Another suggestion to determine character stats

[Bonifarzia]

I notice there is a lot of discussion of how the stats system (str, agi, end, int wil and cha) could be (re-)implemented, be it about character creation, final limits and plasticity, the process of gaining training and practice, correlations and synergies, prerequisites for or benefits from skills etc. etc.
Some of these threads have very good approaches, but there is little consensus.

As this is an interesting topic, I have put a bit of thought in this and came to only one solution that seems convincingly simple and yet leads to some effects that seem to be desired according to the various discussions. So allow me to put this here for discussion and excuse if a very similar thing has already been subject to older threads I could not find here.

For now we have a profile of base stats for each race, some additive contributions from character creation and a process of training that boils down to simply paying some tria and PP for reaching the same hard limits on each stat.

My suggestion is to establish a construct where each implemented skill has its own profile of stats, similar to base values for each race. Whenever the character gains another skill rank, his stats are evaluated as a weighted average over the profiles of the race and all trained skills. The base values from the characters race may have a weight of 100 and each skill will have a weight equal to the trained rank. (Just to explicitly define weighted average: The players profile of stats is then equal to the weighted sum over all profiles, divided by the sum of weights.)

In the way I express myself, this might sound fairly complicated at first. But think about it, doesn't this seem an intuitive and simple way to couple stats with skills?

Obviously, it is not a simple task to define well balanced profiles for each skill, but it would allow for a lot of flexibility and introduce some natural concepts of progressive, practice based training, synergies between certain skills and stats, as well as relative limitations to opposing or contradictory skills. Training a certain skill may improve some stats  while degrading others, for which the character currently has higher values than those associated with that trained skill. In return, some stats can be useful when gaining practice for skills. If we assume significantly higher stat values on the skill profiles than on the races base stats, the latter will act as a blind count. This means that training a larger number of skills can enhance your overall stats, but the benefit will get indeed smaller and smaller with each additional skill. If a character wishes to have a highly specialized stats profile, he, she or kra will have to stick to a suitable set of skills. Someone who rather tries to learn a broad variety of skills will have well averaged stats, but none of them will be exceptional. One problem I see with this is the current implementation of character creation. With the proposed system, character creation should be limited to adding starting values for skill ranks. If boons on stats were added, one would have to take into account a profile of starting stats for each character, which seems highly impractical and encourages specific choices of character creation.

Thanks for your feedback...

[Sen]

An interesting idea that could also play a part in the master of all trades-problem. But I think there should be some way to do stats only training. For example would someone who runs a lot have better endurance or someone else plays whatever role that would lead to a musuclar body without having a special profession.

[Bonifarzia]

Agreed. Notice that it takes much less effort to gain only a few ranks in three or four skills than to reach a three or four times higher rank in a single skill. Yet, in terms of increasing averages, the benefit will be the same. Thus this simple concept does not exclude training strategies focused on the development of certain stats. However, the player needs to have some a priori knowledge about the effect of a certain skill on all stats.

[verden]

Perhaps I misunderstood, but it sounds like simply a different training system. Why not just have a character creation interface where a limited number of points were allotted to various stats and skills when the character was made?

[Bonifarzia]

I would not call this a different training system, because it does not imply anything about how progression is made for skills. The focus is on the definition of stat development and for this, we do not really have a training system yet, I would say.
Speaking of the allocation of stats and skills during character creation, which seems a commonly used concept to determine a characters development for a lifetime .. I am not a friend of such systems, because they lack an important feature: Plasticity.
But maybe I misunderstood you here. If you mean an allocation of starting values, its basically what we have right now, and all that would need to be changed for this to fit with my suggestion, is to allocate starting skills only and have those determine starting stats.

[verden]

How do you define plasticity in relation to a character generation system? But you are saying that a pool of points would be used to build skills, and then the skill choices would determine stats? Its an interesting approach, it could be tried.

[Bonifarzia]

To answer your question, I have to distinguish clearly between character creation (generation, choice of initial values) and progression (development of character attributes over a longer period of time). If the former determines the latter and no further significant choices can be made, we have the system I disapprove and hope you do not refer to. With plasticity I mean the opposite, the characters freedom to adapt to new situations and evolve dynamically, as it is the case with Plane Shift, even though skills or stats cannot degrade.

To come back to the topic and clarify things once more: The key of this suggestion is the mathematical approach to handle a dynamic evolution of character stats such that they reflect the life history of practice based skill training, where the races base values can be raised asymptotically and opposing contributions are averaged out.

[Nivm]

 I'll just tack this on here.

Skill Atrophy; it helps even if you hate it.
(Predit: please remove the "but if we hate it isn't good" from the title, I don't mean it that way yet can't think of something else differentiating.)

 I found four threads on the matter in this subforum, but they are diffuse in subject and at least three years old ("I'm this many!"). So I wanted to make one that compiles all points raised, at least on this aspect of skills. There are a whole lot of other skill points (ignore pun).
 The concept is that one's skills (and possibly stats) will atrophy (decrease), when left fallow, for the purpose of rounding out the game.

