Warning: A bit rough around the edges, and rather long. Still, I would appreciate any feedback.
Humans learn by making mistakes. It has been said "a fool learns from his own mistakes; a wise man learns from the mistakes of others." Through non-interactive medium we have been able to learn from the mistakes of others, whether they be real people or fictional characters. With the rise of video games as a truly interactive medium, we are able to create safe spaces for experimentation, wherein the player can learn from his own mistakes, without having to suffer the consequences. Thus, the player's dead self becomes his own "other." This cycle of failures and second chances allows the us to advance and become better gamers. If the skills we learn can be applied to the real world, they can make us better humans as well. If this cycle of failure is so important to our success, why are we punished to much for making the "wrong decisions" in games?
Even in branching RPGs, there is a conception of what is "wrong" and what is "right." The difference between these games and those which are older is simply that the newer generation of games leaves the definition of "wrong" and "right" up to the player, rather than having it be something which the designer predetermines. We as players will keep multiple saves so that we can access different routes of the story if we don't like the decisions we have made which took us there. Why shouldn't we have this power? It's something which the rules of multiple save files implicitly dictate, if one is only willing to read between the lines. Game developers obviously know this, because they limit the number of saves even of console games which use memory cards for storage, or on the PC where the only limit is the amount of hard drive space one has. This is a much higher limit than what was previously possible with only simple cartridges. There is no real reason to have these artificial limits anymore.
Except they could break the game.
But games like Radiant Historia, and Braid which came prior, are beginning to reanalyze the ability to redo one's mistakes. In Braid, it is not necessarily to keep multiple files because you can never get a "game over," there are no dead ends. It should be conceded that the Linux version (actually I think this applies to all the PC versions of the game) has a glitch which prevents the player from getting the "true ending." However, that is a design flaw. I am fairly certain that the designers did not intend for that to happen. It was simply something they overlooked while play testing. They knew there was a secret ending, so when they tested they tried to get it. Ignoring that exception, there is no death, no final screw up, no irreversible failure. You simply learn from the missed jump, the bed of spikes you overlooked, the level you didn't throw, and try again.
In the case of Radiant Historia, failure is actually beneficial to your progress. There are certain "time nodes" that you can revisit again and again by using the power of the White Chronicle to travel back in time. Some of these nodes, especially the major ones, have branching paths. These branches in the storyline are made by triggering other events in the past, or making key conversation decisions, or a combination of the two. However, sometimes one run though the timeline is not enough to make the correct decision. Occasionally, the player will be forced to hit a "dead end," warp back in time, and then try again. The new information gained in this way will unlock a new conversation option, which allows for further progress through the story. Other "dead ends" give the player more information, but do not provide additional information to the character, Stocke. These branches are still helpful, as they assist the player in locating potential spots where the facts don't add up, and more investigation needs to be done. This indirectly points the player in the right direction, like a good mystery should.'
The time travel mechanic in Radiant Historia thus allows for the player to see how the story changes based on the various decisions made throughout the course of the game, but does not penalize the player for making the "wrong" choice. Even though your mission is to save the world, actions which cause the world to end are still critical to understanding the motivations of characters, and the importance of the relationships between them.
Another area where relationships between different aspects over the course of a game is important is the concept of the skill tree. Who is to know, at the onset of a game, which skills are good and which ones are bad? In Deus Ex: Human Revolution, the cloaking seems like a great upgrade. Likewise, so does double takedown. Simply the name alone is evocative of the double hidden blade executions found in Assassin's Creed 2. It seem like the two together would be even better. However, the recharge system on the the skills limits the times when you can use the two abilities in tandem. Additionally, while the game emphasis choice, there is occasionally a boss fight which can not be avoided. These bosses must be defeated using direct combat skills. There's no room to run, to try to lose your enemy in a maze and circle around, acting as a silent killer. Suddenly, the player is forced to fight head on. No way out but though. But how is one to know this from playing the opening level? There's no indication. A skill reset would be nice, but none are forthcoming. The player could start over, but that would be cumbersome. If only you could redo the parts you didn't like, and keep the parts you did.
Does this seem too idealistic?
And yet that's exactly what I feel I am getting with Radiant Historia's decision system. There's endless freedom to jump back in time and try new conversation options, or simply to relive the cutscenes or battles that you particularly enjoy. But the traditional skill system has none of this. It's straightforward. Don't like your character? Start over. When experimentation is so key to sucess, why stifle it in this manner? Modern game designers themselves realize the value of iterating on an idea until the best possible form is achieved. Why then, would we deny this ability to the players of our games?
As I said before, because it would break them.
At least, that's our fear. It doesn't have to be that way. Simply allow players to respec for free, but not in combat. Allow various different augments to be collected and on hand, but only allow a fraction of them to be enabled at once. MegaMan BattleNetwork was great at this. You had the freedom to find and buy a variety of different upgrades for megaman, which could alter his health, attack speed, damage, and buster charging ability. On top of this, the game had various cards which held power ups to grant mega man a multitude of different bonus powers in combat. Some of these could even be combined for even more devastating abilities. You were free to try various combinations until you found the one which best fit your play style. This was only further improved upon in later games, with more in depth methods of tweaking these attributes, as well as adding a few more things to customize your play experience. Layers of these customization unraveled over the course of the game.
The problem of strict progression seems to be most prevalent in games with richer stories. This does not mean that rich stories are bad. In fact, I think MMBN has a pretty rich story, but it's still an excellent game series. The problem lies in our stems from the hero stories of non-interactive media like books and film. While these stories are by no means bad either, they focus on a set character (or group of characters) which advance in a set pattern over time. In non-interactive media, the creator knows what is best for each character, and can thus assign correct powers to each one. But in an interactive art form, the creator must take the player's feedback into account. This is similar to a director accepting feedback from the actors. However, game designers can't simply ask the player directly what sort of person he thinks his character should be, so that the audience will respond properly to the character.
This is because the audience is the actor. There's no time to "practice" a game beforehand the way an actor can, so that you can make the experience of playing seem cooler to yourself. The performance is the play itself. Remember that actors and directors must work together to make a good movie. So too must the designer work with the player to create an immersive and enjoyable play experience. Fan feedback is helpful for successive versions, but is not fast enough. The designer must allow the player to iterate on his or her own and find what works best. Let us then re-think balance while including the player experience as well.
Given a multitude of options, skill points are not the only factor in determining the depth and breadth of skills a player can acquire. There are real-world factors, such as the time the player must spend to get those points, which must be considered. Minecraft is fairly good at this, making the more powerful materials take more time to acquire. Even if the player were allowed to access a wide variety of skills over the course of trying things out, he would most likely not use all of them in combat. Some would be simply forgotten, or they may not fit on the hotkey bar or analogous control mechanism. Thus, the player's physical skill itself limits what can be done, not just the skill of the player character. It is possible that people like the professional Starcraft players will arise, who are capable of playing so well, or so rapidly, that they can do things no other players can do. These are merely an edge case. At any rate, the highly competitive will create additional rules imposed by the metagame, or by outside parties, which limit the ruleset for their style of play. This is not to say that balance should be ignored, but simply that a well-balanced system will not break simply because there are players capable of playing extremely well.
With this is mind, perhaps a good skill system should only limit modification in higher levels of play? Only when the player is clearly locked into one style of play, one way of doing things. But if games are to be art, art must challenge our thinking. This method will only serve to make the player's thinking stagnant. What if a new strategy is developed which shakes the metagame? Should the player be forced to start over to radically modify his or her character?
These are the sorts of things I consider as I make games.