I''m picturing a party watching a fellow thief trying to make his way across a thin beam so he can lower a drawbridge for the party. Halfway across, the player stops paying attention and the ball slips to the left. The party watches in horror as the 3D figure is bending backward frantically waving his arms, one leg in the air. Then the player finishes sipping his Jolt, looks at the screen, drops a brick and starts tapping the right arrow. The party breathes a sigh of relief as the thief seemingly regains his balance and continues on his way.

Interesting.

So I''m assuming then that a high balance skill will keep the ball more stable. I''m thinking that tapping the arrow keys should not be implemented, since those faster on the arrows with their fingers would do better. Simply holding the arrow would apply the balance skill to the ball, releasing at the right time of course, and if the challenge is too great (high wind, earthquake, etc.) and the balance skill isn''t high enough, the character falls. It also keeps those boring ol'' feet on the ground types from getting across to the rewarding spots that those with balance can reach.

I don''t think applying balance should be all that hard. It shouldn''t take too much reflex, just attention. It does add that extra dimension, though, and adds a tactileness to the need for balance skill. Those with enough balance could even let the ball slide a little as they fire a bow at a passing enemy and continue on their way. A low balance score wouldn''t have the time to knock the arrow and fire before the ball sways too much. Besides, after the arrow fires, that ball is going to shoot one way or another. (Imagine a player firing arrows to keep his balance, making his way across a rope - funny picture)

Another vote for the idea.

I also like the ideas for allowing the player to bypass a mini-game if the skill is high enough. Maybe combine it Waverider''s idea of stopping in the middle of the balance game to fire an arrow. So, if you were really good, you might be able to just cross the narrow beam/rope, or you could do something else while crossing, but might have to worry about keeping your balance at that point. A high lockpicking skill could just be a "jiggle-jiggle click" - sort of like thief, maybe even including multiple lockpicks, while a low lockpicking skill could involve a more difficult mini-game (guiding the correct pick to the correct pin).

As far as the "hacking" in SS2, I don''t think it was that bad, but it could have really used a boost in the player-side difficulty to make it more puzzle-like. If it was made so that there was only one or two safe ways to make it through, and a lot more danger squares, it could have worked. A highly skilled hacker probably could just hit nodes randomly with impunity, but a less skilled hacker would have had to spot the safe path

Anyone remember the lockpicking in Hillsfar? Perhaps some of the ideas could be utilized there (but broken picks *should* be replacable at a cost).

As for balance being a worthwhile skill, I think it is. Sort of a thief-acrobat kind of feeling. Combine it with something similiar to the rope arrows in Thief (but make it so that they can lead the rope horizontally and vertically - based on the angle that the arrow is fired), and it could be a really cool and useful skill.

A stealth skill could be something along the lines of the light meter in theif (and/or an improved version of the sound meter in Soldier of Fortune 2) - perhaps split into hide and move silently. No skill in these areas would provide no meters at all, and once the player takes some skill points, they get the meters - the more skilled they are - the more accurate the meters would be. Someone unskilled could try to find the appropriate places on their own, but might not catch everything about the lighting, floor material, etc - and would probably have to be more deliberate in their actions, rather than just having a meter turn dark when they were hidden.

Overall, I see this not so much determining success or failure in a skill, but more of being a distraction if other things are going on. That is, the challenges should be easy enough if someone can focus their attention on them, but when you''ve got guards firing on you, it takes someone with skill to pull it off. As the character becomes more skilled, the tasks take less time, and require less input from the player.

I like it

I would vote for mini games whose emphasis is related to the nature of the skill being represented.
So in the case of physical skills, use a skill based mini game like in your example.
To simulate an intellectual skill, use a more intellectual mini game, a Tic-Tac-toe ? a memory game ? IQ test type of problem ? Mastermind ?
Of course, it ould be quite challenging to represent some skills. I am thinking of strength for instance, it''s the old problem of making an athletics game without using the "type till your keyboard explodes" methods, how do you solve that one ?

The other point is the link between mini-game difficutly and character skills, but I guess that one isnt too hard to solve, simply make the difficulty of the games inversely proportional to the character''s skill (low skill, high difficulty), and to the action''s difficulty itself.
In this fashion you give the player who has a crap character a little chance of still making it.

I wonder how you could use that for combat itself (since combat is usually the meaty part of an engine, I consider it apart from the rest of the skills)...
I remember a cool game called Cobra Mission where the figthing was basically "click on the picture". The trick was simply to find the hot spot. A bit lame I agree, but it worked damn well for me... go figure.
Maybe a more complex but similar idea could work ?

--------------------------------
"millions to one chances usually occur one times out of ten"

I really like this balancing idea.

I just played a demo of some inline-skating game for the PS2 and I think it had some form of balancing (not sure what the maneuvre is called, but when you ride on just one foot, you have to press up/down to remain balanced).

So far, pressing arrow key left/right (or up/down) seems to be the accepted way of controlling balance, but I think you can make it much more interesting by using either a mouse or a joystick controller to do the balancing act (leaving arrow keys for movement), because that way, you have a vertical and horizontal balance control at once.

My preference though, would be a dual-joystick controller, where one joystick controls movement, while the other controls balance.

The joystick/mouse can control balance much more delicately. Move joystick/mouse a little bit to the left -> character''s center of gravity (because that''s what balancing is all about) moves to the left a little, perhaps by stretching the left arm out. Move joystick/mouse all the way to the left -> character''s center of gravity moves to the left as far as possible, perhaps by stretching arm out as far as possible, throwing left leg up in the air and curling the entire body to the left.

Ah well, I''m just a big fan of the dual-joystick

PS I vote for the idea of combining player and character skill in determining success and difficulty of balancing.