In all Rpgs characters are constructed with attributes. These attributes can be anything from Intelligence and Strength to Sword Fighting or First aid. I can see that using a generic experience system allows newcomers to the game to adapt to the system very quickly. The hard bit in coming up with new XP systems that make the game better without straying to far from the norm.

After some quick thinking myself I can see a way the XP system can be improved whilst maintaining a way that would still allow new players to enjoy a fast learning curve. Hopefully I can communicate this idea as clear as it is in my head. Well here goesc

First, leave player statistics (Str, Int, Com) alone. Donft improve them, If they must be improved make them only minor improvements or temporary. Now this takes away a level a game complexity, as players donft have to watch these all day long with concern.

Secondly, bring the players concern about character improvement into issues that they know will effect their character. Examples would be Swordmastery, Healing, Magic (defense), Magic (aggressive) etc.

Herefs an idea how the mechanicfs could work for this new system:
1. Characters have Awareness points that they must invest into a Skill (Sword Mastery) or Skills.
2. When XP is awarded it is divided up into the skills where the character has implaced the Awareness Points (APfs).

Using this system a player can easily gDesignh there own character as the Character progresses. A player can move the APfs around at will whenever they want. Also using this system, Character Progression Levels can be removed entirely. Personally I hate Character Levels as I think they make Character Comparisons to easy. Meaning all you want to do all day is get to the next level which is...well, very boring!!

If we continue to make RPGfs that revolve simply about level upping then this genre of game will die if not through boredom then by being taken over and absorbed by other genrefs of games like Action, Adventure or mindless shooters. What you say?

I haven''t read many of the posts on this topic so stop reading if this is repeated.
I like what Planescape:Torment did for their experience levels. If wasn''t kill kill kill to get up levels (although i guess if you did enough you could), but i would say that 60-70% or more of experience points came from completing side quests or obtaining certain information.

There was also a negative side effect to all the killing(if this is what you did), the high sorceress(or whatever) of the city the story takes place in, was the baddest being in all the known universe. If you killed to many people/beasts/whatever inside town, she would come and whip your a** with no chance of beating her. I''m not saying that she was too tough for me to beat, I mean literally, there was no way to beat her.

So you could have where if the player killed to much, punish him/her.


"Now go away or I shall taunt you a second time"
- Monty Python and the Holy Grail

OK, some good ideas here, some bad.

First of all, you learn a lot from succeeding, potentially more than you learn from failing. Learning how to swing a sword or hammer (smithing) is as much a ''feel'' thing that must be learned from repeated success as it is learning from your failures, if not moreso. Muscle memory for balance, combat and many other activities is very important. A good smith can tell if he''s doing it right just by the feel of the hammer hitting the metal, he can tell if the metal is too soft, too hard, everything. A warrior can tell if his sword is glancing off the armor that he''s trying to get through or if it''s cutting true. While he will learn if he''s succeeding or not and potentially learning from failure, he gained this ability to even figure it out by being successful for so long and consistently.

I''m more of a person that dislikes the leveling systems, not because they are stupid, but that they generally affect the wrong things. Levels (as they are implemented in most table top RPGs and CRPGs) should be only good for one thing, adding hit points. I''m developing a system where all skill usages are relevant only to the skill and you can become a complete swordsmaster only through training (though it''s not easy to do without combat) and all that engaging in combat really does that can''t be done elsewhere is add hit points.

There is so much more to actual fighting experience that can''t be taught elsewhere and that''s all relevant to system shock. A combat experienced swordsman will in all likeliness beat a swordsmaster with no combat experience and not because he''ll know more about how to use the sword because he won''t. It will be because he''s willing to take a hit that the master swordsman will give that won''t be fatal to get in a fatal blow against the master swordsman. The master swordsman will then in all likelihood go into shock at least for a moment and that will give the combat experienced warrior the time to finish him off without regard for skill. The combat experienced individual knows that a sword slash isn''t the end, it''s just another wound that he''ll have to recover from, while the inexperienced master doesn''t have that actual experience to draw from that says that the wounds aren''t all fatal. He''s going to panic for a moment and try to look at or grab for the wound (it''s only natural) and that will be his undoing.

Anyway, a wizard who never fights in actual combat but supports the fighting will also have similar experiences, but they won''t be quite as strong and he won''t have that ability to ignore significant wounds the way that a more experienced warrior could. However, most experienced combat wizards have been in close-in fighting and will understand that wounds aren''t the end of the world from their experience. As a result, I will give combat experience (toward more hit points) based upon the amount of damage that a character takes. Yes, it will be more beneficial to take more damage and it can be abused, but every system can be abused to some extent. The danger of getting killed (and getting no exp from the creature, you only get exp if you survive and win) should deter this to some extent. You also won''t be able to gain experience from creatures that you significantly outclass in skill. They aren''t a threat and you won''t be as concerned about what is happening in the combat as you would be from a creature that is closer to your experience level (not level in the traditional sense, more like the way 2 network administrators should be at the same skill level).

In other words, you won''t get better baking by swinging a sword and just because you swing a sword a lot and well doesn''t mean that you''ll be able to take as much damage as someone who is not even necessarily as skilled as you are. You might be able to wear them down if you''re exceptionally good and cautious, but it would be a real fight where you could lose at any moment due to a bad move. You could have a sword skill of 500 and get beaten by an individual with a skill of 300 because of the difference between combat experience.

