I think that the partial failure is always taken into account in games. E.g. shooters let you restock your ammo and replenish your health before you enter a room full of bad guys or after exhausting boss fight, before venturing any further.

There is a certain type of game where “no failure path” should be thought over and created by the designer. It’s the stealth game. Thief series is a good example for that. The game rewards the perceptive player, who observes the environment, finds a weak spot in a “tight” security and move unseen from location to location. The combat is also possible, but the margin for failure here is really small. Usually you can pick up a fight with one-two guards and you’ll be lucky if you don’t die.

On the other hand, a bad example of a game punishing the player without giving him proper guidelines first is Cryostasis. I wrote a few words about it here: http://t3edleveldesign.wordpress.com/2010/01/23/my-game-design-rants-1-cryostasis/

This is the reason why we need to give to the players: second life, second chance. On one hand, number of lives/opportunities should be enough for the players to make progress. On the other hand, the number should not be too big to let the players lose sense of strength toward each game. For example in the game we just designed, we introduced an element where players after getting that element, becomes much easier to win. Then we introduced many negative elements around that good element. So if the player decides to go for that good element, he takes risk by going through the negative elements around the good one. We think this might complicate the game in a good way.

Breaking the mental model is another important factor in designing the sense of failure in a game. Here is where the creativity starts working. For example board game, we must create different mental models for the users in each game. If in Monopoly it was good collecting cards and spending money, we designed a board game, where it’s better to save your money. When the player enters the game first time, the established mental model drives the player to use the old strategy “spending money buying the cards” then the player fails the game. Then he works out that he should save some money for surviving the battles, and then he might win. This change in his mind as a progress sets into the player’s mind and forces him to play again and again to try new strategies to win. This is basically the core value of the game that keeps the users playing again and again.

Therefore in order to attract the players, we need to set up the rules the way that could allow users fail and start again, then fail again and start again. This is a little bit that people might enjoy: making mistakes that could result in a failure, but you always have a chance to start all over again. Moreover, some players do not play ordinary games again if they play through the game and win, but if we design the game with a lot of strategy in it, then players would be able to experience much more every time they change their strategy. This is the key to fully engage the player into the game.

As for the player, he will be awarded more and more challenges, and be concentrated on engaging into the game rather than the actual winning result.

One advantage of allowing a player to die is that he has the opportunity to observe better players as they’re (more successfully) navigating through the chaos. It’s not only about the immediate feedback of “if I do this, I die.” The spectator also gets a look at other people’s strategies for dealing with the given problem.

Great post, by the way. I enjoy this blog quite a bit.

The regenerating shield not only made sense in terms of the game world, but it also meant that getting down to 15% health wasn’t going to have you reaching instinctively for the quickload, the scourge of virtually all “hunt the healthpack” FPS games up to that point.

Having finite health and a regenerating shield allowed the game to flow much better, and encouraged the player to learn about the way enemies attacked and adapt their tactics accordingly.

I agree with your main point wholeheartedly: failure is good. It’s character-building and makes you a better gamer, because it gives you a chance to learn from your mistakes.

Personally, I think the move to “failure is bad” game design and rewarding the player for the slightest inconsequential thing they do within the game (really, don’t get me started on “achievements”) is a reflection of the cultural shift from the expectation that success had to be earned to today’s instant gratification society, where people expect everything laid out on a plate for them.

I started gaming in 1984, when I was 8 years old. I’m not saying gaming was better then (certainly not from a technological perspective), but we were certainly much more tolerant of in-game failure. Perhaps this is more due to the technical limitations of videogames of that era, that were just making the transition from the arcades to the home, but I do think that there’s a cultural and societal element to it as well. I shudder to think of the reaction you’d get from an 8 year old now if you sat them down behind something as inherently unforgiving as Manic Miner, say.