There's an article posted on Kotaku that caught my attention and I found it interesting, mainly this:

"Where the evolution in failure lies, Juul claims, is in the guidance a game gives the player toward understanding the tools and the world. Where games of the 1980s could let a player make an irrevocable mistake and then keep playing, unaware, for hours, modern games are more likely to telegraph intent and requirements. With feedback from the game, the player experience is less likely to result in any failures. He cites Guitar Hero as an example of a game with instant feedback, and adds:

I think that games have become easier in one way; there's that guiding line from your current state to incompetence to becoming competent. There are more signals. There used to be a longer time where you're just sort of practicing blind. Now I think there are many more attempts at steering you toward improvement.

(...)

Then there's the question, especially if it's a mystery, of whether you have figured out who did it, for example, who committed the murder. So you might say that with other art forms it becomes sort of a social thing. It's not that you're denied access to the ending or something like that; it's more about that social performance at the end when you're discussing the end. So you'd say that … games are interesting since all of that is sort of hard-coded into the game; they'll tell you physically if they're doing well or not doing well."

Why not allow the player to reap the consequences of his own bad choices? Few games that do come to my mind.

Fallout 1 and 2 are one of the best examples, and a more recent one, even if to a lesser extent, Mass Effect 2, it allowed you to end the game with all your crew, part of your crew or complete the game but die in the process. That was one of my favorite features in Mass effect 2, choice and consequence, but I do think that it could have been a feature more explored and more open, not only in Mass Effect but also in other games.

I think we like to fail and then succeed. I could never accept an ME2 ending with dead crew members because I want to eventually overcome the challenge.

Also when challenge makes your free time more annoying than fun it has gone too far.

Fallout 1 & 2 four life...

In my play-through, Garrus was one of my crew that always had a spot on my party and was with my custom Sheppard all the way through Mass Effect 1 and 2, and in the end he died and that to me was really, really awesome, the idea of having to suffer the consequences of my bad choices later in Mass Effect 3.

In Fallout 2 you could make so many quests uncompletable by making the wrong choice or by killing a main character, I miss that in video games.

In Reality is Broken, Jane McGonigal cites some research that shows that gamers are actually "failing" around 85% of the time. She postulates two things: 1) that gamers, when having fun, actually would prefer it for the fun not to end (this is obviously a difficult balancing act for a game dev, because they must be having fun while failing) and 2) gamers do enjoy the challenge, really, as long as it looks like success is possible and the only hold up is them learning something, gamers in general react very positively to a challenge.

In those games is was more subtle and I did like it. I have no issues with quest decisions having consequences. Actually I want them to. When ME2 presents it as a challenge to kind of get the best score though, I want to be a winner.

I just read the articel, and he might be right.. i instantly thought of Dwarf Fortress while reading it, where in failure really is the fun part :P

I think a distinction needs to be drawn between making a conscious decision during gameplay which has unintended, negative consequences later in the game and not realizing that you were supposed to perform some action/obtain some item which stops you from finishing the game/leads to the otherwise negative results. I would argue that the former is often good design, while the latter is almost always bad design.

The "good" kind of failure is more about something like (in the article too, but hardly the first time it's been used as an example) in Guitar Hero, where you hear the song start to be off tune and the crowd starts to react poorly to you.

I don't see an actual reason for someone who isn't completely obtuse to argue that "silent" failure, that reveals itself hours or days later, is a good idea. It was what broke many an adventure game or RPG (some RPGs could not be completed without the right party makeup). For instance, Bard's Tale III: The Thief of Fate does kind of clue you in that the Thief class is important, but there's nothing like getting to the final battle and finding out that you cannot actually win it without a Thief (regardless of your characters annihilating every wave sent against them).

Often it's a balancing act between the excitement of "I nearly got there!" and the frustration of "But I hardly missed it! Again! ARGH!". I find the most delicious failures are the ones that make me chuckle with how miserable an attempt that was (poking fun at me or doing horrible things to my character) and the ones that encourage me to try again with the perspective of actually making it this next time.

