|
Post by BitterFlashback on May 29, 2014 1:43:27 GMT -5
So I almost posted this in Code Discussion but it's not technically for Arm. dcdc and I made comments in the shoutbox I kept thinking about. because it's absurd.
It led me to wonder what kind of skill-gain system people would want. For a game without levels. the most obvious two systems are Arm's fail-based and its inverse success-based. With success-based you gain a point for a success, but either the amount of successes you need to push a skill to the next point begin increasing dramatically or successes only count against more challenging things. I could also see success-based being limited to how many successes per-time-period would count.
So what would you want in a game? One of those? or something else?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 29, 2014 5:07:22 GMT -5
I would like a system that is based more on successes than failures. When you manage to do something correctly, it should be a better example of how to do it correctly than fucking it up. I think that the successes required should scale with the skill% you have currently. Like...
1% skill You make 1 thing successfully. Your timer goes to 2% One hour later. You make a thing successfully. One hour later. You make a thing successfully. Your skill% notches up to 3%, ala almost the fibonacci sequence, so that as you get better, it takes longer to get better, but you could do that with a high amount of successes and still be good because of a skill timer built in of 1 hour (give or take, depending on int/wis modifier, etc) between crafts that count toward it.
|
|
Jeshin
GDB Superstar
Posts: 1,451
|
Post by Jeshin on May 29, 2014 7:53:15 GMT -5
The 1st thing you need to understand is why failure is chosen over success or some other factor. If you want a simple way to do scaling difficulty then failure is the easiest solution. You fail less the better you are, ergo you skill up slower the better you are. You get a nice simple progression from many failures = faster skilling up at lower ranks to fewer failures = slower skilling up at higher ranks.
If the intent of a system is to have a scaling difficulty in gains then I believe failures are the optimal simple choice. If you want an alternative system then you can do difficulty of craft/attempt. Essentially you gain skill based on the difficulty of the attempt. For crafting this would mean you always have to be progressing to better more complicated crafts to continue to skill up success or fail. For combat this means you would need to be challenging higher DC rolls to continue to skill up.
Combat Example: Tregil has an armour of 20, so when you roll-to-hit it sees 20 and assigns a 20 difficulty to it. A carru has an armour of 60, so when you roll-to-hit it sees 60 and assigns a 60 difficulty to it. Eventually you stop getting gains from fighting 20 armour/difficulty NPCs and have to fight higher challenge ones to continue to progress.
|
|
|
Post by jcarter on May 29, 2014 8:20:38 GMT -5
success/increase based on accomplishment vs relative skill level would be an interesting metric to see implemented
i.e. newbie cook success on travel cakes - average skill increase newbie cook success on making the perfect rare steak - good skill increase near-master cook success on travel cakes - little to no skill increase
|
|
japheth
staff puppet account
Posts: 31
|
Post by japheth on May 29, 2014 18:35:02 GMT -5
The logic behind skill-ups only occurring on failure is fairly easy to understand of course - as your skill rises, failures become rarer and so your skill-ups become slower. There is also an implied understanding that this means more skilled people would need to seek out harder tasks to do so they are more likely to fail.
What I think the original design logic misses obviously is the cleverness of players at maximizing their grind, and the bizarre incentives that arise in the system - for instance the examples given in this thread thus far. The original designers might counter that such behaviour is not really bizarre and you see examples of it in literature and real life - think Luke Skywalker training to deflect blaster shots while wearing a vision-impairing helmet, or someone going for a run with weights on.
However, I think there is just one final step that the original design missed that salvages this entire mechanic - it shouldn't then be failure that gives skill-ups, but success. The real game should be making the task as hard as possible, and then succeeding at it. So ideally, the engine would evaluate any skill check for its total difficulty penalty against the current level of the skill, and if it is sufficiently difficult they are eligible for the skill-up roll if they succeed.
Tuning the level would be necessary but certainly doable. I would probably use some ratio of the penalty vs current skill level (say, penalty must be half the current skill level) but also use a flat offset to account for some learning at very low levels of the skill.
|
|
|
Post by BitterFlashback on May 29, 2014 18:56:05 GMT -5
Hm. All good stuff so far. Im still in the plannnig phase for a mud codebase and wanted to settle on a direction to take with skillgains. (Obviously I'm not looking to do character levels.)
