THE DEATH PENALTY

Posts

Pages: 1
LockeZ
I'd really like to get rid of LockeZ. His play style is way too unpredictable. He's always like this too. If he ran a country, he'd just kill and imprison people at random until crime stopped.
5958
Hey, I did a thread about this topic back in 2012. A lot of this opening post is gonna be just copied and pasted from that thread. But it's been a long time, the site has different people than it did five years ago, and a lot of us have grown and think differently about games than we used to. I've decided that it'd be cool for the people who are on the site now to participate in more game design threads, so yo, let's talk about video games. Specifically, about dying in them.

Dying in a video game sucks.

Or does it? It depends on the game; on how many tries it takes you; on whether you know what to do differently next time; on whether the death was caused by something completely in your control or not; on what you lose when you die.

It's that last factor that I want to talk about. In this topic we're talking about what to do to the player when he or she dies.

Game over, you say. Duh.

First of all, that's only one of a large number of options. And second of all, there are several different factors that can make a game over more or less punishing. What I really want to delve into is why are punishments for failure necessary, why are they problematic, and how can you maximize the player's feeling of accomplishment while minimizing frustration?


Let's take a look at various punishment options.

Game over with periodic save points
Like I said, there are different factors that can affect a game over - the biggest one is how saving is handled. With saving allowed only at save points, which appear periodically through the game, the designer can designate a longer stretch of the game as a "single non-stop challenge" that you can't restart from the middle of. Very few games allow you to save in the middle of a battle and restart from that point, right? Because that would be super cheap, it heavily encourages abuse of enemy AI and stuff. In traditional RPGs where most normal battles are not really very dangerous, but dungeons are essentially wars of attrition where your goal is to get through all the battles without using up your resources, you can make the same argument for a dungeon that you do for a battle. So the player has to restart the challenge - which is the whole dungeon.

The big downside here is that the player can get sent back 30-45 minutes, which can be extremely discouraging. Losing the better part of an hour to a game over feels like bullshit in cases where the player only made one mistake near the end, but has to redo all of it. If the player didn't just make one mistake, but actually is having trouble with managing their pacing and with the challenge as a whole, they're likely to lose several times before finishing, which means that they'll be stuck in the same dungeon for hours and hours.

I've seen this used as a way of disguising the wild difficulty swings in games that have bad balance, since players can go through a very easy dungeon without realizing how easy it was until after they beat it. Don't do that. Just fix your damn difficulty.

Game over with save-anywhere
Basically, here, we swap the "game over sends you back too far" problem for the "you can start from the middle of a challenge and cheese it by brute force" problem. A lot of people prefer this to the above. In games where you are fully or mostly restored after each battle, I can't come up with a reason not to allow save-anywhere. The overwhelming majority of modern commerical games use this method in combination with frequent automatic quicksaves, so that the player doesn't have to remember to manually save after every battle. Adding a Retry command to the game over screen is usually a no-brainer if you're using this method, especially in a game where battles take place on a separate screen from dungeon exploration.

Game over with automatic saving at checkpoints
This is kind of a middle ground between the above two. At first glance it doesn't seem that different from periodic save points aside from the fact that it's automatic and thus you don't have to remember to save. But in practice, this system is often used when you only have a small number of enemies between checkpoints. It changes the individual challenges from being entire dungeons full of battles to being rooms or corridors full of battles.

Respawn at nearby point without losing XP
FF6 does this. So does Earthbound. So do a lot of MMORPGs. Essentially here, you get sent back to the last save point or to the nearest church/graveyard or some other sort of nearby respawn point, and have to redo the battles between that point and where you died, but you get to keep any XP you got. So it's a little easier the second time. And if you die again, you'll have even more XP, so it'll be a little easier again the third time. And so forth. I find this to be extremely nice, myself, because it helps out players who are having trouble without making the game any easier for good players, and more importantly because it makes the time you spent not feel like a waste. Sometimes there's also a small penalty - usually gold, as payment for hospital fees or for reincarnation services or for armor repairs, but it's typically a trivial amount.

Respawn with heavy penalty
I've seen this used mostly in online games, such as FF11, but also in single-player games where the world "persists" through your death. Dragon Quest takes away half your money and sends you back to the starting town when you die; many later games do the same thing but send you back to the most recent town you saved at instead. It's not the worst thing, but don't combine it with save-anywhere, or people will just reload to avoid the penalty.

