"Designers were faced with a twofold challenge. First, they had to fit player failure into increasingly complex, fixed stories. The solution was to make failure independent of the story - you could die as often as you liked. The second problem was figuring out how to penalize failure without requiring the player to replay substantial areas. Around the same time, LucasArts, faced with a similar conundrum in the adventure game genre, removed death entirely. But RPG designers could not give up killing the player, in part because cheating death is such an integral part of fantasy stories. Instead, they relied on saving. If the player saved his game regularly, death would not force him to replay much. And if death ended the game, failure didn't cause any story problems because restoring a saved game "undid" the death and reset the story."
Marty M. O'Hale explains why death is overdone in "Killjoy."
I think Marty is 100% correct here. The reason that rogue-likes appeal to me is because they've managed to capture the elusive quality of the reward being the journey not the goal. A racing game wouldn't be half the fun if you were constantly reloading the make sure you hit an S-Curve correctly. His comments ring true to me for more than just role playing games, which are surely guilty of this problem. First person shooters suffer from a lack of immersion when you reload every battle. (or worse, every exchange of fire)
The job of the developer isn't an easy one here, Do they remove quick-saves entirely? Should they instead break up their game into 15 - 30 minute sections so that it can still be played by people without a lot of time? I'd personally be satisfied just to see more games implement Diablo II's hardcore mode. It would force the developers to make sure the game could be played successfully through without failure through intelligent planning and execution. It caters to both crowds in this, because I'm sure that there are people who don't want to deal with making a new character or starting over if they make a fatal mistake.
Speaking as one of those players who likes saving whenever (My FF12 games go: Load, teleport to another city, save, get the quest in that city, save, teleport to the quest area, save, get in the dungeon, save, etc.), here's what I propose to appease both camps. Just graduate the save frequency with the difficulty level. Easy/Normal is save anywhere. Hard is maybe a console-style checkpoint system, where you can only save upon hitting a certain point. Ironman/Hardcore is the one where you can never save ever to prove you're awesome. It also means the advanced difficulties are actually more difficult, rather than simply giving enemies autoaim or noclip or whatever.
Or, another option, you could make a penalty for loading a save file. That is, if you load the game from the main menu there's no problem, but if you hit the "load" button while in-game you lose ten percent of your HP and golds, or something (and the save file itself is overwritten with these new data). That way, quicksaving still works as a way to end your session, but quickloading isn't quick at all. Plus, remember, your player does still need a way to reverse catastrophically bad decisions on their own part. Taking that out would be unfair in any game with static content.
That wouldn't be a solution. That would just be a way to get your players out of the habit of quicksaving every twenty seconds.
The core problem is that many video game engines simply don't have the room for the kind of creativity needed to save your neck in a bad situation. There are encounters in any given RPG where it's mathematically impossible to not take a lot of damage. Get rid of those. The player should always be able to think, fight, run, talk, or otherwise find their way out of an unexpected threat.
Precisely where to balance it is an issue, but I think you'll find more success in having a variable (user-specifiable) difficulty level, so that the player always feels no more threatened as they want to feel. Eventually they'll realize that quick-saving all the time is boring.
This article reminded me of how much I love Dragon Quest VIII without really realising why.
I mean, just to save you need to go through alot of dialogue, and you can only save at certain places. After saving, they even ask if you want to quit right there. It's the most total opposite of computer game saving systems that you can get.
I agree that designing games/levels that don't kill you every 5 steps is the way to go, but in conjunction you need a limited save system that makes sense and isn't exploitable. One of the most interesting save systems I've seen comes from Operation Flashpoint. The game has automatic checkpoint saves, but it also allows you one arbitrary savegame to use whenever you want during the course of a certain mission. It's a nice compromise... it can be used either as a convenience save (e.g. saving after you've spent a lot of time setting up the perfect ambush) or a post-"omg I just did an epic maneuver that I surely won't be able to pull off again" save, but either way it prevents quicksave/load exploitation enough to not remove the challenge of the game.
Of course such a system would probably only work for linear/mission-based games.
Very fine article.
