PDA

View Full Version : Question about Roguelike Design



BlackShipsFillt
25-09-2010, 10:54 PM
Okay... I've never designed a roguelike before. While I have played and studied a bunch it'd be hubris for any designer to believe that they can mimic the elements they think they enjoy in a game and achieve the same results. Games are very complicated, and roguelikes are arguably some of the most complicated games.

There is probably a lot I don't know, but something I'm struggling over is permanent death.

The question then is: How integral is permanent death to the design of a roguelike. What affect does it have.

Right now we are playing around with a couple of ideas:

Offer a hardcore/roguelike mode and the option of a casual mode
The casual mode could have Diablo-like respawns, ie. you keep your progress but face a bit of backtracking
Or the casual mode could restart the level, basically checkpoints as there will be several levels in each game.
Or maybe the player keeps their progress but has limited continues/lives
The casual mode could also be termed "Practice Mode" in order to steer more serious players away from it.


With endless continues there may be less sense of challenge, like Diablo's you-are-50000-clicks-from-the-finish gameplay, and given the choice would players choose Casual Mode over Roguelike Mode and thereby ruin the experience for themselves.

I don't believe there are any hard rules to game design. Certain mechanics favour certain temperaments. And I'm not ashamed to say that I'm targeting as broad an audience as possible. Permanent death does not feature in Diablo-likes. Death in Diablo-likes is a nuisance and also an embarrassment, but only a temporary setback, and it is easily provable that a Diablo-like would be more popular on this platform than a roguelike.

But permanent death has a certain psychological effect on players. In a long game (say several hours) permanent death provides a lot of grief (that few audiences have a taste for), however in a short game it not only increases the challenge, but instills in the player the idea that the game is a worthy adversary (as in, the game isn't just difficult, but it can also defeat the player).

The iPhone/iPad audience is maybe not as competitive as the hardcore audience, but there is definitely a demand for replayability and most successful endlessly replayable games on the platform are random sandboxy games that include permanent death. Also most of these do not include a win condition.

The gameplays in PocketRPG most likely will last 15 -30 minutes, so while a permanent death is not a big deal, it is significant. The question really comes down to: Would a short Diablo-like work or is it the worst of both Diablo-likes and Roguelikes, and should there be an option for it?

Daemienwrath
26-09-2010, 12:22 AM
In a roguelike the key is giving the player to play random worlds and the ability to take their time to decide on their next move. So while the story could take 15-30minutes for a well versed player it could take days for another player. Permanent death is really ok for the most part as you say gives that little bit of excitement and increases the importance of staying alive, however this is not a requirement for a roguelike :P If you dont want to frustrate gamers provide a possible mechanism to make a save point, either in specific locations or at specific significant times in the game. Possibly add it into your world you create as a story plot item. You could still leverage the loss of points strategy but make the save points rare enough to still maintain a balance between frustration and character importance. In the end its your game and its about finding the correct mechanic for your narration. Remember its not just about the game mechanics, its about the game holistically. I ran into a similar situation when writing a MUD a while ago and i personally decided to go with the perminant death option as i felt it went well with the setting of the game. That being a World of Darkness style rpg.

Good luck

Miktar
26-09-2010, 12:53 AM
The value of a roguelike is the sum of its random generation systems - which are an integral part of roguelikes, be it randomly generated weapons or randomly laid-out maps - and how brutally honest it is to the player. The more brutal, the more someone who wants their roguelikes to be brutal will like it. The less brutal, the more people you'll attract who are put off by games being "too much work".

I think both areas have room for expansion, the roguelike genre is really underutilized: the question is, which group do you want to play to, or will you try to provide for both (tricky, it's basically double the design workload, but amicable if you attempt it).

"Price, Speed, Quality - Pick Two" - as the old saying goes.

dislekcia
26-09-2010, 02:13 AM
Personally, I think that permadeath is the line between roguelikes and action RPGs. To me, Diabloesque games are action RPGs and things like SAIS are roguelikes because they contain permadeath.

The thing you need to be looking at is player value: As Mik said, a roguelike's random systems are important - I reckon that this is because they're what a player derives value from. Because permadeath de-emphasises characters as vehicles of attachment, players get attached to what happens in the world, they come away armed with stories of how cool a particular session was or how strange something was that completely blindsided them. In games that have really rich random systems, permadeath can even be a value-add (beyond mere difficulty) by creating valuable stories for players to tell. Nandrew loves explaining how he starved to death in Crawl once because he was levitating and couldn't reach the food under him on the floor. People have run to find me to tell me how they were doing great against teleporting squids in Spacehack until the-I-Totally-Swear-It-Was-the Millennium Falcon pwnt them from behind.

