Oooh, difficulty in games as a general discussion.
Lessee...first off, I don't know what's "good" or "bad" for that. I do know what, by watching people, is psychologically fulfilling, though, in general. Which might be the same thing? I don't know, people are atypical. Let's look at this, then.
For instance, is "both hard AND tedious" bad design?
Yes and no. It depends on the tedium. People don't mind losing a game, especially if they don't lose very much, although it's objectively always an addition of time. Why? Well, there's a risk/reward to it. You are playing the game, there is a real risk(time loss) to dying.
It's kinda like how there's risks to a lot of sports that give people a kind of...added thrill, on top of your difficulty.
Now, where this line is drawn person by person-how much time, how much risk they like-varies, plain and simple. There are people that freeclimb buildings, which is really quite dangerous, and surely has more of a thrill as such. But most people don't want decent chances to die, a chance of a broken limb is often more than enough thrill for them.
However, unforseeable tedium, actually just wasting time, either to gain enough stats or items or whatever, is mostly only fulfilling if there's another way around it and you're taking that option as a last ditch. You've still won, which is better than losing, but it's cost you a lot of time.
Now, granted, I've seen a very good analysis on how people also just find pride in achievements of large amounts of time(like being proud of your MMO character). I've...never quite gotten that out of grinding-I actually grind because I am easily amused and because I want to see more abilities, more options, more *stuff* accessable, rather than just to smash enemies. But I can surely see the draw.
Also, there's tedium in the sense of things being long, and hence giving you less time to screw up. This is boring, but legitimate, challenge. Is it enjoyable challenge? I...well, too short and a boss fight in a game can easily be worthless. Too long and you're wasting time. A person has a point where they've already proven they've got the dodges down, more or less, and you're just trying to make them screw up.
Is there an optimal length?
One single optimal amount of time lost per failure at a challenge?
Valid question. The answer is...no, but because it matters how many times you're likely to
lose trying something as well as
how long it takes to lose. And of course, it matters how much information you're willing to garner from other people on the challenge.
But for the amount of time lost as a general function? It varies by the person, again, some people have a very low tolerance for it. I think about 30-40 hours lost with nothing to show for it is about the upper limit. A person
might, say, restart a game they lost their near-endgame save for right off the bat, but they are going to be royally pissed and it will hurt the game to them. Ditto if challenge does it.
Now, if you've got a smaller outcome based system, such as how DMC3 lets you just do single missions at a time, that's different. You'll likely have some stuff to show for 30-40 hours of deaths there if you're, say, trying to S rank all DMD missions. Maybe everything won't be S, but some of it will be if you're able to handle DMD at all. That's psychologically much more fulfilling(and shows why people will go and cap every single card in a Touhou game for instance, despite this being hands down the most time consuming non-challenge-run thing you can do.).
What role does probability play in all of this?
A good but psychologically unsatisfying one, in general, but that's because it's rather hard to do well, put in short.
Generally speaking, randomization-within reason-makes either strategy or reaction time skills take precedence over raw knowledge. That is to say, if you know a boss' pattern in FF6, and it's fixed, you can win at X level every time with Y strategy. No originality or skill required, just add GFAQs, as it were.
Randomized patterns require you to react to certain situations, and while a strategy can be wrote out concerning every possible variant of a randomized pattern, people rarely do it, and only really when required due to a challenge being
so hard.
So it's obviously really valuable annnnnd people are fine with it so long as randomization is a fairly major part of the game. Touhou is a good example here again, it relies on randomization in EoSD's case rather heavily, for instance. Randomized bullets don't bother people too much. However, there are limits. People don't mind randomized patterns of bullets.
But they do mind moving boss, upon moving, taking it into it's head to move to the upper right corner of the screen and stay there. Which causes you to struggle to hurt it at all and forces you to judge if you want to stay in your position and dodge until the boss decides to move-if ever-or if you want to try to go after it, which is usually quite dangerous to do sideways through a pattern. There's also patterns I've heard simply called luck garbage(and called a few that myself) thanks to the fact that no matter how skilled you are, they can throw up a wall sometimes, which you couldn't have foreseen.
Randomization is basically like this in general. No one really minds randomized damage spreads, but criticals are a wall you can't foresee, except in a general sense, and struggle to deal with(Except by extreme measures, much like you can just bomb attacks that turn into walls of shots. Lacking an extreme measure is bad design. FE, for instance, really lets you just go out of your way to avoid criticals, or use people that fail at being one-shotted by them.). No one really minds enemies changing up their patterns, but sometimes they'll just make a goal unachiveable or unreasonably hard to achieve and it's quite irritating.
You really need good limits on the ranges of randomization in order to make it work.
cheese out
I cannot emphasize enough how good of an idea it is to have game mechanisms that allow you to do this. Difficulty should be on a nice sliding scale, and optimally it should show(ranking type mechanisms are good! They make you feel sucky for using cheese! Although it can be taken too far, Megaman Zero games added in a major game feature and then had using it at all splatter your rankings for all time, permanently. They actually stuck with this for two games. I don't get it.).
If you want to make grinding a requirement for your easier way through, or whatnot, go for it, people pay their time for their easier game.
Quick Time Even related deaths
Good example of a completely patterned event with little or no randomization. You just screwed up execution of mashing a displayed button. So yes, it's not very fulfilling, as you not only know what you have to do, the only real issue is if the game catches you off guard. Too simple for what it is.
...it's also a much worse execution of a rythym game, but that's neither here nor there.