I'm personally going to rule out using failure skillgains. Or maybe Ill have both failure and success as possibly giving skillgains, but failure giving less. I get the logic to using failures: they eliminate a ton of work. But I dont like the side-effect. I actually do have some crafting skills IRL. So I always felt a little silly playing Arm and setting myself up to be at a disadvantage. "I could use this jar of tanning solution. Or i could branch leatherworking some time this year."
Perfect example of this is learning fine glove making. In real life. As a beginner you're supposed to use high quality leather/fur. Seems like a bad idea because youre new and you'll probably mess up? That's a reasonable question. But there is a damn good reason you dont start on cheap material. You need to learn to do something the right way first.
Before you learn to compensate to work with what youve got you have to know what you should have been doing if you werent in this situation. If you start with bad material youre not going to learn what youre supposed to be doing. You'll be learning stuff that may not even be applicable to using the good material you intend to work with. And since youll be learning the wrong way to do things first you'll have to unlearn them later. You also won't figure out if you have a talent for making gloves or if you just suck until you move on to the material you should have started with.
|
|
|
Post by mekillot on May 29, 2014 22:05:09 GMT -5
Ultima Online had a nice functional system.
Skills raise, on a chance system, on success. The chance stayed the same until the soft cap, and it wasn't that low. Skill increases get smaller as skill goes up. Towards the higher end the increases were like .05, but there was no skill lock out timer. After something like 70 or 80 (of 100) you hit the soft cap. I think the chance of a skill up decreased a lot. All characters could learn any skill up to 100. They only had 700 total points at any given time though.
Skills could be locked, or set to slowly go down. As the player wishes.
|
|
|
Post by BitterFlashback on May 29, 2014 23:30:08 GMT -5
Huh... I think I may go with combining the ideas from japheth and mekillot because Im feeling inspired. What if you had a system where... lemme just fudge some numbers so what Im saying will be easier to follow. Every 25 points is a skill gain. When you do something theres a difficulty level. let's say you cant even attempt something more than 20 difficulty higher than your skill. To curb abuse. Points for a success would be: if (skill > difficulty) and ((skill - difficulty) > 5) points = 0
if (skill > difficulty) and ((skill - difficulty) <= 5) points = 1
if difficulty > skill points = ceiling((difficulty / skill) * (difficulty - skill))
To translate that from pseudocode... if your skill is higher than the difficulty but the difference is 5 or less, you get 1 point. If the difference is greater than 5 you get nothing. Otherwise, we divide the difficulty by your skill then multiply the quotient of that by the difference of the difficulty and the skill, and round up the product. ... to translate that into hard numbers... Say you're building a cabinet. your carpentry skill is 10. A cabinet has a difficulty of 25. So if you succeed we... (25 / 10) * (25 - 10)2.5 * 15= 38 points rounded up You make 1 cabinet and now youre at 11 for carpentry. Huzzah! but now the same cabinet now only nets you 32 points. By the time your carpentry is 24 youre only earning 2 points (we round up) per cabinet. I might have to adjust the numbers and formulas a little... but I like where this is going. I think points should be reset upon skill gain but Im kinda on the fence.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 31, 2014 0:01:42 GMT -5
I would like to see crafting that was handled differently than I've seen in other muds.
I want it to take realistic amounts of all the things it would actually take, and for things to be more interchangeable in the recipes.
Interchangeable how, you might ask?
Okay, here's a fairly good example:
You can make resin from pine sap. To do so, you dissolve the sap in acetone and strain it to remove the impurities.
But you can also dissolve it in moonshine or concentrated alcohol.
Why? Because what you need isn't moonshine, and it isn't acetone. It's pine sap and a solvent, pretty much ANY solvent.
So you could use pine sap and any item flagged as a solvent to make pine resin.
Automatically opening up the crafting system to a more broad and realistic implementation.
|
|
|
Post by BitterFlashback on Oct 31, 2014 0:36:08 GMT -5
Yeah. i think i discussed how i was going to handle crafting recipes to make them self-updating in IMs. I definitely agree with this approach as well; chemical types I mean.
|
|
|
Post by Redacted on Oct 31, 2014 2:36:14 GMT -5
I would not do skill gains on increases or failures at ALL. Or if implemented, only on general skills that are accessible to everyone.
But I'd do away with that system entirely, because it reduces advancing your character to a grind, in any way shape or form. This is how I would do it, though it may not be possible using the current iterations of code. And I can't believe I'm suggesting this because I wanted it to be a sort of proprietary thing, but since I'm not actively involved in any active game design, go wild.
