When our hero fails to complete a particular task, falls off a cliff, or loses an enemy battle, the player often gets a game over. They are forced to restart from their last save point if they feel the need to continue playing.
However, the act of getting a game over has several unwanted side-effects that we don't realize. In some cases, it may result in players spending unnecessarily large amounts of money on healing items or overly-conservative item usage just to simply avoid the situation to happen again. The most obvious side effect (which is more of an effect to be blunt) is frustration and wiping the game off the face of the player's hard drive. With amateur games it's best to assume the player is either overly-critical, impatient, or both. I won't go as far to say the Game Over feature isn't useless, or that players should never lose, but it is a bad choice from a design perspective and it's usually the easy way out. Let me explain.
When creating a Minigame, for example, it's very easy to test for a losing condition and call a game over. It's an abrupt halt and can be done in a single command. I like to say that it favors the designer over the player. However, if you lose the game and you are given an option to retry, skip, or even give up and get a game over, the player gets much more options. This is definitely not as easy to implement, but it would make the player much happier knowing they get options. It favors the player over the designer.
But when we're making games, and we aspire that one day people will actually play through them, we have to remember that getting a game over is discouraging. This issue arises a lot and I think it's better to just be forward about this. A playable game is a fun game.
However, there are boundaries on the other side of the spectrum that definitely need to be considered. Are you offering a challenge to the player? If not, they probably won't be having much fun unless they are enjoying the storyline. Gameplay has a much more immediate and direct influence on a player, I feel. A person playing a game might give the game a chance if at the very least it is playable and the storyline isn't too great. However, good games often go unplayed because of a simple design mistake. Don't be a victim!
Another pitfall that really stinks as a player playing your game is the undirected approach. In a lot of games--both commercial and amateur--you might find yourself not knowing what to do. Things to consider: people often miss subtleties in dialogue. It's important to be very direct about what your intentions are for your player as a storyteller. Make sure players know what they have to do or it isn't too obscure. Not remembering what to do and not having any sort of indicator or person to tell you what to do can be very distracting. Sometimes certain things need to be said twice, because often we read things so fast they go overlooked.
Also, keep in mind, the player is not omniscient. By that I mean, if you tell your player to go to someone's house, remember that they don't know the layout of your village as well as you, the creator, do. This I consider a game over because you basically are stuck and if you don't get any sort of tips soon chances are you won't play much longer. Be obvious with this. I've yet to play a game that wasn't too obvious what to do to the point where it really took away from the experience.
These are two small aspects I'm always taking notice of more and more, because as a game gets longer in some ways it may be difficult to keep them playing if you've made a poor design decision, and if you do it's definitely not the end of the world. Off the top of my head I can think of three decisions in my mind I will return to and fix simply to avoid a game over or two. People will like your game if it doesn't piss them off. I know it's obvious but think about it.