Delete saved game on death
This is as brutal as it gets. The idea here is that the only type of save that a player gets is a quicksave; you can save at any time, but the save is deleted upon loading your game. And if you die, you start the game over. As RPGs go, this is most popular in short games (Gauntlet, Desktop Dungeon) and in roguelikes such as Nethack. I guess the idea here is the whole "can't restart from the middle of a challenge" mentality taken to its logical extreme: the entire game is effectively a single nonstop challenge. To me the cost outweighs the benefit here so heavily that this option isn't even on the table. If you can justify the use of this in games longer than an hour, I'd love to hear your point of view.

Dead characters are permanently lost
Fire Emblem is the classic series that's famous for permadeath, but if your entire party dies then it still gives you a game over. That's not the case in X-Com and Darkest Dungeon - when your party dies, the game auto-saves and sends you back to your base. When your characters die you're forced to continue the game with other characters, which often requires an extra hour or two of levelling the new characters up to the strength your old ones were at. In many ways this is only very slightly less punishing than deleting the player's save when they die. I'm personally not a fan, but it can sort of work if you have a game where the characters are very much expected to die a lot (as is true in both of those example games). One would expect a very low level cap in such games - the level cap in Darkest Dungeon is level 6 and in most X-Com games is even lower.

Limited number of lives/continues
This is super rare in RPGs. Like, to the point that I've never seen it in my life. Ultimately you have all the same problems as the above "delete saved games on death" method, but the player is less likely to encounter them. Some games do play with spins on this, giving each character on your team a limited number of lives before they permanently leave the party (SaGa series), or giving the player a limited number of continues to retry the current battle and if you run out you have to reload from the beginning of the dungeon (Wild ARMs 3). If done right, maybe this can make "cheap" deaths feel less cheap - because they don't feel like a complete death, they feel more like... losing some of your HP.

What are your favorites? Why? What are the biggest problems do you have with the others? Almost all of these have both good and bad points, so I guess it's largely a matter of which good points you value more and which bad points you find more irritating.
Two heaviest penalties for death in RPGs I can think of, not counting strategy RPGs with permanent character deaths, are in Azure Dreams and Breath of Fire: Dragon Quarter.

Azure Dreams offers no permanent level-ups for the main character; you're reset to level 1 every time you re-enter the main dungeon. You can only take five items into the dungeon with you when you enter, and you can only leave the dungeon if you have a specific item which appears randomly, so unless you find an extra one and sacrifice a valuable item slot to take one with you, there's always a risk that you'll get killed before you find one. It's a pretty tricky roguelike format, and you're rarely ever completely safe.

The main progression that allows you to get further through the main dungeon in successive playthroughs comes from leveling up your monster familiars, who don't lose their levels upon re-entering the tower, but who can only fight for a limited amount of time before running out of magic, and by gradually improving your equipment, which both from finding pieces scattered randomly through the dungeon, and from finding occasional items which permanently boost their stats. These permanent boosts add up a lot, and are likely to become your primary source of progression. The highest base attack power for weapons you can find is +7 or +10 for a pair of extremely rare weapons (Why list both, when +10 is clearly higher? Because the only weapon with that attack power comes with a substantial cut to your accuracy, so many players will prefer not to use it even if they find it.) But even without using cheats or exploits I've discovered through later play or game guides, I've built up weapons with attack power dozens of points higher than that.

When you die, you lose all the items in your inventory. Not just everything you collected on your trip, but everything you brought with you into the dungeon. Since leaving behind your best equipment means probably not being able to make it further through the dungeon, it's very likely that dying will cause you to lose a huge part of the assets which allow you to progress further in the game.

This one can be circumvented with a relatively simple exploit though. It's a PS1 game, and you can prevent the game from overwriting your save file when you die by simply taking out your memory card while you explore the dungeon.

Breath of Fire: Dragon Quarter has a system that's pretty conceptually similar. Basically, the game is only a few hours long for a single playthrough, but you're practically required to restart the game multiple times with a New Game + system which carries over elements from partial progress in order to make it through. Dying rather than deliberately restarting doesn't just knock you back to the beginning of the game, it eats up pretty much all the bonuses you get from the game progress you've accumulated. Unlike Azure Dreams, this one has some pretty strict systems to prevent you from avoiding this by just taking out your memory card or copying your save file.