Contrast this with what players get value from in action RPGs: Neat character builds (that could only exist in a setting where the sorts of hazards a character is going to face are known), items and - ironically - corpse retrieval runs. You don't tell people about Snot Snot the Snot and how he killed your thorns pally 4 times in a row - that's just not compelling because everyone's run into something named that while playing Diablo 2 at least once. An action RPG without character development decision-making is a tricky thing to pull off. Permadeath is thus anathema to that form of player-derived value.

But there's something to be careful of here. As much as permadeath is another chance for players to take a dip into the random and find a truly unique setting/world/event/thing, if the game is capable of creating situations that are patently unfair, they will be upset by this. We get that in DD all the time: People feel very strongly about the dungeons that are dead-end runs because they're hemmed in by higher level monsters from the word go. Because the system is so unforgiving, (but winnable!) people feel really ganked when there appears to have been nothing they could have tried differently - as long as there were options they hadn't explored yet, they'd have another stab at it.

So you need to decide: Are you building a complex set of interacting systems that create a fun play experience for players, or are you sculpting a setting in which they can get attached to characters in. That'll answer your permadeath question.

dislekcia
26-09-2010, 02:21 AM
In a roguelike the key is giving the player to play random worlds and the ability to take their time to decide on their next move. So while the story could take 15-30minutes for a well versed player it could take days for another player. Permanent death is really ok for the most part as you say gives that little bit of excitement and increases the importance of staying alive, however this is not a requirement for a roguelike :P

Completely disagree with this. If it doesn't have permadeath, it's not a roguelike. Plus the idea of a story is almost alien to roguelikes in general: You don't want your players to be able to experience the entire world in one play session, why would they come back then? Roguelikes give players vague goals (Get the amulet of yendor) and players invent their own stories or find little tidbits buried in events, action RPGs give players stories (Deckhard Cain was totally in league with Tyrael to dupe you into destroying the worldstone) as sequences of known goals (Collect X rat spleens to make the potion to cure the witch to curse the monkey to...).

Miktar
26-09-2010, 04:48 AM
Play enough Spelunky and ADOM, and you'll get what roguelikes are. :P

Fengol
26-09-2010, 06:43 AM
There's also the option of carrying "some" attributes between permadeath; like Dead Rising's health meter and Desktop Dungeon's gold.


Because permadeath de-emphasises characters as vehicles of attachment, players get attached to what happens in the world
I've never thought of it like this before; it's absolutely brilliant! I haven't been a fan of rogue-like permadeath but I think it's because I've been "trained" by other RPGs to be thoughtful and attached to my character and I'm actively involved in his growth and morale choices.

That said, a way to disenfranchise the player from the character is to make the character creation screen weak (i.e. simple button selection and no customization), automating skill leveling (the player can't control the direction of growth) and make the acquisition of skills based on pick-ups rather than granted by level. Desktop Dungeon does a good example of this by having radio buttons for race/class selection; leveling deterministic but undirected by the player and access to spells and power-ups are only available through the level generated pick-ups.

Back to dislekcia's point; the trick is to re-train the player to not be so attached to his character. Another idea is to speedily go from character death through new character creation and back into world. Because of the nature of the mobile device I suspect players will likely put the game down after dying; so quickly getting into a new game must be as easy as loading a saved game.

BlackShipsFillt
26-09-2010, 08:04 AM
That said, a way to disenfranchise the player from the character is to make the character creation screen weak (i.e. simple button selection and no customization), automating skill leveling (the player can't control the direction of growth) and make the acquisition of skills based on pick-ups rather than granted by level.

Yes, this is the route we are going. There are three characters and that is the only choice the player needs to make at the start. There is no stats-screen at any point, the inventory affects all the customisation and the stats simply scale as the character levels up. That part of it is definitely Roguelike.

Dislekcia, I aggree that Roguelikes require permanent death to be true Roguelikes. Without permanent death the game would definitely have to be termed Diablolike. The other thing that also makes a Roguelike for me, like you said Dislekcia, is that each game offers a different experience (through its randomness and complexity).

However this game does not have to be a Roguelike (it certainly won't be a classic Roguelike as it has a third person action hack 'n slash interface and action gameplay). Kind of like Daemienwrath kind of said, I'm trying to figure out what the audience wants, and see if we can adjust the game to be like that without ruining it.