One caveat to the save game recommendations, though: instant retries are not only used by players seeking to avoid tedious deaths. For some gamers, the real world can interrupt or disrupt any gaming challenge at any moment and reloading from a saved position is the only option. It's all very well saying "apply a small permanent penalty for failure" when the task in question was fairly attempted, but I don't want a small permanent penalty applied every time one of my five year olds runs between me and the TV during a critical sequence !
For linear games, the solution here is nothing complicated. Clever level design is all that is needed. Just look at Half-Life, its sequels or any similar games. By carefully thinking about each encounter, the designers can ensure that it is always challenging while always being completely surmountable. A good player will feel like a hero, constantly facing off against extremely challenging opponents and almost always surviving.
As a result, such games are great. You'll battle and battle, that health recharge point comes at the perfect time after your fifth consecutive battle with a competent opponent.
Nonlinear games could use systems that modify the amount of damage enemies could deal based on the ease at which the player a hero could heal. The sense of attrition can ewar on a player and make things exciting.
Trying to limit what the player can do with the saves is not hte answer. Players get to make their own choices about how they play games and aif they choose to do it in a way that ruins it for them, that's their fault. It'd be nice if you could set options to facilitate such things but a little discipline can make a game much more fun.
Anyone else remember the 'good ol days'? On the original Bard's Tale (C64), the manual told you to use your favorite copy program to copy your characters. If your party wiped in the game, they were dead. Your latest game save loaded up with a party or corpses. Now, you could revive them by creating another character, getting enough gold, and then taking each one to the temple, but it wasn't easy or fun. It did, however, make things much more tense during big battles, especially when you realized that it had been a while since your last backup.
it's hard to say what the 'right' solution is here, but something that would bring that level of risk/reward/punishment back seems to be the place to start.
Here goes the dreamer in me:
What I would love to see is a game where in most situations, "Death" simply set you onto a different storyline: Get beaten by orcs, and you become a prisoner of thiers, now you either cooperate, and try to get promoted out of slavery, try to escape, or fight back, and actually do die. The dragon you fight woln't kill you unless you have reached a certain level of power, but it will take anything "shiny" (coins, gems, and anything else that the game has marked). Vampires woln't kill you, but they make you a charmed slave (same as the orcs, but you can't fight back). Get blown up by pirates? Nope: you have an automatic escape pod. However, you now better hope you can get away from them safely, or negotiate your release, or whatever.
Basically, you get another chance. In fact, you get as many chances as you want. However, each chance you blow means that you lose something, whether wealth, connections, or abilities (That wizard that beat you spared you, but took several spells out of your spellbook. You're alive, but short a few spells. The Ghost drains a level before your panicked flight takes you out of the house. And so on.)
Back to reality:
I think the short term solution is make some sort of cost for saving. Nothing major, but that adds up for compulsive autosavers. Perhaps a cost that starts at zero, but increases every time you save: more so if you are in an easy area, and decreases the longer you spend in one consecutive stretch of meaningful play (not just standing around), without saving, as well as after a long time (say, an hour) away from the game. This penalizes people who save over and over again without thought, while not overly effecting people who save before walking away from the game, or people who save only before critical sections of the game.
First, unlike what the author wrote, quicksaves were not invented in RPGs, but FPSs.
Second, the above suggestion for limiting the number of saves according to diffculty is a good one, and is used in the latest Hitman game.
Third, to Dom, if you have interruptions in your game, you can pause the game, to continue later on (especially true in slower RPGs, less so in fast action games).
Fourth, ZacQuickSilver, nice dreams, but I don't see how they can be done without the improvisation of a live GM (until such AI is created, of course). In addition, there should be no penalty to saving in my opinion... only to loading, and only if it's done in mid-game. (could result in players quitting the game and restarting it in order to cheat the save system, but every system can be manipulated.)
In general, increasing the feeling of consequence of your actions is very improtant, as mentioned in the article.
