[...] Could you clarify that for me?
1. In dwarf fortress mode, I would thinkt that the warfs in my fort have seen most of the common things done in the dwarven culture thay come from. here migth be some {forbidden knowledge} (like a military secreat) that is not part of the common culture even thow it is a skill posessed by some of that culture.
Actually, when it's put in that context, you could also cover social castes with these mechanics as well. If alchemy is restricted to nobles, or many professions are restricted for a dwarf of a particular religion to have (historically, look at Jews) then you could declare that forbidden knowledge to those specific dwarves as well. Of course, you'd need legal repercussions as well, but it's another interesting use of this system.
2. With the need for training to advance a skill, I was thinking about the rule of diminising returns (DOR cap). So when I start off adansing a skill, my effort becomes less and less until I learn nothing new from using the skill, it has become rutine. Now, if I were to meet somone and talk about the skill, and have some training and perhaps some of his reflections from his use of the skill, I could start to anew to advance by using the skill. Thusly I can not trow dirt at children to get legendary at thowing dirt, and then go out and kill the hydra bu using my skill from trowing dirt at children.
3. there is some mood, variable. and the system knows if the dwarf feels this or that, so using a skill under non-common sircumstances sould fulfill my desiger here. I see this as a nother tier on the rule of diminising return from using a skill. I also think there sould be a limit on how much somone can benifit from training, so there is to use the skill in an uncommon manner that could give a boost and brake he diminising of return cap (DOR cap).
4. Some more history for creating art and stuff. could also give the ability to train others or oversee the training of others, could also give boost on the DOR cap
Am I right in thinking that each tag after #1 could be stacked multiple times, and that they'd have extra effects for doing so? If that's the case, how and and what rate do they wear off (if they do?)
2. I think 2 could be modeled just by having teaching as another source of experience and doesn't need to have special privileges over the DOR cap. Repeatedly teaching, after all, doesn't have any magical ability to avoid becoming uninformative or boring in the same way that repeating another activity does.
On the other hand, it is true that a teacher has the capacity to let you look at previous activities in new ways. I think that, as far as realism is concerned, it's equally viable to let it also have some additional effect on the DOR caps of non-teaching-related activities. I really don't think either view would be demonstrably wrong, so IMO it's mostly a question of game balance and heavy analysis can only be done after implementation.
3. It might be possible to simulate this by adding a "Not in combat/stressful situation" diminishing returns tag (and possibly weighting it multiple times.) Therefore, the more you practice outside of combat, the slower your skill gain is. And in combat, the influence of the tag completely vanishes. Actually, you might be able to do the teaching tag like this too.
4. That is interesting. Maybe appreciating/examining an object could reduce the DOR caps associated with crafting that object, and examing engravings depicting an item/process could do the same to any skill associated with that object? If that's the case, we could make a fortress renowned for its masons, and merely looking all the art would serve to also help to teach masons how to better hone their skills?
This would be an excellent example of something that shouldn't flat-out give experience to everyone, but still could serve to inspire craftsmen to greater heights.
favord skill, ok. here I was thinking that the Skill for example, Shield use could have several generic tags, like say hauling, arm use, hand to hand, and range combat. (or something) ok, now the Skill of shield use has a favorit generic tag like for example arm use, thus using the skill benefits more from the tag arm use then any of the other generic tags. (the bottom line, the favorit thing is not for the dwarf but for the skill it selfe)
I misinterpreted this (and a lot of other things in your post); my apologies. I was thinking this would be something like Fallout's tagged skills, something that just forced a bit of complexity into the skill system without doing much more.
Hmm... this is an interesting idea. I'm still not sure what to make of it. I'll think about it, try to see if I can see exactly what it entails.
I liek the idea that the personality models cound infuence a happy or sad thougth by using a skill that is opposide the traid.
My conception of that was that personalities would affect how effectively a dwarf was working at the task. If happiness affected how effectively dwarves worked, that would fulfill both of the ideas and have a few other benefits too.
Grinding seems to have a simple solution: give each task a difficulty level (maybe equal to chance of success), and give experience based an difficulty and current skill. If a novice thrower throws a stone to an empty place several steps away, he will have some experience; but if a legendary thrower does the same he will get none.
That does make grinding harder, but that can just be solved by switching to repeating a task of the higher difficulty. For example, once you're getting 0 experience for throwing a rock at a tree 5 feet away, move to 20 feet away to get experience again.
One problem with it is that it would also make legitimate skill gain harder to do as well. Instead of needing to simply find a place to use the skill, you'd also need to make sure that it was a level-appropriate challenge.
Hmm. I'm looking at the problem from a slightly different angle. I wasn't looking for making the game more interesting, but instead more realistic. So in RL you can train by repeating something over and over, but it must be at appropriate difficulty level. Look at the sport for example: professional sportsmen train many hours a day every day, and there is no other way to win.
To repeat something that you obviously already know, the basic problem is that it's boring for the player. Jumping into adventure mode to do 100 push-ups before your next dragon hunt is dull, but if the skill advancement system is to be kept similar to the one used when doing a quest you can't just let the player automate doing it (because otherwise they'll spend a few in-game weeks to max out every skill in the game.)