This question wouldn't even come up if it were just up to me. If it were just up to me there would definitely be permanent death, I feel the 50000-clicks-until-completion approach limits replayability, which is fine in a 6 hour game, but not a 20 minute one. While there will be high scores, if the player is seriously trying to get a high score then the player will restart the game when they die anyway. And if the player does not care about high scores and the game does not pose any real challenge then once they have seen the three characters they will not be inclined to play further.

In any good Roguelike (like Desktop Dungeons, Spelunky, Crawler, SAIS etc), or even any good replayable action/shooter/platform game (like Crimsonland, Geometry Wars, Tilt-to-live, DoodleJump, Canabalt or Super Meat Boy etc), I have ended up playing it over and over again precisely because after I died I wanted to get further on the next try.

But there is a lot of pull towards Diablolike at Tasty Poison (and from the iPhone audience). The idea of permanent death seems to be unfamiliar to hardcore RPG players. There are very few success stories of any sort of Diablolike Action Roguelike hybrid that I can point to and say : Look at that, see, it sells!

My fear is that making the game always winnable, which makes it a short, random Diablolike, reduces the players urge to replay it. The audience really craves a Diablolike, certainly, but I fear that a short random Diablolike may just end up weak.

Fengol
26-09-2010, 09:13 AM
Why does seem that mobile = casual and casual = easy?

I suspect that mobile = short gameplay (I don't have market data to back this up) but I don't agree that mobile games must be accessible and entertaining to the 40-something white male who, for the past 15 years, has worn slacks, shirt and tie. Sure they're the ones most likely to have the latest hardware (just like Porche is driven almost exclusively by the over 50) but if they're your target market make a good golf sim or performance car racer.

If you start by saying "I want to make an RPG for a mobile device" then you're dedicated to the genre first and will make it accessible to as wide an audience as possible second. With this mission statement you're going for evolving the genre instead of sales.

To back up my claims I always tote Why "Casual" Doesn't Mean "Easy" (http://thegameprodigy.com/why-casual-doesnt-mean-easy/) from The Game Prodigy and now recently Finding Casual by Defining Hardcore (http://www.gamasutra.com/view/feature/6143/evolving_the_social_game_finding_.php?print=1) from Gamasutra.


A game that starts easy is accessible, even if it ends hard.

I think Tasty Poison (http://www.tastypoisongames.com/) should reflect on what their mission statement for the game is which should direct these questions. There are a lot of options and it depends on what you want from the result.

P.S. Just to be fair; the definition of casual gameplay is an ongoing discussion at Luma Arcade (http://lumaarcade.com/) as well.

BlackShipsFillt
26-09-2010, 11:14 AM
Sorry, I wasn't trying to suggest Tasty Poison wanted it Pocket RPG easy, or that their concerns were as a result of a misunderstanding of the casual market.

Imagine if in Diablo when you died the death was permanent and you had to start from the beginning. Okay, so that is obviously a bad idea. So what I meant was that to someone who mostly plays games that last 6 - 40 hours long (like Diablo, Neverwinter's Nights, Mass Effect etc) the concept of permanent death sounds very unappealing. If permanent death does not feature in any of their favourite games why then would they include it in a game they are involved in?

It isn't so much a matter of easy vs hard or casual vs hardcore but of the previous experiences and expectations. If they have little positive experience with the awesome effects of permanent death (that Dislekcia mentioned) then they are very unlikely to favour it. If we were making a longer game we would be going full Diablolike.

There is a large group of gamers on the iPad/iPhone market who are starved for lengthy more serious RPG's. This group of gamers are certainly interested in Pocket RPG and if we get them on board they will be very vocal in their support. But they are interested in Pocket RPG because they play Diablolikes and MMORPGS where permanent death is not a feature. I think the idea of including the ability to continue arose from trying to appease this group (and there is definitely some sense in that). My concern is that rather than pleasing them we displease everyone.

My concern is that we receive a bunch of reviews along the lines of: "I had fun trying each of the 3 characters, but then had no desire to play further". There are also unlockables to keep players playing, but if those are the only reason players keep playing then we've done something wrong.

There is of course another large group who may be very interested in Pocket RPG for its brevity, challenge, replayabilty and to some extent its simplicity (because a lot of RPG's require a lot more management work just to play). This is a much easier group to cater to only because their expectations are lower and we don't have the budget to make a Torchlight. This is the group of people who I feel will appreciate the effects of permanent death (even though death itself always hurts a bit).

Of course right now, as the design document stands, permanent death is optional. But I think this is one of those situations where players may not behave rationally and they may choose endless continues even though they end up not enjoying the game as much.

