Intro
Hi, I'm developing Dwell, a game where you explore a giant world, build villages, overthrow other villages, and survive the wilderness. It is a top down 2D ORPG (servers can hold upwards of 1000 I believe). The game is permadeath. You have 3 lives, but after the last one you lose everything.
The situation
I've been trying to find a way to add permadeath without the player feeling annoyed or defeated at the end. Ultimately, I would like the player to be disappointed, but motivated enough to continue. I personally like a RotMG approach (a leaderboard with stats, account-wide stats not tied to your character, but nothing affecting gameplay). My artist suggested adding a Path of Exile death system where your character is teleported to a different server (aka league). I don't really think that sort of system could work in a game like this.
The question
How should I add meaningful death in a permadeath game so that the player is still motivated to continue?
I find the question is a tad worrying. It almost seems like you're pursuing the idea of adding permadeath for the sake of it, rather than considering what is that permadeath would bring into the game.
Permadeath is not a mechanic you can expect to be able to pour into any design, ending up with a better result. Permadeath needs to support the rest of the experience, otherwise it might as well end up ruining the experience as a whole.
As an example of that, take the genre which most prominently uses permadeath to its advantage: Roguelikes.
In (most) roguelikes, the core of the game is a combination of the following: tactical combat (most often than not turn based, rather than real time), decision making, resource management and exploration; Also multiple, possibly very elaborate systems are thrown in for good measure, in order to make the game space significantly larger and allowing for emergent gameplay.
Permadeath in RLs is actually supporting (and enhancing) all the other mechanics. It makes the design (and most importantly the gameplay experience) better as a whole.
The same goes with games like Spelunky, RotMG, The Binding of Isaac, Rogue Legacy, etc. etc. (games often referred as Roguelike-likes, and for good reason).
For instance, permadeath in all of those games doesn't really feel like an overly severe punishment for failure mainly because, even though you're required to start again from the very beginning) the game is throwing at you a wholly different experience every time you start a new game. As long as you're learning something new, or you're making some progress in therms of player's skill, dying doesn't feel like an utterly meaningless defeat.
This is
very important. And in fact permadeath in all of those games is ultimately making the learnining process all the more important. Coincidentally, the fact most of threats the player is going to face are procedurally generated also means the player can't simply "beat" the game by memorizing map layouts, enemy locations and solutions to puzzles.
So, the real question I feel you should ask yourself would be: "Would permadeath add to Dwell's experience as a whole?".
You mentioned, in Dwell, the player explores a big world.
If the player is to permanently die, would he have to re-experience the exact same content from the very beginning? This could easily lead to player frustration (notice how most action/adventure games which feature linear/static content opt for a save/checkpoint system instead).
Is the player expected to learn a whole set of skills (meaning player's skills, like in knowledge about the game's system, or maybe execution skill)? If that's not the case, and starting over is merely a matter of "grinding your way" through already experienced content in order to being able to face whatever killed you the last time and take another attempt at beating the same challenge... you get my drift.
So yeah, I felt like it was important pointing out the very basics about why pemadeath could be a good idea for your game or not. This is only the tip of the iceberg, and you can certainly find a whole lot of useful (and more in depth) info about he subject by using google.
Assuming permadeath is a good option for Dwell, it's important you implement a system of sorts which is keeping progress acquired via previous games from being lost.
Most traditional RLs (think Rogue, Nethack, Angband, etc.) can get away without this mainly because the player is usually acquiring new knowledge with each following game. So all the progress is out-of game player's knowledge, rather than anything like an in-game reward.
Spelunky goes a similar route. Beating the game is hard, especially if you're a new player, but you're acquiring new knowledge and your platforming/whip-swinging skills are getting better (even if marginally) with each game. Furthermore, the game is full of secrets (maybe the first time you beat the game you don't know what the purpose of the Udjet Eye is. And possibly even when you've beaten the game multiple times you've never been to the Worm level, or Hell).
The Binding of Isaac uses an interesting approach. As you play more and more, new items are unlocked and can from a certain point onwards randomly spawn in the game's dungeons. You're unlocking some new creazy item with basically every playthrough, so the player feels rewarded even though he technically lost. Hundreds of unlockables are in the game, so this system also heavily taps into collector-type players' psicology.
In Rogue Legacy, the gold you acquire in a game can be spent to upgrade the shop (these are permanent upgrades) and buy upgrades for your next character (these you lose upon dieing), so you can effectively re-invest resources gained over the course of a given game. A very important caveat of this is, you can't store earned gold in a bank: any amount of gold you don't spend is lost as soon as you start a new adventure. The important consequence of this is you can't simply hoard gold in order to build yourself an overpowered character.
RotMG is kind of a unique beast. The game doesn't feature an ultimate goal. Unlocking classes and completing class quests is a short term incentive, diminishing the severity of death. But even then, there's a slight problem with having to pot your character up again every time you die (unless you manage to hoard lots of potions and storing those in your vault/mules). Furthermore, the game kind of gets repetitive soon enough, despite the use of randomly generated dungeons. You always face the same combination of enemies in every dungeon, and the enemies/bosses always use the same patterns.
It's complicated. Leaderboards are only relatively effective in giving RotMG players an incentive to play the game again after a death, if you ask me. My experience with the game is, you generally want to play more just in spite of the fact that the game is fun and very addictive way (and I don't necessarily mean it negatively way). Other than that, the main goal for a long term player is hoarding more and more riches, growing a powerful pet or becoming a celebrity among the game's playerbase (all of these usually end up playing into each other). But I can assure you a huge percentage of new players end up leaving the game out of frustration in little time (both the game's difficulty and the community can be very harsh to newer players).
As a rule of thumb, leaderboards or other ways for the player to track his progress over the course of several playthrough is a nice feature. Those let the player withness for himself if he's getting any better at the game by exposing some kind of metrics.
If many different classes are in Dwell, you might require the player to hit a given mileston for him to unlock all of them. This gives a short term incentive with trying new classes out and, as long as reaching said milestones is easy enough, makes the player feel rewarded despite losing.
If exploration is a huge part of Dwell, you might add an entry in an hypothetical player's log/diary upon reaching specific landmarks. This would be persistent over many playthroughs, so the player might be willing to explore a different part of the world after creating a new character.
You mentioned players in Dwell build villages. Maybe players retain ownership upon whatever they built?
Ultimately, the "right" way to make death less severe of an outcome is going to be entirely dependent on how exactly Dwell works. I can only point you out in a general direction, but only you can work out the finer details.
Hopefully this long-assed post was helpful.
Developer
