike the probability, which is a number that does measure something very precisely (the average proportion of successes in a long sequence of independent attempts http://en.wikipedia.org/wiki/Binomial_distribution ).

      So whatever real point you may have here, I don’t think using language like The chance of failure or success is ALWAYS 50% is really helping. As Zukhramm points out, chance and probability are used interchangeably by many, and frankly, I think this use of language is absolutely fine.

    4. Newbie says:

      So the chance is 1 in different options available, such as Yes no or maybe the chance is 1 in 3? That seems odd…

      The probability is a much more useful method and infact the whole crux of the problem with statistics is the language allows you to lie under the pretence you are doing maths that means something.

      Chance is a way of decieving people by them thinking the probability
      Infact the probibility (I put chance the first time) of getting heads on a coin is greater than 50% but the chance will be 50%… so again lies…

  13. Mari says:

    Funny you should bring up difficulty. I am currently playing a game which shall remain nameless to spare me some ridicule from people who had a different experience with the game and I have hit the difficulty Mt. Everest. It is not a curve. Well, it was a curve. Until it suddenly became a vertical line. I am 100% certain that many, many other people have managed the challenges I face and moved well beyond them but it doesn’t fix the fact that I’m “playing” a game where I’m unable to progress nor figure out how to get “better enough” to return to progressing. It’s intensely frustrating to me to the point that I haven’t played in a week. At this point I have about a hundred hours invested in the game and it depresses me to think that it might be all for naught because I can’t figure out how to move past this obstacle.

    1. TSED says:

      When this happens to me, I just start googling or youtubing the game in question (particularly Let’s Plays). Often there’s some little trick and you smack your forehead and move on, happily.

      For example, that just happened to me. I’m playing through Soul Calibur 4 (way behind the curve, I know) and got to the Tower of Lost Souls floor 45-46. These things were brutalizing me SO BADLY because they had incredibly huge damage and defense bonuses and were completely immune to ring outs and grapples. I was able to get to floor 46 but I simply couldn’t fight them to death, no matter how hard I min-maxed.

      And then I saw someone on YouTube fight them. They’re immune to grapples, but not to ATTACK grabs. And then I won first try by spamming Nightmare’s down-A+G. I was so very, very relieved.

      Anyway, difficulty? Pffft, Dwarf Fortress.

      1. Someone says:

        Dwarf Fortress isn’t actually all that difficult, as long as you don’t handicap yourself it’s almost as easy and tranquil as Sims.

        The difficult part is getting into it and learning to operate the game’s tools, and if you keep a wiki handy that won’t be a problem either.

        1. Will says:

          Older versions aren’t difficult, 2010 is a goddamn nightmare.

          1. Someone says:

            Why? It’s no longer possible to end up with no water or soil, the only difficulty is getting rid of forgotten beasts, and most of the time you might as well just ignore them.

            1. Mari says:

              See? This is why I refuse to name the game I’m stuck in. Difficulty is so subjective. For those of us stuck somewhere obviously it is difficult. Otherwise we wouldn’t be stuck. But for those who got past it, clearly it wasn’t. The thing that irks me to no end is to be on one side of the curve and get ridiculed by people on the other side for not being as competent as they are. Sorry. Trust me, if there were a switch somewhere inside me to make what seems absurdly simple to YOU just as easy for me, trust me, I’d flick it. In the meantime, thanks for not making a frustrating experience even more so by mocking my difficulty. KWIM?

              1. Someone says:


                I was talking about Dwarf Fortress and how everyone keeps banging on about how difficult the game is. Will up there said that 2010 version was more difficult than the previous ones and I asked him why, because I think it’s a lot easier and want to know what’s his take on it.

                If you thought I was being smug or something – sorry, didn’t mean to.

                1. Decius says:

                  DF is, has, and will be either trivially easy or literally impossible. Until victory conditions or performance metrics are established, nothing you do can change your performance.

                  DF players establish their own performance metrics, therefore they effect the difficulty.

      2. Mari says:

        I’ve looked it up. 90% of the “hints” I’m finding don’t apply to me. OK, yeah, it’s an MMO that I’m playing. Almost all of the hints and tips are either for multi-human groups (vs. me playing by myself) or for character classes that I’m not or for character skills that I don’t have and the only way of getting is to get PAST this point. Other than that, apparently luck is a factor which I haven’t had yet. Which means I get to DIAS it until I get lucky or until I wheedle some poor sucker (my husband) into playing with me in hopes that two human intelligences can succeed where one fails.