So... I think I'm hearing strong support for permanent death, though I do know that Game.Dev LOVES Roguelikes (as do I). Does Game.Dev feel that endless continues, or the option for endless continues, could ruin a game like this... or is it more a matter of flavour and I am being overly concerned?

P.S. I hadn't seen that Gamasutra article about what is hardcore and what is not before, going to give it a read now. Thanks!

dislekcia
26-09-2010, 06:26 PM
I don't know if you're hearing strong support for permadeath or for good player engagement. I think in the end it comes down to what players are going to be playing the game again for and how much budget you have to extend that:

If players are playing to experience a story, then focus on the story and making the character interactions within the story coherent and compelling, different character types should have different interactions and reveal different parts of the story.

If they're playing to see new areas and marvel at the graphics, then make sure different player characters see areas in at least different ways (if not get unique areas some of the time).

If they're playing to get engaged with the characters and to push the third person action mechanics, then you need to design the characters in such a way that there aren't enough resources in the game to fully unlock everything a particular class can do.

If you're sure players will get a benefit from permadeath, then include it, but make sure that you've got enough novelty in the world to warrant returning multiple times. Design with death in mind... In many ways this is completely against the sort of story that I've heard you talking about for the game so far. Story breaks random generation :(

To be honest, I don't think that Pocket RPG needs permadeath. If you look at the character classes in a roguelike, they're all made using slight variations on what your standard character is going to do - they're not very different when the game starts out, but as people play, they diverge because they're predisposed towards certain types of decision by the slight tweaks to the system's starting state. If you're going to explicitly design characters and class skills, then focus on making exploring those really fun. Permadeath doesn't work in a setting like that because players will always be going over the initial decisions of which tree to go with that they've made already last time.

Miktar
26-09-2010, 06:44 PM
A good rule of game design - make sure that every action the player takes, is rewarded in some way. But when you punish, punish honestly. Permadeath is a very honest puishment system, and it adds value to the experience a player gains from playing - and they keep playing, because their interaction with the game is rewarding, due to good design.

dislekcia
26-09-2010, 09:21 PM
A good rule of game design - make sure that every action the player takes, is rewarded in some way. But when you punish, punish honestly. Permadeath is a very honest puishment system, and it adds value to the experience a player gains from playing - and they keep playing, because their interaction with the game is rewarding, due to good design.

So let's flip that around and ask: How can a player be punished honestly in an RPG without permadeath?

I think that answering that will go a long way to addressing the "50000 clicks from the end" problem that PhillyT is battling with.

Miktar
26-09-2010, 09:56 PM
Ironically, Demon's Souls does a good job of answering that question. It's one that Diablo 1/2 has answered before as well.

Give the player a possession they care about, because it means something to them in the game world (Diablo, it's rare items spat out by the random generator, scarcity = demand, and demand for power-items with low drops, is always alluring - look at the MMO market using that to great effect right now, and in Demon's Souls, it's physical experience points, spent upgrading a character, which you can lose all that you have collected once starting a mission, by dying before you can get back to your dead body to retreive - hence the simularity to Diablo's way of doing it), but then have the threat of temporarily losing it for a length of time be the punishment. In this system, the fact that you actually WILL enact the punishment, and that it DOES happen, doesn't end up feeling much like a punishment at all actually, in the mind of the gamer - because hey, the game totally told you what was going to happen. Which means the game was being honest, and you can respect that.

So, there's that route. I'm curious to see if there are others I hadn't considered yet.

BlackShipsFillt
27-09-2010, 06:37 AM
If players are playing to experience a story, then focus on the story and making the character interactions within the story coherent and compelling, different character types should have different interactions and reveal different parts of the story.

There will be no formal story. We'd like some battles with monsters to be memorable and interesting, but no formal story elements (as in text or speech) that would appear with each play.


If they're playing to see new areas and marvel at the graphics, then make sure different player characters see areas in at least different ways (if not get unique areas some of the time).

There aren't going to be big graphical changes, like new areas, brought about by playing repeatedly. There will be a unlockable monsters and unlockable weapons but the game will not have as strong a sense of progressing (or exploration) that Diablo or Torchlight does. Each of the three characters will play quite differently, but they will all receive the same settings. There will be randomness in the design of the settings... but all the settings will be seen within each game. I'm not exactly certain what you are suggesting, custom areas are a lot of work. Hopefully there will be some rarer areas, but that isn't a given at this point.


If they're playing to get engaged with the characters and to push the third person action mechanics, then you need to design the characters in such a way that there aren't enough resources in the game to fully unlock everything a particular class can do.