As a character uses their skills, the game keeps a tally of the skills they use and how often. Every set time period that goes by (this can be anything, minutes, days, weeks, months, years, wtfever) the game takes the skills most used, again, could be the most used skill, or it could even be a section of maybe the five most used skills, or three (my idea was three) and prompts the player to choose which of those top-used skills-per-time-period they wanted to raise. Players would actually be forced to make choices on the paths they'd want their character to advance on, though they would also not start the game with drastically low stats and also be able to set their own general proficiencies on the skills they wanted to spend them on.
It would also provide a way to control both the rate and way people chose to learn their skills, and allow them to pick up new skills and progressing at them at a rate feasible to the game world.
This would clearly not be ideal in a game where high char death was expected, but would be ideal in a roleplay immersive world where char death would occur at a more reasonable level.
It would also do away with the skill grind, since skills would be up for selection based on what you actually used and the frequency of it, which means what you do in game actually has an impact on what your character is offered to progress in, skillwise. It means that unless you make dresses on a regular basis, your dressmaking skill will probably not be raised very often. But if you find yourself fighting a lot, and expect to do so, despite choosing dressmaker skills, or whatever, you could gradually alter your path to that. Or if you wanted to pick up shoemaking to add to your clothier skillset, you would actually have to set aside and devote time to doing it, AND you have to sacrifice giving up advancing in other skills to get it and keep raising it.
You could even extend time periods between raises as the skill gains in a particular skill increase. So 0-10 goes fast, 10-20, pretty fast, 20-30, a little slower, 30-40, progressively slower-- yadda yadda, thereby helping to somewhat offset skill inflation. So people could learn an assortment of things and be meh to eh in them, but sink time into specialities to be really good, with increasingly better gains/abilities linked to them based on their level.
There'd be other checks and balances, but you could do this realistically in a more intuitive and interactive manner than the current system which basically just rewards those with the mindset and stamina to grind, or having to grind just to get 'ahead'. It also means there's more consequences to your character codedly, based on the things they choose to do in game, and you as a player actually would be partially engaged in directing some of the choices to hone the character into their speciality or identity.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 31, 2014 3:52:04 GMT -5
I like the idea, but I don't like that it only prompts for 3 skills or so.
My proposal for that would be to have the # of skills depend more on int/wis/whatever-mental-stat from maybe a base of 3 at the lowest 5 or 6 at moderate to perhaps 10 at maximum. Or maybe base the skill points you can use to increase skills on those and the baseline is 3 skill points?
I don't know. Dark Isles is the only game I've played with something similar to that (though use could also raise them, and the higher your skill, the more RPP/skill points it took to raise each), and it was awful, because even if you never use a weapon, you could just idle for indefinite periods, rack up the RPP and use it to max weapons skills, etc.
|
|
|
Post by Redacted on Oct 31, 2014 13:59:24 GMT -5
Yeah but DI also actually didn't base it on a set time period. You could raise your skills as normal (by using them) but it was pretty random. Or you could roleplay for roleplay points and spend them. It wasn't time based. It wasn't that great. There's a couple games out there currently that use that model-- I don't like it either. My idea was to get away from automatic skill gains entirely, since it essentially turns it into both a grindfest and penalizes casual players. The method could be tweaked, but the ultimate goal was to create a skill-gain system that revolved around player customization and choices. Also I'm not sure why you would have the skills depend on any stat in that system, but I guess you could set it up that way. Another way of going about it would be to have stats affect advancement, if you want to give out more than one point, which is useful in a multi-tiered skill system as in my example below. You could also have it prompt for any number of skills, but I don't think I explained it clearly. You would actually only be able to raise one of the skills, the other skills would only be part of a choice. So if you had people start off at a reasonable level in the skills they have chosen (based on stats or other skill choices or whatever skill ordering system) you could have it so every character would get an 'advance' based on the skills they actually used IG, the frequency of that usage, and the desire of the player to advance a skill. It would look like this (off the top of my head, skill sheet can look like anything really): Background information: Jane is a warrior/bongmaker type of character. Melee Combat - Primary
Edged Weapons - 100 Swords - 50 Knives - 18
Blunt Weapons - 100 Clubs - 33
Defense - 100 Parry - 42
Gathering - Tertiary Herbalism - 3
Crafting - Secondary Glassworking - 100 Bongcraft - 32 Cooking - 100 Snack food - 99
Over the course of a real life week, Jane does quite a bit of glassblowing, including a few bongs, goes sparring a few times (though she swithches between weapons), cooks a ton for her friends, and picks a few herbs in the forest. Just typical coded activity along with a bit of social rp with other players besides her coded play. At the end of the week, when the next skill increase is due, Jane is logged in and the player receives the prompt: > You are ready to advance a skill! You may choose to increase your proficiency in: Bongmaking, Cooking or Herbalism. Which skill do you choose to advance? (Please use the command advance <skillname> e.g. advance Parry or none if you wish to skip advancement.)