It's admittedly a bit late, but I simply must comment. Best article I've read in a long time, not only because it is insightful and fascinating, but because it presents a realistic solution to the problem raised. This is an issue that exists in many other game genres; I am currently deciding how I'd fix the problem in the Phoenix Wright games, for the mental exercise. Anyway, I hope to read more from Mr. O'Hale in the future.
I do think the suggestions offered by the author are interesting and I would like to see games do that, but there are very few that do or will, and a lot of hardcore gamers would criticize games as being too easy on the player anyway.
I'm not sure I agree with the idea that games are more brutal now because of the saves. Some games certainly seem designed on the assumption players will save at certain points. At the same time, a lot of very old games would absolutely kill you in all sorts of unfair ways knowing perfectly well that the player would have to replay huge chunks of a game. Some game designers are simply sadists.
The first time I quickload the magic is gone.
I don't see the solution in taxing the load, but moving away from the bi-polar cause-consequence worlds that define todays games.
Time to start making games for adults, I'm far from being a definition of an adult, but very games today are realy fun for me. Gone are the days when I could just plug away in some random FPS. After few levels, I just loose interest in killing whole armys of foes and dying as the only consequence of my health bar reaching zero...
I think that quicksaving needs to be done away with as it stands. A system of save points could be used, and to cater for having to end the game suddenly, use the system used in some console games, and a few PC games; the user makes a quicksave, but the game ends, re-loading the quicksave deletes it. This stops the player from abusing the quick save, but still allows you to save and quit whenever you want, without sacraficing difficulty.
I admit you could bypass the system by copying quick-save files to other directories so you can replace the deleted one, but it still takes away the quicksave /quickload mentality, as it's no longer 'quick'.
I think autosaves are almost always a bad idea; though sometimes game design is so broken, as the article suggests, that they become the preferable evil. Save points are usually ideal, so long as they are rationally placed, as they break the game into manageable chapters, allow you to have multiple saves in case you want to replay a fun area (something lost with checkpoints), and allow the developer to very easily scale the game's difficulty-note that some games include singular, optional, very long and very difficult battle sections (usually some sort of arena) during which you get no opportunity to save, or very little opportunity.
I agree with Shannon Drake's idea (which I've only ever seen used a few times in actual games) of changing the save-ability of a game based on the difficulty; I think most games should have difficulties that range from totally unchallenging (to the non-gamer who wants to experience the story or the visuals) to difficult (actually difficult).
I also think that all games should allow you to quit at any time, and pick up at the exact moment you left off. Yes, you COULD abuse this by going into the game's file structure and finding the save file, and use it basically as an autosave; but that's a lot of work. Most of us would rather just try the level again.
This is eaactly the problem I am having with DA:O at the moment.
The combat is so brutally tough that quicksave/load is an absolute nessecity, but I refuse to use QS/L in dialogue, I always stick by my first choices of dialogue (even if I'm forced to replay the same conversation because of a swift death from a fight afterwards). Thus the experience of combat completely disconnects me from my character.
However all of this said, I'd still rather have DA:O's tough combat, than mindless grind (which I feel is indicative of the developer not respecting the players investment of time), at least with each failure I do learn something about how to do better next time.
I've a simple idea that sounds like something that appeals to you, though I'd bet it's been done before:
If a player is to die then they "black out" and see death (in other words a "near-death experience"), and to stave off death, to step back from the brink of death they need to solve some variety of puzzle or hand death his ass in a round beating, something along those lines. Once the aforementioned is completed then they are returned to the very moment that they "blacked out" at say 50% health, meaning that the player has lost maybe 3 or so minutes to make up for their failure, but they're still have FUN making up for their failure, thus meaning dying isn't a massive problem and you aren't being thrown back constantly, but dying and having to consistently solve a puzzle (or kick death's ass) would be something that you would want to avoid because you've got an objective to complete.
But as stated in the article death is a tricky thing to deal with, punish a player too much and they won't want to play, don't punish them at all and there's no challenge, and worst of all there's no blanket solution, especially with my idea, sure it may work in a "Fantasy RPG", but it sure wouldn't work in a game like Portal or MW2, or any real multiplayer game, meaning that even if one genre can fix the problem almost all others will continue to suffer.