Yes, there will be more weapons, spells and enemies than a player will see in one session. Hopefully some of these items will work together nicely (and allow some cunning teching).

@Miktar. I think you're right about the really rare really powerful items, I think we had been considering having most of the items fairly balanced, but some extreme ones may really help the drama (like getting the jetpack in Spelunky). The items won't be randomly generated like most Diablo items, they are all explicitly created and the exotic ones will all be unlocked through repeat plays.


If you're sure players will get a benefit from permadeath, then include it, but make sure that you've got enough novelty in the world to warrant returning multiple times.

I think that designing the game with enough novelty to warrant multiple plays is a requirement even without permadeath. This game is going to be short with highscores, so it has to be replayable, I'm not trying to decide whether the game has to be varied and as new and stimulating as possible with each play, I'm trying to figure out if endless continues reduce engagement or replayability.


Design with death in mind... In many ways this is completely against the sort of story that I've heard you talking about for the game so far. Story breaks random generation :(

The Pocket RPG story is: You start, you kills stuff, finds loot and buys items, you beats a boss (or die somewhere along the way) game ends (and maybe you unlock stuff). I'm a little confused, is that the story I've been talking about?


To be honest, I don't think that Pocket RPG needs permadeath.

I agree, it might not. If it has the things you've mentioned, enough novelty and events that are interesting (like Rodain's story about starving, although that is more awesome narrative-wise than anything Pocket RPG will provide. I've even quoted that story) it can work without permadeath. Also unlockables will give it a sense of progress and make things more interesting the longer the player plays (hopefully). Short repeatable plays may work even if every single one ends in a win, so long as there is a reward for greater successes (like a score or achievements).

I would like cool memorable randomly generated encounters, but like you mentioned Dislekcia, without permadeath every play ends in winning and the in between encounters won't be quite as memorable. There will certainly be few stories about death, every story the player tells will be "I beat the boss with this and this item".

Another point about permadeath is that the fear of dying increases adrenaline which then stimulates memory retention (which could be part of why I have such vivid memories of Spelunky but not Torchlight). Some people like to be excited while playing games and I think at least some of our audience may be made up of such people.

@Dislekcia. That was also a good point about graphical changes, I'm not certain if I understood it right, but getting to a new area in a Diablolike feels like awesome progress. Right now we had been thinking about each game consisting of 3 settings that appear in random orders... But maybe this is wrong. Spelunky might be a good example here. In Spelunky the levels always played in the same order, so the end levels were always the least familiar (and trickiest).

In Pocket RPG we want a much higher success rate than in Spelunky (assuming permadeath). It really adds to the drama/informal story that the harder levels always look like the hardest and are the most mysterious. But of course this cannot be achieved without permadeath.


If you're going to explicitly design characters and class skills, then focus on making exploring those really fun. Permadeath doesn't work in a setting like that because players will always be going over the initial decisions of which tree to go with that they've made already last time.

Yes, agreed, there isn't going to be a skill tree in Pocket RPG, nor any perks or stats screen. All customizations of classes are based off of the items that randomly drop (like Fengol suggested). The classes are explicitly created though, and hopefully we can make each play within a class quite different (through differing random items with each play that significantly affect playstyle).


So let's flip that around and ask: How can a player be punished honestly in an RPG without permadeath?

Didn't Vampire Masquerade have an interesting death solution, that the player had to fight in a netherworld briefly? But it was an easy to abuse mechanic (which could either be problematic or awesome depending on how you feel about death)... Also it is a solution that requires a lot of effort from us.

Also wasn't there a RPG where you became a ghost and had to run to your corpse? I forget.

Maybe there could be a in-game penalty for dying... for instance the game allows you to continue, but you have to bargain for it, possibly losing a treasured item or a bunch of gold and possibly not even being able to afford it. The problem though is that if this happens a bunch of times the game would get harder and harder (although retrieving a corpse did get more and more difficult the more times the player died in Diablo and that seemed to work).

Non-permanent death solutions:

If it is a netherworld solution it is quite a lot of content creation and could be tedious after a few deaths.
If it is a checkpoint system then the random level has to be rebuilt exactly the way it started. This also amounts to players redoing stuff, which amounts to work, which is a bit of a nono.
If it is a fetch the corpse solution then it is also a little bit of extra work for the player and the tiny cost of corpse rendering.
If it is just a continue-from-the-nearest-entrance solution it is easy to implement but also makes death almost meaningless and includes a bit of work from the player (running through empty corridors)
Continues could be limited, which is easy enough to implement and makes death at least a little stressful as then it is possible to lose.
Making continues cost some gold or experience could make death more meaningful.
A sound effect (like an evil laugh) or an image could convey a message of defeat and make death more meaningful.