Skill advancement is an additional carrot to make doing anything skill-base more interesting in most RPGs. It's possible for Dwarf Fortress to shy away from that trend, but I'd assume that Toady would still keep to a system that lets you easily upgrade skills as a result of doing stuff that you're intended to be doing (like killing monsters, making useful items, or whatever have you - I haven't seen any direct quotes from Toady that counteract my assumptions.)
One idea is that there would be another form of skill gain to be used when waiting/doing things over long periods of time. This kind of training would take months and be less effective than learning skills under stressful situations, but NPCs could use it too and it would save you from doing nothing over that time. This would be a good bit more realistic than the current system, but I wouldn't expect that it should replace the standard system entirely (because being a virile youth beating up dragons would be pretty hard to do otherwise.)
It would require some way to wait/ do long-term things, but I really can't foresee Adventure Mode working without that, so this could be simply tossed on top when it would be added.
Using skill when needed (for a quest) doesn't always mean it was educational. If a king asks a legendary hammerer to punish a peasant it doesn't mean that the hammerer will learn to fight better.
Agreed. One big problem that any advancement system has is in determining which activities are educational which which aren't, something which is really hard to determine and frequently unreliable. You might figure something out by repeating an activity that you've done before in a different way; I don't see how that could be modeled well. The best idea I've got is adding a small random chance to earn an amount of extra experience (which does very little good to it from the perspective of a game).
I'd personally argue that a lot of these problems stem from storing knowledge and everything else that goes into a skill as a simple integer rather than something that behaves more like a web of concepts and mental/physical conditioning (muscle memory triggers). Of course, modeling that for one skill is a challenge because you'd need pretty specific ideas for each skill to be based from; doing that with the sheer amount of skills Dwarf Fortress has would take a long time.
As for DOR cap, I like the idea of having to do several different activities to advance. I once thought about a more simple idea of having 'theory' and 'practice' level in each skill, first is gained by talking and reading (maybe teaching/writing on higher levels) and the second by using the skill; when there is a gap the progress of the higher is lowered (down to zero when the gap is big enough).
It's complicated and slightly more realistic, it just seems to me that you'd need to synchronize the two extensively, which would force a lot of micromanagement on the part of the skill leveler and on the game designer. You could tweak it to the point that it achieves a nice balance and doesn't force the player to go through lots of hoops just to get better at a skill, but IMO the gain in realism would be minor, especially compared to the amount of work required. You can't completely separate the theory from the practice in almost any skill, and as a result the implementation will still reflect that basic problem regardless of what you do.
I think your suggestion is great for a full rpg game. The problem is that DF has its fortress mode (which is much more important for me), and it seems hard to implement such system for a whole fort of ai-driven creatures.
I'm aware of that, and that Toady doesn't want disparate systems for modes if he can avoid it. But I don't think that it will be a significant problem.
One of the most important ideas here is to keep it automated (or easily capable of automation). The demands on the hardware for a complex skill system are insignificant compared to that required for, say, pathfinding. If no choices are required to be made for the dwarf when leveling up skills, there won't be a problem in that regard in switching it to Fortress Mode.
The other major problematic notion is that advancing skills will require/encourage special behavior from the dwarves. This could be pretty nasty on the skill implementations requiring training, goals, or other specific behavior.
I'd guess that the goal set required by the Hybrid Anti-Grinding System could be comparatively easily designed that this wouldn't be a huge problem (things like fulfilling mandates, maybe making a masterwork, definitely making an artifact, killing a named creature) could all be goals that trigger the experience allocation, and you could just consider a goal to be automatically achieved once a year so that the dwarf churning out endless identical mugs or shoveling manure still ends up learning a few things from it (just not too much.)
It'd be harder for the Goal-Based Skill Uses idea, I think. After all, there's not much difference mechanically between building a megaproject and repeatedly building and deconstructing a wall that the game recognizes.
As for having to wait before repeating the task, it seems that it would be good together with some other changes, not alone. BTW maybe skill rusting will force you to do multiple activities during the day.
It might be; it's likely that a hybridized approach to many of these ideas mind end up proving to be the best one.
And I really hope that skill rusting doesn't work that quickly. If it does, then taking a month-long journey to another town, say, would be likely to completely mess up your skills.
...So an "optimum" training regimen might be:
jog, swim, throw, juggle, press weights, run stairs, rest, lunch, weapon kata, haul wood, wind sprints, pump operating... repeat.
That is: you might still be 'grinding' at this point, but it can be more realistic and more interesting for the keyboard warrior involved. Additionally, if you do decide to "train hard" for awhile, different people may well train differently...
The Diminishing Returns for Similar Actions would model the improvement of this training regimen, since it'd be doing different actions with different records of repeated activities.
That being said, it's still basically grinding, and if you're controlling the character while repeating this it's still rather boring. Training for long periods of time (an idea earlier in this post) might let you design the training regimen at the same time you determine the skills you're training; I'd foresee that as a pretty good reason to incorporate this training idea.
Call of Cthulhu had the d% system, and improving skills consisted of some combination of succeeding then failing a skill check IIRC. It's been a while so I'm going off memory, but I didn't recall it being that ingenious, and that it tended to create as many problems as it solved (percentage skills not scaling too well with skill difficulty, for example).
Rolemaster I know pretty much nothing about, save for legends of its complexity.
Wow. This post is an authentic Wall'o'text.
Author
Topic: Skills Diversity (and other things) (Read 2644 times)