Both games play this for pretty much the same payoff. The final bosses of both require you to reach what would be Game Over conditions anywhere else in the game in order to finish. Lots of games pull this, but it never has the same level of impact as when the game has taught you to instinctively freak out at the prospect of getting a game over. Whether it's worth putting up with for the entire rest of the game for the sake of that depends a lot on how you feel about games which force you to operate under high stakes.
I've probably said these things in other threads already, and they'll probably still be somewhat off-base or useless given my not really being a frequent RPG player, but w/e:
Game over with periodic save points
...
The big downside here is that the player can get sent back 30-45 minutes, which can be extremely discouraging. Losing the better part of an hour to a game over feels like bullshit in cases where the player only made one mistake near the end, but has to redo all of it. If the player didn't just make one mistake, but actually is having trouble with managing their pacing and with the challenge as a whole, they're likely to lose several times before finishing, which means that they'll be stuck in the same dungeon for hours and hours.

Yeah, when you force the player to potentially replay longer segments of your game, your design really has to be immaculate and consistent to hold up to repeated play. I play a fair amount of arcade games in MAME, and a common thing that I don't care for is how the first stage or 2 can become boringly trivial after a few plays, but then the final stretch might present this ridiculous wall that requires so much trial and error or downright luck to overcome that, if I were to start over from the beginning every time, I would completely burn out on the early game before ever beating the end-game. And of course, different people have wildly different amounts of patience.

And then of course there's the shitty thing some games do where you have to rewatch the cutscenes every time

I've seen this used as a way of disguising the wild difficulty swings in games that have bad balance, since players can go through a very easy dungeon without realizing how easy it was until after they beat it. Don't do that. Just fix your damn difficulty.

I'm a bit confused by this statement. It seems that, if you wanted to hide erratic difficulty swings, then using a more frequent saving system would be better, since you'd then be able to rapidly retry and brute force a hard part and completely forget about easier parts before it. Or, do you mean more like that people try to say that their dungeon is "hard" purely because of some ridiculous boss they throw at the end of it, without putting much though or effort into the dungeon design leading up to it?


Delete saved game on death
...
If you can justify the use of this in games longer than an hour, I'd love to hear your point of view.

Well, you know how the typical system of making players get through ~30 minute chunks can create a feeling of tension/stakes, and force the player to achieve some level of game mastery to progress? Well, this is like that, but just many orders of magnitude more extreme. That's pretty much what I remember seeing people say they liked about games like, say, Tekki/Steel Battalion, which had lives, but would still delete your save if you ran out, and the game went on for like, 60 hours or some shit. Which is of course what a larger proportion of people disliked about it.
There's not a huge market for this level of intensity, but there is some market for it.


What are your favorites? Why? What are the biggest problems do you have with the others?

My preference is doing 10~30 minute chunks of game in one go, reverting fully back to a previous save upon failure, nothing holding over. I just prefer to take games as they are designed, without too much "brute forcing" or any opportunity for grinding a challenge away. I say 10~30, and while I enjoy ~30 minute arcade-like games, I actually prefer games more like Silent Bomber or Ginga Force where the game saves after every stage, but the stages are pretty meaty (5~15). I just don't like super-frequent saving like Meat Boy or what people usually use emulator save-states for, because that feels like brute-forcing sections rather than really understanding the design of the game. Of course, that only applies if the design of the game is actually good...

What I really want to delve into is a) why are punishments for failure necessary, b) why are they problematic, and c) how can you maximize the player's feeling of accomplishment while minimizing frustration?

This is going to vary a huge amount between people, but for my money:
a) I believe a game is as much about being able to lose as it is being able to win, so if nothing happens when you lose, than, as far as I'm concerned, there's no game. Of course, "losing" could just mean losing points and not actually "dying" (see: gallery shooters), but yeah, there's gotta be something to lose other than just prolonging the inevitable.
b) and c) seem like they're almost the same subject, so, I'll say that it the obvious statement of not having cheap shots and wild difficulty swings, and generally not forcing players to repeat the exact same shit too many times. Like, you mention roguelikes deleting your save file, but they usually feature a heavy amount of randomness, so starting over may not be as repetitive as it would in other kinds of games. Likewise, it's for that reason that I'd say for a puzzle game, "save after every puzzle" is the only system that really makes any sense, unless the puzzles are randomly generated.
I can't remember what game it was, but I liked the idea of picking your old corpse clean with a new character. Especially if you get to respec the character and don't lose too much progress.