        1. Jarenth says:

          Is it EVE Online? Because the difficulty curve of EVE Online is notoriously messed-up, to the point where wrong choices taken during the first few weeks — you know, that time where you’re still trying to figure out what’s what — can cripple your character progress pretty much forever.

          On the upside, you get to shoot lasers at space-rocks.

          EDIT: And no, I’m not really expecting an answer; this was just an exuse to link a funny picture and complain about EVE.

          1. Mari says:

            LOL That is, in fact, a funny picture. And that notoriety has kept me from even considering the slim possibility of playing EVE. I would be the five hundred millionth corpse on that pile. And bald. Which is not a good look on me.

          2. Andrew says:

            I remember trying the trial for EVE. And yeah, that graph about sums up my impressions. The game is amazingly open if you can get into it (or so I’ve been told), but I couldn’t even begin to learn the ropes. Also, it irked me to no end that spaceships had a top speed. Seriously, what’s up with that?

  14. ccesarano says:

    It is always a relief to read this blog and know someone else that does not hate Prince of Persia 2008, but saw new potential. Maybe for me, it is because PoP’08 was my first in the franchise ever and thus had no basis to go off of.

    I think what makes PoP different is that I always found it engaging and relaxing. Having a video game that is relaxing to play is almost unthinkable, but with few exceptions I could sit down, turn on PoP’08 and just relax. It requires just enough thought to keep me entertained, but not so much that I can’t just let myself ease into my comfy chair after a rough day of work/classes/what have you.

    Seeing how poorly people reacted, I’m certain this sort of “relaxing engagement” design concept won’t become more frequent. Which is a shame, to be honest.

  15. Kale says:

    Nuts, I’m taking down a stage at that time tomorrow. I’ll miss the phone call.

  16. silver says:

    Maybe I’m just dumb and naive, but it seems to me like there’s easy solutions to some of your major problems that have been technically possible for decades:

    1) save anywhere. then it’s your fault if you have to repeat some piss-easy challenges before the hard jump. and you’ll know where to save next time. any console game that lets you push a button to get to a menu can have a save option in that menu.
    2) well-placed auto saves. helps mitigate problems with #1.
    3) skip-able death animations

    That won’t save you from marketing the wrong difficulty to the wrong people, but it will solve about a zillion complaints unrelated to that.

    1. Veloxyll says:

      It would. My DS can do on the fly saves, there’s no reason that modern consoles can’t. It saves a lot of frustration when you can do your saves yourself. Autosaves are still cool to have for those times when you forget but getting them perfect is less of an issue when you can save anywhere.

      Skippable deaths are good too for traditional games.

    2. Rosseloh says:

      Quicksaves are my favorite thing in CRPGs. I practically abuse them (and have to periodically clean out my save folders when they get too big, if the developer didn’t have it overwrite)!

      Too bad for developers who haven’t realized that the lack of quicksaving, or at-will saving, doesn’t actually increase difficulty.

      Looking forward to the “interview”, Shamus!

      1. acronix says:

        I´m quite sure they aren´t trying to ramp up difficulty, but game time. If everytime you die you have to repeat 20 minutes of content before trying again, then the game suddenly can become a 10 hours game instead of a 6 hours one, if you have enough of those battles.

        1. Rosseloh says:

          And when I realize that is when I set it down. Unfortunately, by that point they usually have my money already.

          Fortunately we aren’t completely lacking in competent storytelling. It’s just not in the “mainstream”.

  17. This-is-Hip-Hop says:

    Im really excited to hear your interview on the Dessert Bus play, im planning on buying the shirt to show my support, not to sound stingy, but I always like to get something for anything.

    1. FatPope says:

      Mmm dessert bus…

  18. Meredith says:

    This is a good topic and I always like the reasonable approach you take with it.

    I just don’t understand people who think a game isn’t fun if they don’t die x number of times per level. That’s not fun, it’s hellish frustration and a waste of time! I don’t mind a failure I can learn from, but anything that’s up to random chance or having to jump from just the right pixel can go die in a fire.

    I’m with the people above who would rather have it be slightly easier and completely fun than a massive challenge and not much fun at all. Of course, too easy can get boring, but I find most games err on the side of too hard or they ramp up the difficulty too fast.

