Crazy idea for a skill-based system (It's crazy I tell you!)

So, I'm working on developing a new MUD and I've been working on designing a skill-based system for it (Kudos if you've figured that much out). Outlined below is what I've got so far, please, poke as many constructive holes as you can in it. I've got some of this written down at home (where I am not) so I may come back and tweak the parts I don't remember perfectly. Now, without further adieu:

This system will have an as of yet unknown number of skills, but I will probably settle for a number somewhere between 20-40 depending on what makes sense. The skills will be along the lines of Plate Armor, Chain Armor, Leather Armor, Dagger, Longbow, Lockpicking, Stealing, Hiding, and so on.

Now, one of the things that bugs me, personally, about many games is the endless grind. I know many people enjoy this and by not tailoring the game towards that crowd I will be losing some of my potential audience, that is fine because I will be gaining a different sort of audience, it's a trade I'm willing to make. To that end, the system is going to be designed to reduce the amount of time a player has to spend training any given skill.

Imagine a bucket, if you will. Each trainable action the player successfully performs, or even fails at in certain cases, will fill up that bucket to varying degrees. At the bottom of the bucket is a small hole, draining that experience from this bucket into the actual amount of skill that player has. Think of the bucket as a buffer. Now, it may only take a player, who is training a skill "perfectly," 10-20 minutes to fill that bucket up, but it will take a week for it to drain completely. At any point during that week the player can top-off the bucket if they wish, but they don't have to. The bucket will also drain at the same rate, no matter how full/empty the bucket is. <update #1>It will also continue to drain whether they are online or not.</update #1>

10-20 minutes seems like a rather short time, but the odds of a player being able to train each skill at a maximum rate is rather low. Add in travel time, spawn time (if talking fighting critters), possible failures and we have a more realistic training time of 20-40 minutes on most skills. Taking that low rate of 20 minutes and multiplying it by, say, 20 skills a player might want to train and it would take a player 6.6 hours per week just to keep their desired skills training at the maximum rate. That seems like plenty to me.

Now, as in many/most games, the more you level up a particular skill the longer it takes to gain the next level <update #2>(Though I would want to throw in a cap on how long it takes to avoid running into the problem of 6+ months to level up once in a skill)</update #2>. This presents a bit of a problem in that the players first in (assuming they stick around) will be so far in front of everyone else that nobody else could catch up, assuming there is no skill-level cap. There are ways to mitigate this, of course, such as diminishing returns and so on. This still leaves the developer with a problem though, the development of content for the highest end of the spectrum that will never be seen by a decent portion of the population, especially in the system I am proposing where how fast a player can advance is so heavily regulated.

The solution I've come up with is interesting, I think anyway, but I am not entirely sure how most players will react to it. It's pretty simple really, lets say Player A has been in the game 5+ years and they have 1000 levels in lockpicking, which is the highest of all players. Player B enters the game for the first time and has, surprisingly, 0 levels in lockpicking. What I propose is a bonus system where Player B would get a learning bonus to lockpicking that is proportional to the highest achieved skill-level by any player in that skill, which in this case is the 1000 ranks Player A has earned.

What this would do is reduce the amount of time it takes Player B to level up their lockpicking. This bonus would diminish over time the closer and closer they got to Player A (who we can assume is still training), so that they would never be able to actually overtake Player A but they would be able to gain access to the highest levels of content a lot earlier that they might otherwise never reach, increasing the amount of content available to them as a player and thus, hopefully, increasing retention. In order to make lower level content still viable for new players, however, some sort of cap would need to be introduced so that Player B couldn't level up their lockpicking faster than, say, one level every 10 minutes(this number is just tossed out there without any thought put in); this would only matter at the very very very lowest of levels or if the game has been running for a very long time and a new player joins. So while a new player might advance through the lower content areas faster than the oldest players, they would still have to earn their way through the ranks.

Obviously there would need to be a lot of balancing and tweaking, and I may have forgotten a point or two, but all in all I think that pretty much covers it. I certainly plan on adding quite a bit of other content for players to play around with since they won't be spending all their time grinding but that is a whole other series of topics.

If it generally requires only 20-40 minutes of active play per week to keep your bucket full, and there's no skill-level cap, then newbies would never be able to completely catch up with players who had started before them (unless those players stopped playing for weeks at a time). On the other hand, the top players who had invested the most time into getting their skills so high might well feel their efforts devalued when they see the newbies advancing so much faster than they did.