For example, you start as a knight, you get to stage 2 and die to a zombie. Restart and spec into cleric, jump to stage 2 (since you already passed stage 1 as the knight), grab your old kit and continue to stage 4 where you die to a werewolf. Respec as hunter, jump to stage 4, grab your old kit and go through until you die in stage 8. Go knight again, jump to stage 5 to grab a rare drop for knight, jump to 8 and grab kit, continue onward.
Isrieri
"My father told me this would happen."
6155
Limited number of lives/continues

This got me to thinking about Final Fantasy and its Tent/Cottage items. I think they've phased them out after IV, since all they do is give you a partial or full HP/MP re-fill. But in the first game, they also let you save your game anywhere on the overworld. That made them useful in a sense, but you still couldn't use them inside dungeons. Therefore they kind of sucked. I never spent much money on them that wasn't better spent on a heal potion.

So what if you took that idea and ran with it? Instead of allowing players to save anywhere, or at registered points, they carry their saves around with them (lets say you still get hardsaves at towns or similar safe zones for convenience). Say you start the game with five saves or so, and you have to find all the rest along the way. That would make those save items ridiculously valuable.

Although I think if you went that route the temptation to drop save items in all the dungeons would be too great. BUT it could offer a risk-reward system. They're functionally the same as savepoints if you can just find them hidden in chests, but if they're worth a lot of money then selling them becomes an option, to make the game easier in a different way. The challenge thus becomes a bit more dynamic.

I can't imagine a game where losing all your 'continues' sends you packing back to the beginning in an RPG. The genre didn't become really big until after saving data to the console became possible. Saves at regular checkpoints with hard-to-revive death states is, I think, the way to go.


What I really want to delve into is a) why are punishments for failure necessary, b) why are they problematic, and c) how can you maximize the player's feeling of accomplishment while minimizing frustration?

Personally if I'm going to play an rpg video game, I want that game to try and kill me. Turn-based affairs are my favorite and in those its all about prep and planning - If you don't plan well enough, you get stomped and have to use a different plan. All challenge comes from the dying, and the oft-overlooked aesthetic aspects of death (ie; the in-game consequences of characters actually losing) are also pretty vital.

But dying sucks in games that take hours and hours to play to completion. Moreover rpgs are by design meant to be completed and have their stories experienced. Combat gameplay is kind of superfluous if story is your only goal. If you want challenge, you need a penalty for sucking. Dying and being punted to a save is a lot better than a character getting downgraded, being lost, or the enemies getting more difficult or some other wacky alternative I can't think of, because if those were the penalties I'd probably reload my save. Just make dying not delete all those hours of progress. Taking you back before the boss or a segement of the dungeon that killed you is the best move in my experience. If the random overworld encounters ever kill you its usually because you wandered somewhere you weren't supposed to, or again, you didn't plan well enough (or the game could be badly designed i dunno).
I looked up my answer from 2012. I guess it hasn't really changed much. I like to be able to save anywhere at any time. Preferably with multiple quicksave slots (in case I hammer that key by mistake mid-air falling into a death trap) and also autosaves at certain points (in case I forget to hit the quicksave button some time) and of course manual saves that can be named in the case of branching stuff or something awesome I want to be able to quickly get back to at a later date.

That's the general one. The other ones games are often specifically designed for. The Hotline Miami/Super Meat Boy "instant respawn and try again" is what the game is all about. Roguelikes are roguelikes.

I guess I'm generally against death penalties. Because most games these days are not designed around the idea of dying (only the idea of almost dying).

However death penalties I like and would like to see more of are setback penalties. Instead of ending the game or making the game unwinnable, losing actually makes a game more fun. Now this rarely, if ever happens and it is incredibly difficult to pull off. But it is very satisfying when it works. It's mostly in open-ended games it works. Darkest Dungeon for example. You get a party wipe or better yet an almost party wipe and you have to grind some lower level stuff for a bit while also maybe retaining that one awesome dude that managed to survive. That creates a story about the veteran who went through some shit. Similar happened to me in Xenonauts. My main squad got hammered and one or two experienced guys survived and the next couple of missions was all about keeping those guys alive and see which of the new cannon fodder would make it. (contrast this with Firaxis' XCOM where a similar situation essentially made the game unwinnable)

Or take Crusader Kings 2 where you can lose some territory or have your kingdom split up in a civil war and that's the stuff that actually makes the game good. You lose, but it creates more excitement.