    More games should have selectable difficulties and quicksaves are a must, which is why I’m glad I play on the PC and not a console.

    1. “I just don't understand people who think a game isn't fun if they don't die x number of times per level.”

      While I obviously can’t speak for everyone who enjoys hard games, for me and a lot of other people it’s not so much that you need to fail a certain number of times but that I want it to feel like the game is punishing mistakes and rewarding a good performance. As far as I’m concerned, it should be possible to win any level on your first playthrough without prior knowledge, provided you’re good enough at the game, and any trap that requires you to fail once before you know what to do is a major design flaw (I Wanna Be The Guy gets a pass here because it was basically using extreme unfairness as slapstick comedy). I also completely agree with you about random chance, which needs to have its influence minimised as much as possible. On the other hand, if I waltz into a fight without bothering to keep my guard up then the enemies should take full advantage of this, if I misjudge a jump I should fall into the hole, and if I perfectly pull off a sequence of moves then the result should be noticeably better than if I had performed merely adequately.

      As an example of a hard game that I thought worked really well, I’ve recently been playing Super Meat Boy. On the harder levels you might die dozens of times, but there is never any point where you have to fail. Every death involves being hit by a hazard that was clearly visible in advance, and it is always obvious where you went wrong and what you need to do in order to be less rubbish on the next attempt. And then when you do pull it off, you win, with no chance of some random event beyond your control screwing you over.

      Conversely in Resident Evil 5 I had far fewer deaths, but it was frustrating because there never seemed to be any clear link between playing well and winning the level, since success depends too heavily on hoping that your pet lemming Sheva will avoid doing something unbelievably stupid. Ultimately I had the same basic problem with it as Prince of Persia 2008 in that two radically different performances could have the same result – success in PoP and failure in Resi.

      1. Jarenth says:

        I second this comment.

        For people who enjoy the challenge (and I occasionally count myself among them) the prevailing thought is really ‘if there’s no risk of failure, what’s the reward of victory?’. The feeling of challenge and the idea that you managed to overcome that challenge on your own wits and skills in the main rewards in these cases — and if that’s not what you’re in the game for, it won’t work on you.

        For instance, I play games like Mass Effect and the earlier mentioned Darksiders on the highest possible difficulty setting, because in those games, the challenge is the point of playing. God knows it’s not the story in Darksiders anyway.

        Conversely, in Minecraft Singleplayer, I will almost always turn monsters off because they don’t represent a challenge in that game, they represent a distraction. My aim in Minecraft is to build, not to get random-exploded every night.

  19. thebigJ_A says:

    The game that got its difficulty perfectly right for its intended audience was Demon’s Souls. (It’s tutorial was great, too, as was mentioned in the referenced video.) The game was massively challenging, but perfectly fair. It had very strict rules, but crucially, it never ever broke them. You just had to learn to play the game on the game’s own terms.

    It was also a risk/reward thing. Having a single currency (souls) for everything, from buying weapons to leveling, and then putting that currency at risk when you died was pure genius. If you died, you’d lose all those souls, but you had a chance to get them all back if you could make it to the spot where you fell and pick them up without dying again. Here’s the important bit, though. You could return to the Nexus (hub) at any point and spend souls. You had to decide whether to press on and earn more souls for spending, risking what you had, or playing safe, but having less to spend. If you lose a ton of souls, it’s your own fault.

    God I loved that game. Too bad I borrowed someone’s PS3 to play it, and they aren’t around anymore. And I’m too poor to go buying a console for one game.

  20. thebigJ_A says:

    When you edit a post, the text jumps around in time with the seconds ticking. It makes it very awkward.

    1. Jarenth says:

      Comment +1.

  21. Daemian Lucifer says:

    When it comes to challenge,I wanna be the guy is the perfectly designed game.

    As for the oldies that were fair,how about from another world?Not as much pointless repetition as in prince of persia,but still a very difficult platformer puzzle.Or lost vikings,though that one could get quite frustrating in the longer levels.