Also, a good system would force people to choose a profession, as it will take nearly as long to become a master armorer/weaponsmith as it would to become a master swordsman. You can create alternate characters to do the different things, but they would each take large amounts of time to get up in skill and avoid the ''power character'' syndrome where people get every skill maxxed out that they can (like in EQ).

Call Of Cthulhu, a system based on H P Lovecraft''s Cthulhu stories uses a system where skills are expressed as percentages, representing chance of success, and significant successful use of a skill earns a ''tick'' for that skill. When characters "level-up" (by which I mean those points in a pen-and-paper gaming session where experience would be distributed and level increases processed or equivalent for other systems eg Shadowrun Karma distribution), any skill with a tick trades it in for a chance of improvement - roll a skill check and, on a failing roll, the skill is increased. Note that this system has several of the qualities discussed by people above - it makes improvement less likely for skilled characters, it only improves skills that have been used, and it only rewards "significant" skill uses, though that does rely on the ability to recognise significance.

About rewarding success vs rewarding failure: I disagree that it''s clear cut that failure teaches more than success. To give an exaggerated example: if you want to turn on a computer, pressing the power button is going to teach you a lot more about how to do it than any amount of typing on the keyboard or maving the mouse around. In fact, in real life, succeeding in something for the first time teaches you a large amount about how to do it - and in more complex situations can include skipping over several steps that learning by failure would take you through. Of course, repeating the same solution teaches you very little, so repetition beyond the point of mastering a given skill shouldn''t provide bonuses. The way people learn in real life is by trying new/different things. When I tried 10-pin bowling for the first time, I kept knocking over the three pins on my right. When I tried changing the way I was throwing the ball, I got very variable results, and could probably have kept trying and failing for a long time. Then I tried one thing different - rather than changing how I threw the ball, instead I moved my feet, turned myself slightly to the left and went back to the old three-pin motions, and suddenly I was getting 8-10 pins every time... Now if you look at my situation in terms of my abstract "bowling skill" (measured in terms of number of pins reliably knocked over) I was initially at skill level 3, with little or no variation in outcome, then when I started experimenting, my basic skill level effectively dropped to 0 or 1, but with a large range of actual outcomes, before jumping suddenly to a stable 8 or 9 again with very little variation. But my increase in skill, while it required my temporary drop in skill level while I experimented relied not on my failures, but on my success - the moment of change in skill level was dependent on when I succeeded, and my change in skill level wouldn''t have been affected if I''d spent months experimenting with variable bowling or if I''d just tried moving my feet to a different angle to start with. Any skill system that purports to be based on real life should grant improvements based on successes more than on failures, but not on "sure thing" success or failure, but on unexpected successes and failures - and should probably allow players to choose between having given skills stable or unstable - stable skills don''t improve their average result over time, instead becoming more reliable/predictable (so a simple implementation would require the random element of a skill check to average close to 0, and just have its weight reduced with skill use), while an unstable skill - one where the character is trying new things - should sacrifice some base level, and have less reliability, but in return, any successes should modify the base level (down as well as up - playing something like Half-Life on easy after geting the hang of it (but not mastery) on hard tends to encourage bad habits and make playing it on hard again that much harder as you get used to being able to get away with various things)

Actually, the system I''ve just invented sounds like it could work pretty well - skills have both a base level and a randomness level. Skill checks are determined by a formula similar to: [success] if [threshhold (including modifiers)]<[base]+[randomness level]*[random number] with [random number] averaging zero. Players can choose whether practice goes into reducing the randomness, or changing the base level. If the randomness is being changed, every use of the skill reduces it a fixed amount (possibly distinguishing success and failure). If the base level is being changed, an immediate (temporary) negative modifier is applied and the randomness level is (temporarily) set to maximum. Any successful outcome modifies the base level (permanently) by a fixed proportion of the random portion of the result ([randomness level]*[random number]). When a player switches back to changing the randomness factor, the base skill should lose the temporary modifier, and the randomness level should revert to its previous value (possibly over a number of skill uses, but more rapidly than the normal rate of change) since the previous practice should give at least some bonus to the new techniques within the skill.

This system obviously discourages genocide as a means to character advancement since, while it does offer a chance to reduce randomness in skill checks, any attempts to actually improve base skills by trivial use will on average do nothing, and some of the time actually reduce skill level.

I''m with Keith. Can an admin lock this thread? Three pages is too much to catch up on. Let this one die, and start up a new one. Many of these ideas could be resurrected, but let them be references. This is too much to post to.

I think that a game with a good EXP system is Bahamut Lagoon (a tactical RPG). Characters get rewarded for any action that helps their team, whether it be healing party members, casting a movement spell, attacking from far away, or actually getting into battle. Sorry if I am missing a major point, I read to the half of the second page, and then just got bored and wanted to make my post.

I disagree I''ve been reading this thread for the past 30 minutes, although I admit I haven''t nerely digested all the ideas presented. I feel threads of this calibur shouldn''t be let die. If we lock it, it gets buried in the forums and others intrested in the subject will probably never find it.

I for one am happy someone dug this up.