> advance bongmaking Your hard work has paid off! Bongmaking has increased! The time period between skill advancement would be a purely OOC construct based on OOC time, not time played. This means whether you play 1 hour a day, 4 hours every few days, or 8 hours every day, the skill gains are normalized-- as long as your character does one coded action that is contingent on their coded skills between skill gains, they will get to gain or advance in the skill, since the gain is entirely based on a proportional frequency of use, and not how many days played you have, or exactly how many times you went out to chop things up with swords. In order to have a character proficient in their skills they have to use them. Not even that often, but they have to actually use them in game. Which means you CAN sit around in a tavern rping and singing all day, but you will only get an option to raise a singing skill, and never your combat skills if you don't bother to use them. My idea for this revolved around a desire to create a basis for a system that: - Did not favor regular players unreasonably over even the most casual player.
- Allowed players a large amount of input when it comes to molding and advancing a character, but...
- Also base how the character evolves codedly on what they actually do and accomplish in game via code usage.
- Be able to pace out skill gains reasonably to avoid both skill creep and skill hoarding while still allow characters a huge amount of flexibility as to what they can do, skillwise, without admin intervention/skill setting.
- Formulate a system that would reduce or remove skill-grinding or spamming without arbitrarily punishing players. It simply would make spamming or grinding useless from an OOC perspective. Skills could be used IG as they made sense instead of locking players into specific routines for solely OOC purposes.
- Was fair in rewarding characters for actions actually performed. Which means casual players would not be gimped compared to religiously active ones, codedly, but people who did not log in at all or who did no coded actions in game, would also not get any advancements in skills.
Keep in mind this doesn't really address roleplaying issues or other methods of bolstering skill-gains, it was solely a system building exercise to see if an alternative system to the current shitty hns oriented ones in nearly all the current iteration RPI's could be created without losing the coded-backing of IG action that makes a mud a mud. It also doesn't go into the best sort of skill system or how to set one up, but solely represents my ideas on how skill gaining would work overall within the scope of a mud that implements it. It also divorces skill gain from skill or failure with a skill, because it simply assumes doing both leads to growth in knowledge and finesse in being able to do it, hence advancement. There was something else I wanted to write, but I completely blanked, so I'll leave it at this.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 31, 2014 15:39:29 GMT -5
See, I thought you meant could raise 3 skills, and was suggesting that it could be a variable number of skills you could increase based on your ability to retain information (aka wis/int/whatever-mental-stat). I like a lot of the rest of the premise, but I just think unless it's per OOC day, it should be more than a single 1 point skill increase, given the time differential and the amount of activities you might be doing in it. Like... you can do more than one thing a day, you can learn more than one thing a day, if you're smart enough, you can learn a LOT in a day. I love the idea of getting rid of the grind and starting at competant skill levels. It makes starting a new pc hurt a hell of a lot less.
|
|
|
Post by BitterFlashback on Oct 31, 2014 16:21:30 GMT -5
Youve definitely suggested an interesting system. I like a lot of it. But Id say my main issue with it that it lends itself to massive abuse by griefers. one of the advantages to a system that requirs an investment of time is people can't casually create the RP equivalent of suicide bombers.
With your system, i could spend 10 minutes per account/character and have a small army of characters to bulldoze people who spend hours of free time playing. it literally rewards the least invested players the way it's set up.
You touched on an issue i've been complaining about for a while though. The absurdity of having a permadeath MUD where you always restart as a noob. (Ignoring staff-induced skill boosts.) i've been planning something to address this- based on your behavior as a player, time played since you last lost a character, and skill levels acheived, you'd get some number of skill points to spend at character creation. Anyone could play anything, but better players have to waste less time building up.
more on this later.
|
|