Indeed, it's one of my favorite kinds of roadblocks. Essentially, "get good" rather than "you need to do this, then this, then this". Of course, they can only be used in places where the designer is okay with the player getting through it earlier than expected. Very hard to use a beefgate to force the player to see some important part of the story before proceeding, right? This places it more in line with a somewhat immersive roadblock (it's only a 'game over' in the most technical sense, not in any substantive sense)... and that's good by me!! Sounds like a great solution to have a bit of a sense of danger, without forcing the player to lose an hour of progress. In my opinion (which should be taken with the understanding that I'm one of the most anti-roadblock people out there), yes it is bad. Everything inside me would scream, "I'm a Level 84 Hero! Why can't I grab the cat and fling it behind me with such force that PETA will show up in my next random encounter? (Or simply walk forward and make it run away?)" It would add insult to injury when I do the damned four-part sidequest to 'earn' cat food so I can bait the cat out and get into the casino, and then I walk into the casino and everyone is just sitting there playing casino games happily, like they had no problems getting in. And I couldn't even ask for the manager and kvetch at him! (I sound like a lovely person right now, don't I?) I dunno. In an extremely silly game, I guess it would be okay. But in a game that takes itself even the slightest bit seriously, I'd prefer something like @Tai_MT's proposition of a "Members Only" restriction for an exclusive casino, and/or placing the casino somewhere where you might have to do a sidequest or learn a skill to get near it but once you do you can get in... ...Or the (for some reason rare and apparently blasphemous, if most JRPGs are to be taken as gospel) approach of simply letting the player into the casino early on, with no restrictions or roadblocks whatsoever, so they can enjoy it as a side activity all game long! Where there's no compelling reason to stop the player from doing something, I tend to like letting the player do it. === Not at all! Your perspective as a player is important, and very much worth taking seriously, even though (like mine) it's not representative of all perspectives. Signposting is definitely the biggest issue here. You shouldn't have to get halfway into a challenge before you know you're not supposed to be taking it on yet. Graphical signs, changes in music or screen tone, character dialogue (especially), really tough (and different-looking) monster upfront, a change in the location name, help text that spells out the player is taking on additional challenge... all of this helps and (unless you go for the help text) you probably need to do more than one of these things to make it clear. As far as rewards for the challenge, generally you want to reward the player with a single element of either power or variety that they otherwise wouldn't be able to obtain until considerably later in the game. That usually feels satisfying. Getting a spell that deals 3x the damage of what you've been working with so far (even if it also costs 5x as much) will make the player feel a rush of power for a while, and eventually they'll learn spells on par with it (in the midgame) so it's only gamebreaking for a little while. Alternately, you can open up something like a character or a special cutscene that is only available if you take the challenge on before a certain point - this is extremely satisfying as a reward, but it also opens you up to dissatisfaction from players who didn't take it on as a "missable". I'm not sure they need to be scaled. When the power of most things scales up throughout the game, all you really need to do is place a later-game reward at the end of your balanced-for-later-game (but available in early-game) challenges. These are two cardinal sins! Anything that is handled via "mechanic" (rather than used as, for example, a bias of the RNG to present random story-hook events that may come your way) should be clearly spelled out for the player. Even if it doesn't tell you "there's a 0.01% chance each step a Grue will eat you", the Cursed status should say something like "Stay away from dark areas - unknown and deadly dangers lurk there, and they want YOU". Sounds like it! I think you make a really great point about how audiences from different eras will read something like this!! Obviously, when designing games today, it's the most important to keep the expectations of modern audiences in mind (as anyone playing a new game you make has probably played a few other games in the last decade). But when looking back at games to understand why they did what they did, the historical perspective that you mention is a great tool to have front and center in your mind. One interesting point to consider - if the player is literally told (or understands) "You will die if you go any further without a light source", then why not just have the artificial roadblock? What player in their right mind would intentionally press forward knowing they are about to take an automatic Game Over? I'd argue there are players that would still want to do this (mostly the "Spade" types of Bartle's Taxonomy) just to see how it happens. So it's not a bad idea to signpost it in very obvious terms, and then still let the player be self-destructive if they wish. Total agreement that the genre and especially the "purpose" of a game should be considered when evaluating whether certain mechanics are necessary or just annoying. I personally think that being able to stumble around and solve puzzles in darkness (even on your first run) would be really cool. Not only as an optional extra challenge, but also as an additional way forward if you found you couldn't solve the light puzzle. You asked something to the effect of "why would I do something self-destructive like bumble in the darkness", but that second reason (additional way forward) is a compelling one, especially in an age before GameFAQs. It may be the only way you get to keep playing the game. Now, as far as your second run, where you've already seen this exact same puzzle before (and completed it and died much later on and had to restart because it's a roguelike), I would argue there is absolutely no engagement for me or for most other players in re-doing a puzzle I already figured out. It becomes busywork. A mandatory chore that falls well into the 'Apathy' or 'Boredom' sections of the Flow Diagram. Giving the player who has already done it a "shortcut" (solve it in darkness) would actually be a great mechanic for a roguelike, in my opinion! As a quick tangent - in my game timeblazer I do force the player to replay the entire (five-minute) dungeon every time they fail a boss battle, and that's a very intentional choice intended to break up too much battle time in a row. I feel that this works only because everything you do in the dungeon has a skill element to it. Dungeons are made up entirely of minigames - either action-based ones, which have an obvious skill factor and replayability associated, or perception/puzzle-based ones, which are randomized each time so that it's still interesting and challenging even after you've solved it once. (Worth noting is that it's impossible to "get stuck" - you have 60 seconds to solve any puzzle, and if you fail you still get to move on.) If you're designing a roguelike and you want to force players to solve puzzles that they've already solved on previous runs, randomizing the setup of those puzzles is one way to keep it interesting and challenging in each run. Absolutely! The expectations that the game has created in you, through its previous gameplay, is one of the best signposts of all! I still feel that other signposts are necessary where you're using a new trick or one that runs counter to most things that the player has done so far, to avoid the kind of feeling you mentioned earlier with Dark Souls' "newbie traps". (FF3 does repeat the instant game over thing a couple of times, but not enough to make it an expectation - and it felt especially unfair the first time I encountered it, in the swamp.) Total and wholehearted agreement!! I don't like obviously-contrived roadblocks. The more natural a roadblock is (not only in its environment, but in its role in the plot), and the more "negotiable" it is (allowing a skilled player to find ways through or around it), the more I like it. The one thing I'll disagree on is your question of "why not just place this second path through the forest later on in the game?" - I think that a lot of players enjoy having this optional challenge available (especially if the rewards are good). I'm one of them! It's a rush, it's a visceral sense of risk and danger, and often represents one of my 'high moments' in a great game. I tend to get very immersed in games. I forget about the real world around me. I forget about certain contrivances, like Saves, if there aren't any save points to remind me. I generally remember about once an hour. (If I'm getting game overs a lot, I become more cognizant and try to save more, but I find doing to excruciatingly boring and immersion-breaking.) Now, I live in South Florida (lightning capital of the world), so yes, sometimes power outages screw me. It sucks, but that's force majeure. It doesn't hurt as much to retread ground because of a lightning strike, as it does to retread it because of an instant-death blindside. I can laugh about a lot of things, but I genuinely feel frustrated when I have to do this. Whoa, that's surprising, in light of what you said before about "as the dev, consider whether you need the roadblock at all". I certainly don't like these either, but I think the more common form of roadblock is that the dev wants to make sure that you experience all their "hard work" doing a menial sidequest that, ultimately, wasn't really necessary to advance the plot.