• When will it happen? - Some want it to happen all the time, but a majority want it to only happen when you are logged in; this makes sense, since your character effectively ceases to exist when logged out. It should be a slow process; in real life, if you spend a whole day learning archery, it takes a ~month for that skill to degrade noticeably, a few years for the archer to forget, and decades for the feel to completely vanish (if ever). This needs to be converted to game-time based on skill increase times; we also need to define just when the timer (or each skill timer) is active (such as times where the player can't counteract it).
• How much will it happen? - After an in-game month of being logged in, a skill will start losing practical experience just barely (it will take you down a level, but the instant you use the skill again the level will be back), and this will be your early warning. As more IG months pass by, the rate of practical experience loss will slowly increase. After a couple years, you'll be probably be losing multiple levels. A year after that, you will start losing theoretical knowledge, the rate of which will also slowly increase. However, the rate for practical experience will level out and reduce to nothing, leaving a certain amount permanently, but the theoretical knowledge rate will just keep increasing. The rate is cut down every time you use the skill; but it wont go back down to zero with just one use if it's high enough. Those who have put more work into the skill will have an advantage, because the atrophy isn't percentile; the rate is based on time and a constant.
• And skill synergy? - Using skills similar to the ones you're not using will slow the atrophy, and possibly even start gaining levels. This is somewhat broad topic.
• Why? - Creating procedural professions; further encouraging people to focus on certain jobs. It will cut down the "jack-of-all-trades" problem people have found cheapening. Good fighters will be those who focus on being good fighters, being the great or greatest of something will mean more. I recall their were other problems this solved as well.
   Procedural professions; further encouragement for people to focus on certain jobs. Newbies don't like Super-People.
• But why?! - You might not like to lose things you've spent time to get, almost none do, but this will solve more problems than it creates. Besides, if you want to keep those skills, it's not hard or annoying to just use them again every once in while. You might even visit people you usually don't, or see that something's changed (or be enraged that it hasn't).

I found it kinda funny that any old thread mentioning atrophy was locked, but that's mostly because of the vehement arguments about levels in general. Or perhaps because some mod locks threads when they get old.

 I'd like to know exactly what I missed here, and hopefully those who influence or control the game mechanics do to.

[RlyDontKnow]

how does that one fit in here Nivm? it's a completely different suggestion...

as for the original topic: I think we all agree that stats shouldn't be trained they are trained right now and the basic idea your brought up in here bonifarzia is how it should be imo.
however as you already outlined: it's not exactly trivial to balance such a system.

also there is still quite some way to go on the engine side to make it work at all as some parts of the skill system are still hardcoded C++ instead of scripts which would make things a lot easier (i.e. right now practice points given for a skill are hardcoded - for such a system it'd be a lot better to make this go to scripts instead which are invoked upon using the skill).

anyway, if you truly want to make this true, feel free to eleborate this idea more
also you could apply for the rules team (maybe as contributor for now?) if you want to get into this area more and want to help us make things better - I'm sure it'd be appreciated.

greetings,
RlyDontKnow

[Bonifarzia]

how does that one fit in here Nivm? it's a completely different suggestion...

An interesting idea that could also play a part in the master of all trades-problem.

Indeed, an entirely different subject and probably much more effective for the problem Sen mentioned than my proposal. But my idea is rather independent from that, the two concepts would not hurt each other.

as for the original topic: I think we all agree that stats shouldn't be trained they are trained right now and the basic idea your brought up in here bonifarzia is how it should be imo.
however as you already outlined: it's not exactly trivial to balance such a system.
...

Thanks for this helpful comment, RlyDontKnow. It is good to have some feedback on the technical aspects as well, for which i thought there would not be as many difficulties as with more sophisticated systems. I will think about your suggestion, as the development of new rules is an interesting subject, but I am well aware of the requirements to actually contribute something -- I just noticed that page got updated. Interesting, I see the rules department only on the diagram, but I assume it has the same prerequisites as the settings department. Mhmm, now it is me who goes off topic. (I will probably send the one or the other PM or Email when I think I do have something to add at this point.)

[Mrokii]

I support Bonifarzias idea 100%.

If I remember right, PS should be a game where RP is encouraged. But the current setting doesn't encourage that in any way (other than making training most anything hellishly painful and boring).

The current system does nothing at all to prevent powerlevellers. *Any* kind of limitation to that (so not somebody can not become an expert in everything at the same time) would help, and I think Bonfarzias idea is flexible enough. As I understand it, it would allow players to specialise in a few areas without any restrictions, and it would also make it possible for them to change their mind, so that they could become experts in other areas. This would involve "natural" degrading of other skills or abilities, depending on how much somebody would train "new" abilities.

I think this idea is great for several reasons:

1) It would encourage people to cooperate much more than the current system (simply because they *can't* be expert in everything, so they need the help from others).

2) I would help individualising players tremendously. I would combine Bonifarzias idea with a tab in a players' description-window (like the current CC-tab for example) in which other players could read about the characters' abilities. It could contain things like "This individual is highly proficient in playing musical instrument" or "This individual has only basic mining abilities". There could be lots of these "virtual" levels of expertise, all adding to the individuality of players. There is certainly a point that these abilities may not be obvious to other players (they wouldn't "see" that others are good miners). But this could be circumvented with more "general" descriptions, maybe something like: "The marks on his/her hands leave the impression, that he/she is an avid miner". This in turn may lead to shorter descriptions for players as they wouldn't have to describe their level of expertise (as some do).

[Bonifarzia]

This concept is an idea to define stats based on skill training and racial traits, not less and not more. It would not make it impossible to become a master of all trades, at least in terms of skills. But as we learned from the raised caps for stats, having certain statistics at a  value of 200 or 400 certainly makes a difference. This could be exactly the difference between broad or highly specialized choices of skill patterns. About information in character descriptions, I would say everyone should be free to choose what shall be stated there. Be it with writing a description or making use of myplane. That does not mean that I consider the "appears to be" phrases useful and reasonable.
I think Mrokii's first point is very important. Maybe not in context of this subject here, but in general for Planeshift, and in particular for the progression system. It would be wonderful if one day the dull grinding business could be replaced by something that encourages cooperative gameplay in an entertaining way. But now I am going off topic here.

Thanks for your support, Mrokii.