I don't think redoing stuff in a game is ever a good thing, but this may be my bias as I have a very low tolerance for grinding. I believe that this is part of why so many successful casual games are so short. Most of the continue possibilities that I can think of for Pocket RPG require at least some backtracking from the players part.

Backtracking or redoing is certainly an acceptable price to pay in a 6 hour game, but that doesn't mean it is ideal in a 20 minute game. Although it may work.

I know that non-permadeath would spoil or at least not benefit games like Desktop Dungeons, Crawl, Crimsonland, Spelunky, SAIS, Doodle Jump, Farstar, and most other short sandboxy games. My fear is that Pocket RPG shares enough mechanics in common with these for the same to be said of it.

But then it may just be my bias... I enjoy challenging casual games and I cannot think of any examples of short sandboxlike games that don't include permanent death and have been successful (If any of you have examples they would be greatly appreciated and probably allow me to sleep better at night).

Daemienwrath
27-09-2010, 07:42 AM
Completely disagree with this. If it doesn't have permadeath, it's not a roguelike. Plus the idea of a story is almost alien to roguelikes in general: You don't want your players to be able to experience the entire world in one play session, why would they come back then? Roguelikes give players vague goals (Get the amulet of yendor) and players invent their own stories or find little tidbits buried in events, action RPGs give players stories (Deckhard Cain was totally in league with Tyrael to dupe you into destroying the worldstone) as sequences of known goals (Collect X rat spleens to make the potion to cure the witch to curse the monkey to...).

Permadeath is not a requirement for a roguelike just that traditionally they do. This is not written in stone. The main feature of a roguelike is a randomly generated world and quests which can be replayed over and over without the same outcome. So the player will never experience the world in one play session as the world is random, your arguement is flawed. With regards to story there needs to be some sort of main concept it is by no means alien. The story is the setting and general constraints of the world that you are putting the player into. The goals that you give those players need to fit to some sort of narrative if they dont they have no value. Also what is stoping one from creating a more complex quest generator. Something which can generate chained quests to achive a larger over all goal to say "Collect X rat spleens to make the potion to cure the witch to curse the monkey to". Simple answer is nothing. You could still maintain the ideals of a roguelike without permadeath as well as having a more complex quest generation inside a rich narrative.

dislekcia
27-09-2010, 02:36 PM
Permadeath is not a requirement for a roguelike just that traditionally they do. This is not written in stone. The main feature of a roguelike is a randomly generated world and quests which can be replayed over and over without the same outcome. So the player will never experience the world in one play session as the world is random, your arguement is flawed. With regards to story there needs to be some sort of main concept it is by no means alien. The story is the setting and general constraints of the world that you are putting the player into. The goals that you give those players need to fit to some sort of narrative if they dont they have no value. Also what is stoping one from creating a more complex quest generator. Something which can generate chained quests to achive a larger over all goal to say "Collect X rat spleens to make the potion to cure the witch to curse the monkey to". Simple answer is nothing. You could still maintain the ideals of a roguelike without permadeath as well as having a more complex quest generation inside a rich narrative.

I gave you my perspective on permadeath and roguelikes, with the reasoning behind why I disagree with your assessment. I know full well that there aren't any Roguelike Rules™ written in stone anywhere, it's one of the reasons that people keep debating if DD is a roguelike or not. To me, personally, there has to be permadeath for a game to be a roguelike, without that it's a game set in a randomly generated world. My argument is that permadeath dissociates the player from their character and that without permadeath, the patterns that permeate any randomly generated system will not become apparent to players. Detecting and then USING those patterns is a very strong human motivator, so it's one of my big indicators of success in a roguelike.

Randomly generating story is something that I've spent quite a lot of time doing, it's one of the things that we were really proud of in Spacehack. Adding a story to a randomly constructed world isn't nearly as simple as you made it sound - there's a vast difference between a strongly constructed story and something put together by a machine ;) We had to use a lot of shorthands to allow the player to invent their own complexity on the random framework, so we'd never hit something as interesting as Starcontrol 2, for instance.