  22. rofltehcat says:

    So… am I the only one who enjoyed 2008 Prince of Persia?
    I really liked it, it felt much different. Most jumps weren’t any difficult and they removed most of the punishment for failing (well, you lose a little time). Yet it was still fun.
    I also liked its graphics style, the music was great and storywise I enjoyed it over the sands of time series, maybe because it was about a livelier theme, don’t know.
    Well, I only played Sands of Time on the pc and Warrior within on the psp but I got the other games lying around here somewhere… I liked Sands of Time but judging from Warrior Within, I’d prefer playing a sequel to 2008 Princeo of Persia over another Sands of Time title.

  23. Zaghadka says:

    The folks over at WiiBrew just listed “Desert Bus for Wii.” It won’t let you turn it off without hitting the reset button. The Bus is everywhere.

  24. pkt-zer0 says:

    I thought this article was pretty lacking. I feel you could’ve made a lot more interesting points here. Difficulty is hard, yes, but that’s why I think it would’ve been worthwhile to go more in-depth on the subject.

    You talk about choosing the right difficulty for your audience, and cite the Prince of Persia reboot as a mistargeted casual game. But plenty of games let the player choose the difficulty setting, and scaling down to an auto-win difficulty (catering to the casual audience) is usually rather trivial. You say repetition and time-wasting is bad, but that’s not even necessarily related to difficulty, just poor/lazy game design. A game being too easy games could also be considered a waste of time for that reason.

    Can you design a game for the hardest of hardcore that can be scaled down to a variety of skill levels while remaining enjoyable? Are there design choices which simply don’t scale, but are worth including anyway? What are some clever designs that can scale up the challenge, and not just the punishment? These are the sorts of questions I’d liked to have seen tackled.

    Here are my answers, for the sake of discussion:

    1) Starcraft 2 (or competitive multiplayer focused games in general) shows that gameplay balance tends to remain consistent throughout various skill levels. So adjusting challenge is simply a matter of matching people up with an equally skilled opponent (human or AI). How well does this translate to purely single-player games, though? I’m not sure.

    2) The Void has you exploring an alien world and trying to survive in it. Your actions can have significant long-term consequences, but since the rules governing the world are also initially unknown to you, it’s possible to screw yourself over unintentionally. Yet writing this off simply as unnecessarily punishing would be a mistake, because these are the mechanics that fit the game’s theme. A dissonance between the theme (you’re in a hostile alien world…) and the mechanics (…but you can get away with anything regardless) would cheapen the experience.

    3) The Tales of Destiny remake would be my example here, but first for something people are more likely to be familiar with: Ground Control. It’s a fairly standard real-time tactical combat game, but the brilliant part is that turning up the difficulty increases damage dealt across the board. That means that it’s easier to take out enemies on a harder difficulty setting with an ambush or a well-placed artillery shot; and vice versa, getting caught in such situations practically guarantees failure.
    Tales of Destiny on the PS2 has its combat mechanics based around a bunch of elemental attributes, with different weapons/armor/spells/skills having different elemental affinities. By using the correct element, you can score a critical hit which stuns enemies and deals more damage. Enemies can score a critical hit in a similar fashion. So, what the higher difficulty settings do is turn up the critical chance and damage multiplier, making them way more deadly, as well as increasing enemy health, making criticals a more important factor to victory.
    In both of these examples, the end result is that the harder the difficulty, the smaller the spectrum of tactical options that will work. On hard, you have to choose the optimal approach to situations, on easy, you get can away with sub-optimal choices just fine.

    tl;dr version: article not in-depth enough, here’s a bunch of related stuff to discuss.

    1. Will says:

      StarCraft and other multiplayer focused game are exactly the wrong examples for ‘hardcore’ difficulty. Any competent player can trash most AI’s, although the higher level AI’s in SC2 cheat their asses off and can be quite difficult to dispatch, they still get flattened by serious players however.

      But that’s intentional; if the AI’s don’t provide you with enough challenge, hop online and find a person who does, thus bypassing the difficulty system altogether.

      1. Amnestic says:

        In all fairness to SC2’s AI, they seem to cheat a lot less than most other RTS games I’ve played. I mean it’s not perfect by any means, but it’s a step in the right direction.

      2. pkt-zer0 says:

        SC2’s AI actually no longer cheats, unless you consider superhuman reaction times a cheat. That’s if the AI programmer is to be believed, anyway.

        Anyway, the point was more that the game mechanics themselves were designed around high-level play, with crazy 300 APM Koreans, and the game still remains fun if it’s just you and a bunch of pals 2v2-ing it up in the silver leagues.