So Death as an opportunity is the best death. And if that can't be achieved then unlimited save slots.
halibabica
RMN's Official Reviewmonger
16948
I'm such an old fart, I think I actually remember the last time this topic was raised. It's a good topic, though!

Someday when I get back into making RPGs, I plan to use a combination of some of these examples to make failure less serious while still impacting gameplay. I'll be using periodic saves along with "Respawn at nearby point without losing XP" and "Respawn with heavy penalty." Basically, if you fail in battle, you're sent back to your last checkpoint, keeping your exp, equipment, and even forward progress. The cost of your failure is monetary and scales depending on the magnitude of your defeat (party level, was it a boss fight, etc.). If the cost is greater than your bank account, you go into debt and can't buy more items until you've scrounged up more dough.
An obvious answer for 2017 might be something like what the shmup Border Down does: change the game for each death. (In Border Down, the game actually becomes harder with each death. There are 3 levels, and you get punted to the next lower level (of hell) with each death, until you punt out of the lowest level and die for good.) That would be a lot of work, though...
I guess by necessity this dives into only games where characters die?

I've worked on quite a few projects now, each with conflict systems of some sort, and dying was never an option. That makes my opinion on what to do after death kind of meaningless, but I'd like to posit that you can build your game in such a way that Life or Death isn't always the outcome. Some games have conflict systems that don't need Death and they still go for a Game Over system. Before deciding where you stand on the Death Penalty, I think it's a good idea to really think if every failure necessarily leads to character/player Death. Then you can just think in terms of Failure Penalty, like Pokemon and stuff.

In terms of Failure Penalty, I like it when the game continues but you gain way less than you would by succeeding. So there's always a benefit, but you still have a huge urge to be efficient with your time because what you COULD gain is greater. This also negates any heavy feeling of loss or backtracking.
Anyone played NieR 1 and 2? Sure, it was at the end of the game that you deleted your save file but goddamn was it for an awesome reason. And not just one save file. All of them.


@Lib: Nier 1 spoilers
The fact that it was there to really drive home the point that both you and the protagonist were making a sacrifice I think was interesting. Also it isn't really at the end of the game per say just the last ending that you got. But that ending requires pretty much (arguably) 100%ing the game in really painful ways that OF COURSE you'd want to have a save file there just to log your achievements. But I think in the end the game argues you don't need memories, evidence or permanence to be a meaningful lifeform in this universe.
author=Isrieri
Instead of allowing players to save anywhere, or at registered points, they carry their saves around with them (lets say you still get hardsaves at towns or similar safe zones for convenience). Say you start the game with five saves or so, and you have to find all the rest along the way. That would make those save items ridiculously valuable.

The original Iron Gaia and some tower climbing game did that, actually, and I hated it in both cases. Part of the problem, admittedly, was that both games were fairly difficult and had highly constrained resources, which is exactly the environment where you'd want to save a lot.

The one place I felt this might work is in the space between save points in dungeons. Allow the player to save anywhere at any time, but only a limited number of times. That way they can still quit when they want and save progress, but not have the advantages of save scumming. I think what makes this different from something like Iron Gaia is that you still have unlimited saves outside of dungeons, and could still have one unlimited final save right before the boss. That way saves aren't so much a resource to abuse but one which facilitates the user experience. That said, I haven't actually played a game with this yet, so maybe it isn't as helpful as I think it would be.


Predetermined Outcomes
I'm surprised no one's mentioned this one yet. One effective method to prevent save abuse is to simply make random outcomes predetermined. In other words, when you save, you'll still get the next encounter in exactly five steps, and it will still be the same encounter. Or your "random loot" drop will still be the same. Or whatever. This allows you to instantly retry challenges, such as boss fights or even random battles, but does not allow you to eliminate chance entirely.

The only example I can think of that did this is Civilization 4. In Civ 4, no matter how many times you reset, the battle outcome was always the same.
Pages: 1