But, interestingly enough, when we took permadeath out of Spacehack in testing, the feel of the game really changed and the way players approached it changed a lot as well. I think the biggest surprise was the drastic shortening of play sessions: People would actually die far more often and wouldn't play multiple sessions in one go, they'd play, die and lose motivation to improve. So yes, you totally can stick with the ideals of a roguelike without permadeath, I think that's one of the awesome things about this medium - people can experiment with tropes and structures, especially as indies. The thing is that when you're pushing boundaries like that, you're creating new genres instead of looking to be defined by old ones. That's why I personally have an easy to answer set of criteria for a roguelike: Random generation, permadeath, characters don't matter. That's not done to limit anything, it's done to help me recognise patterns in games myself so that I can riff on them ;) Quick question: Minecraft. Roguelike or not?

P.S. Blackships, I'll reply to your much longer post tonight - There's actually quite a lot about Pocket RPG that I had incorrectly assumed from what we'd talked about at our visit to Tasty Poison :)

BlackShipsFillt
27-09-2010, 03:16 PM
Permadeath is not a requirement for a roguelike just that traditionally they do. This is not written in stone.

Well there has to be a line somewhere between a roguelike and other genres. Diablolikes have random levels and other random content but are definitely not roguelikes. Diablolikes do not have short replayable sessions though.

So it comes down to: Is an otherwise standard dungeon crawler roguelike without permadeath a:
a) Roguelike without permadeath
b) Diablolike with short sessions

And well of course it is totally possible to mix genres. These genres are arbitrary groupings anyway.

Do any good roguelikes exist that don't include either death or a lose condition?


P.S. Blackships, I'll reply to your much longer post tonight - There's actually quite a lot about Pocket RPG that I had incorrectly assumed from what we'd talked about at our visit to Tasty Poison :)

Awesome! I look forward to it. Tasty Poison does plan on making a lengthier Diablolike in the future (which is probably what you were thinking of), probably based on the same engine just with a lot more content (and 6 hour playtimes). Like I said, all the team here are HUGE action RPG fans (they're all animators remember), it's just me that prefers roguelikes.

I think permadeath is an incredibly interesting concept - even though we see it all the time in indiegames and casual games, triple A studios wouldn't touch it with any sort of metaphysical pole.

I'd be very interested in hearing more about those play sessions of Spacehack where permadeath was removed. I think Pocket RPG may circumvent the problem entirely by allowing a less roguelike option, but that in itself is perhaps problematic (as players are not necessarily rational).

I think it would be best if we test the permadeath difference on Pocket RPG with a group of beta testers. I hope we have time for that.

dislekcia
28-09-2010, 02:49 AM
Awesome! I look forward to it. Tasty Poison does plan on making a lengthier Diablolike in the future (which is probably what you were thinking of), probably based on the same engine just with a lot more content (and 6 hour playtimes). Like I said, all the team here are HUGE action RPG fans (they're all animators remember), it's just me that prefers roguelikes.

I think permadeath is an incredibly interesting concept - even though we see it all the time in indiegames and casual games, triple A studios wouldn't touch it with any sort of metaphysical pole.

I'd be very interested in hearing more about those play sessions of Spacehack where permadeath was removed. I think Pocket RPG may circumvent the problem entirely by allowing a less roguelike option, but that in itself is perhaps problematic (as players are not necessarily rational).

I think it would be best if we test the permadeath difference on Pocket RPG with a group of beta testers. I hope we have time for that.

Ah, didn't realise there were two games being planned. That explains a lot, actually.

Right, Spacehack. The game's been brewing in my head for something like 8 years now, the version we made for DBP 2008 was a subset of the basics that the game needed in order to work: Randomly generated ships, AI behaviors, bullet patterns and maps; Story was handled in randomly assigned related chunks, only a subset would be used in any particular game and only chunks that players investigated would spawn their follow-on sections; Weapons were explicitly designed.

With permadeath, players experienced more random enemy behaviors but also played very cautiously after their first couple of deaths. They developed favorite weapons (everyone likes the railguns and accidentally misspelled pheonix missiles) but were willing to spend upgrade points on what they had right now instead of waiting for later weapon drops. Without permadeath, players would race through areas trying to find bosses so that they could essentially farm for upgrades, when they died to normal enemy swarms they'd simply re-start at the entry portal and head another way, if no bosses were left, it was down to the next level as quickly as possible. Without that sort of milling, we wouldn't have found out that the random AI stuff became samey after a certain combat rating was hit - we never did fix that though.