      3. Newbie says:

        Funny that you said this… I read an article recently that talked about the AI olympics or whatever it was called where now AI’s are tested on their competance in StarCraft 2 rather than chess. Now if you are having trouble go challenge those AI’s I bet you can’t beat them…

  25. I’ve been doing some thought on the topic of video game difficulty as well (though I haven’t seen the video you’re responding to yet).

    Basically, I came to the realization that dying and getting sent to a loading screen to take you back to the last checkpoint is an instant immersion breaker. To take Modern Warfare 2 as an example (as it’s the last game I beat) – there’s a level in the game after the Russian invasion of North America. The level starts out in a series of tunnels underneath the Capitol (though this hasn’t been revealed yet). You hear fighting overhead, and you’re currently more-or-less alone. You don’t have your usual location text in the lower left hand corner. You don’t have your squad there to tell you what you need to do. You’re just there.

    You navigate through the tunnels and can see from what’s going on around you that things are going pear shaped – a makeshift infirmary, a soldier receiving being resuscitated on the floor, and another soldier just sitting in the corner huddled with his rifle. Finally, you emerge onto the surface and into the trench complex and you see that you’re on Capitol Hill, and it’s become a major battleground. At this point there’s very little interface on screen, and I’m totally immersed.

    The thing is, the very moment I die, get a loading screen, no matter how brief, and then have to go through all that again, immersion will be broken and it’ll be a while until I get it back.

    I kind of came to the conclusion that difficulty should take a back seat to immersion, and it’s better to give me the impression that I’m going to bite it, then have me end up dying in game, run into a loading screen, and have to do it over and over again until I get it just right.

  26. Alex says:

    “The 2008 Prince of Persia…

    …Long time fans bought the game and wondered why their beloved well-established platformer of flowing movement and elegant controls had just become Super Win Button Deluxe.”


    That game kicked my butt halfway through. Its’ Simon Says quicktime events became far too long and complicated for me to memorize, and the controls for certain actions always seem to fail when they weren’t supposed to(like when you’re hanging above a steep drop on the side of a cliff about to jump to the adjacent hanging wall, and instead the game thinks you wanted to jump forward into oblivion, even though you didn’t give it that order).

    It became impenetrable for me. I had to give up. In 18 years of playing games, that has almost never happened to me.

    This is the first time I’ve heard anyone say that game wasn’t difficult enough. Are people actually saying this? Is this a real complaint people had? What MLG Cyborgs are disappointed it wasn’t the right amount of unforgiving? The only way I could reasonably call that game easy would be in comparison to Bullet Hell games.

    1. thebigJ_A says:

      In honesty, I haven’t played it, but yes, it is generally considered to have been far, far too easy. Not by some subset of uber-gamers, but by the general gaming community.

      1. thebigJ_A says:

        Maybe you are thinking of the wrong game…? I hope….

    2. Jarenth says:

      That’s an interesting point: people have been lambasting Prince of Persia 2008 because you can’t fail (as in, you get infinite re-tries of everything and there is no associated penalty for failing), but that doesn’t mean you can’t fail (as in, not be able to pass a certain obstacle, no matter how many times you’re allowed to fling yourself at it).

  27. Andrew says:

    Dear Escapist,

    When you make your ‘background’ a link to advertising, you make it extremely unlikely I’ll ever return to your site, even for content from an author I enjoy. Even when that ad was for charity.

    This is not an acceptable web practice.

    Article note: 0.95^30 = ~21% of success. — EDIT: And now I see the vitriol that math sparked… well, math and jerks. Skimmed the comments and didn’t see it. Searched and found.

Thanks for joining the discussion. Be nice, don't post angry, and enjoy yourself. This is supposed to be fun. Your email address will not be published. Required fields are marked*

You can enclose spoilers in <strike> tags like so:
<strike>Darth Vader is Luke's father!</strike>

You can make things italics like this:
Can you imagine having Darth Vader as your <i>father</i>?

You can make things bold like this:
I'm <b>very</b> glad Darth Vader isn't my father.

You can make links like this:
I'm reading about <a href="http://en.wikipedia.org/wiki/Darth_Vader">Darth Vader</a> on Wikipedia!

You can quote someone like this:
Darth Vader said <blockquote>Luke, I am your father.</blockquote>

Leave a Reply

Your email address will not be published. Required fields are marked *