I like how in Iji you first have to learn the hard way in boss battles, but get some hints after failing too often. Or in Fortune Summoners where you can restart right before the boss battles and don't have to go through the whole villain speech again (they just say things like "Oh, you want some more?" which is a bit out-of-character/meta-gaming but making failure much more fun; I wish Cave Story would handle it the same way).

I don't mind failure in games like VVVVVV, SMB and Super Crate Box either because it's so common and you hardly lose any progress. If failure makes the game too repetitive by requiring you to read through the same dialogues again or travel the whole road again, throwing you back more than, say, a few minutes at max, it stops being fun (at least for me). And then there are rogue-likes, where failure seems to be a vital part of gameplay. And I'd say it only works because each new try is like a different game again and no playthrough is the same.

An example say, your squad is going to cross a bridge:

1, The game gives you enough hints that there is a monster in the river, you'll need to perform some quest to kill the monster before crossing. You still decide to cross, you become lunch of monster.
This kind is fine.

2, The game doesn't give you hint. You cross, you get eaten, but then you can begin again right in front of the bridge and you can go search for way to kill the monster.
This is also fine. Not too good for a RPG, but ok for most of the adventure games or action games.

3, You cross, you get eaten. You go and find some playthrough and find out the only "good" way to solve this part is to feed a panda you met in the last town three days ago. Now you'll have to find another route.
As long as the game doesn't give you any hints when you meet the panda, it won't matter if it gives you any hints near the bridge or not. This is still fine in fact, at least it is not punishing


4, You cross, you get eaten. You go and find some playthrough and find out the only "good" way to solve this part is to feed a panda you met in the last town three days ago. Now the only way to progress is throwing one of your squad mate to feed the monster.
As long as the game doesn't give you any hints when you meet the panda, it won't matter if it gives you any hints near the bridge or not. This is the horrible type.

Not much like the satisfaction of doing something after failing a lot of times, if it doesn't take too long.

For that reason, Super Meat Boy is great.


But essential failure that reveals itself much much later is not fine for me. For example, in some Space Quest game you must pick up an item at the beginning of the game or you can't progress two thirds inside.
http://tvtropes.org/pmwiki/pmwiki.php/Main/UnwinnableByDesign


And it could have got worse:

I think perhaps I should clarify what I meant unintended, negative consequences. I completely agree if it is just a case that you had no knowledge even when making a decision so that a priori that decision is flat, then that falls under my category of "not realizing that you were supposed to perform some action/obtain some item which stops you from finishing the game/leads to the otherwise negative results". Even if it clues you in, the amount of information given in making your decision should be commensurate with the mixture of how much impact that choice has and how quickly you learn the ramifications of it.

I was thinking of non-linear story driven games where sometimes your choices impact the game in ways that you the player couldn't necessarily have foreseen given the information at hand (maybe for good or for ill). Where the choices may not so much reflect a reward/punishment mechanic but rather inform the journey your character is on - which sometimes may involve more or less challenge and bad thing happening depending on your choices. This is a little different from, the game is now un-winnable because you didn't have this class type or didn't pick up this item, etc ...

I'm pretty sure you understood what I meant, but just in case ... :)

Personally I wasn't clear if the interviewee in the article was stating that the evolution of instant feedback versus silent failure was a good or bad thing. I would actually submit he's saying the instant feedback for Guitar Hero is a bad thing (because it makes it easier so there is less failure), which, like you, I would disagree with. Most germanely, I think different games and different genres have different metrics and different forms of failing. With Guitar Hero and games like it- it isn't so much failing to play the song, but failing relative to your friends and to the "ideal" play-through which is a different sort of failure from not surviving to the end of the level. Perhaps I misunderstood him - I don't think the article was all that coherently written - but it would seem to me that giving a player a sandbox to discover the mechanics on their own can be okay ... but not if it results in simple frustration (and you know there should still be a manual :P). Failure is one thing, simply being annoying and frustrating is another. Of course everyone has their own limits where failure crosses into frustration, but some things I think are clearly bad design.
Is someone's avatar hungry? :)