We actually had one player (Karuji) wait an entire year to be able to play the game at rAge, just so that he could try to finish it. Without permadeath he'd have won the first time he played it and probably not cared about or enjoyed the game nearly as much ;)

That said though, it takes the right sort of game content for permadeath to work, as I've covered before. I think you could pull off a randomly generated game without permadeath, provided you could inject enough interesting meta-story elements that players unlock over time. Essentially focus on designing strong character elements (not necessarily their dialog or emotions, more like deep skill trees with nice, solid controls that engage players) and then slowly expand what the player knows about the world as they play through multiple times. You could probably hide some content behind those unlocks, but ideally not too much. Some sort of visual hook that shows unlock progress right in the main menu might work quite nicely... Basically I'm trying to think up something that you can use to keep players coming back (control mechanics, rich skills that won't all be maxed out by the end of 1 play-through and meta-unlocks) without needing permadeath to drive that. I think you could figure out enough of a punishment mechanic to not force permadeath on casual players... You do have the added problem of real-time gameplay on a mobile device, it would be unfair if a player could die permanently due to inattention if they get briefly distracted by random life events.

Dying to see your gameplay... You guys heading to rAge at all?

BlackShipsFillt
30-09-2010, 01:38 AM
I think you could figure out enough of a punishment mechanic to not force permadeath on casual players...

Did you mean "casual" players? Or did you mean players that don't want to be challenged so much? If anything I think it is the casual players that will favour permadeath... hardcore RPG fans who are used to investing in their avatars I think are more likely to balk at the idea of starting over (it's the hardcore guys here who do not favour permadeath).

But yes I agree that death can be punishable in other ways and it could still work. We are quite a long way away from testing out the difference.

There probably will be both permadeath and non-permadeath options, at least until I find the perfect solution. It does concern me that if players play the non-permadeath option, and almost all players will start off that way, then they are unlikely to transition the permadeath option (why would they, they've already played the game, the mystery has gone, why play it again just harder).

Let's assume for a moment that the permadeath option is ultimately much more fun/engaging (bear with me)...

Remember Unreal Tournament 1. There was that insta-gib game mode, well after playing a bit of that all the normal damage game-modes would feel impotent. Players are not inclined to handicap or nerf themselves, and so once having experienced great power in a game then anything less than that will feel weak. So once the safety of continues has been felt, permadeath will then seem even more cruel and unecessary. And, since players will most likely not choose permadeath initially, most players will be ruining their own roguelike experience (assuming the roguelike experience is superior).

Am I mad, or is this a huge risk (and it could possibly lead to a huge problem with the design, or is at least a opportunity that could be wasted, assuming a roguelike game would be more fun/addictive)?

ALSO, allowing endless continues ruins any mystery or greater meaning for the battles towards the end of the game, since every game ends in victory all the areas and battles will be equally familiar.

Fortunately if there are still items and monsters to unlock then players will keep playing, but not all experiences are equal, and the permadeath option in Space Hack sounds like the superior experience.

One compromise I am thinking of is this : Limited continues - lives basically. That way the player can afford to lose concentration once or twice, but can also be defeated if they are playing poorly and therefore will appreciate victory more.

Also, as a penalty for a continue, the player would at least have to move back to the start of the level (which won't be that far away), and, as the player's performance is evaluated base on time, this is a significant punishment (if the player cares about the scores and the unlocks that come with them).

Another thing that can be done, but may feel a bit nasty is offering a initial bonus with the "hardcore" option with permadeath. A little extra gold or something at the start, just enough to make a difference and guide competitive players into hardcore mode.


With permadeath, players experienced more random enemy behaviors but also played very cautiously after their first couple of deaths.

Do you mean they started over fresh more often, or that they were paying more attention? I don't like the fact that they may be more cautious. If Pocket RPG applies permadeath and is scored based on time then I would expect a bit of a bell curve of cautiousness, reckless at first, followed by more cautiousness, followed by more recklessness if the player is now experienced and is trying for higher scores.



Some sort of visual hook that shows unlock progress right in the main menu might work quite nicely...

There is going to be a score / unlock screen. Hopefully it will also list a bunch of future unlocks and the player will be able to see them and see what achievements they need to complete them. (for example: Axe of Maiming may come with 3 Blade Master wins, Amulet of Swiftness may come with completing the game in under 12 minutes and a Ring of Seeking may come with hitting 300 enemies without missing once... Though we will have to come up with betterer names for these things)


Dying to see your gameplay... You guys heading to rAge at all?

I'm busy testing it on the device, it's all kinda in pieces right now but were are trying to make a showing-progress video this week.

So sad... will not be at rAge.

Fengol
18-10-2010, 07:59 AM
I know it's a bit necro but I found this quote funny enough to share: "i won’t be telling you about how great permadeath is …. God uses it, it can’t be bad…" - Dr Richard A. Bartle

http://www.raphkoster.com/2010/10/08/gdconline-liveblog-of-dr-bartles-talk/