The hardcore player who logs on for 10-15 hours per day would advance at the same speed as the casual player who logs on for 5 minutes per day. But with no skill-level cap, there'd be no finishing line, and even the casual player would eventually feel he was stuck in an endless grind.

If you wish to avoid grind, I'd suggest allowing players to max out in a relatively short period of time - perhaps a few hours. Or if you really want to have your bucket idea, you could keep the learn bonus fixed, but give newbies bigger "holes" in their buckets, so that they can eventually catch up with the top players as long as they're investing more than half an hour per week.

Before I go into anything else I realized when I woke up this morning that I forgot a piece of the puzzle. The bucket will drain even when players are offline. I've updated the OP to reflect this.



This is an issue faced by players in most (all?) systems without a cap. Those players who have been there longer are simply so far ahead there is nothing a new player could do. Of course I have seen cases where the Dev's realize that training levels X-Y are way too difficult so they change algorithms, which then makes it easier for those lower level guys to advance through that range. It might even be just the addition of new content which makes advancement easier and so on, so this is by no means a new thing...but of course it being institutionalized, so to speak, might very well have a more pronounced effect on the way older players feel about the system.



I understand your point, but a player would need to invest 20-40 minutes per skill per week in order to advance that skill at a maximum rate. This would mean there would be a cutoff at which point it would no longer make sense to keep training skills because a player will have maxed out all their trainable skills. A player who can invest more time would be able to train more skills at their maximum rate vs someone who only has the time to train, say, 5-10, so in a sense they would be advancing "faster" in that they would be able to do more with their character.

For those who can max out all of the skills they have access to in a given week, this is where additional content would come into play (not that everyone else can't or won't enjoy it), what other incentives are there for the players to continue playing? If advancement is the one and only trick the game has up its sleeve, you might get those people who just love the grind but nobody else will enjoy it anyway. It's a trade off.



I thought the bonus scheme I had described covered this. The more disparity between the highest player and "you," the faster you learn a skill. The closer you get the lesser the bonus. In this way a new player will never be able to overtake an old player, but at the higher skill levels this won't be a huge issue as the difference between 20/40 levels in a skill is far greater than, say, 500-600 levels in a skill if the system has diminishing returns. In this way the older players will still, unless they stop training, always have an edge but the newer players will get the benefit of having access to more content in the game.

One of the benefits, for a developer, in this system is that you always know exactly how fast players can advance since you can set the maximum rate. This way you can plan out developing new areas, creatures, skills and so on instead of having to rush to put something out because a small but very vocal part of your population suddenly has nothing to do because they advanced faster than you thought they would.

Realized I should probably clarify this statement: "the newer players will get the benefit of having access to more content in the game."

What I mean by that is that a newer player will get a chance to experience a wider range of content in the same amount of time, based on their being able to advance faster. The advancement of course has to be done in such a way that it doesn't completely break the game or make the low-mid range content obsolete...but it's either that, cap the levels early so everyone has access to all the content or have content that most people simply won't have access to because they aren't going to have 5+ years to invest.

Another benefit, which I forgot to add in previously, is let's say some of the earlier adopters to the game want to start up a second/third/whatever character. Rather than having to go through all the same grind, since the player will already have done that, they will get to focus more on simply enjoying the game.

If I wanted to get REALLY crazy, could even have a check to see what the highest skill-level is that that player has reached on that account for that skill and provide an additional bonus. Again this bonus would disappear as they got closer to their personal best but this would allow a player to, essentially, progress multiple characters at almost the same rate. Allowing them to role-play/experience different playstyles without forcing them to grind nearly as much. The key is making it feel like they have worked for and achieved something rather than it just being nothing but a grind.

Running the risk of turning this into a conversation with myself I would like to point out that I do see a potential problem with the last tweak to the system I mentioned, namely providing yet another bonus to characters based on maximum advancement of another character in their account.

If such a bonus would be based on a skill by skill basis, and assuming that not all characters can gain access to all skills, a situation could be run into where the new player might end up heavily unbalanced in their skills. Let's look at the following scenario:

Player A has two characters, Character A which is an old (3+ years) character and a brand spanking new character, Character B.

Character A is, primarily, a magic user but has trained in other skills. Let's just toss out climbing, for example.

Character B is primarily a fighter type. Since most of Character B's skills have nothing in common with Character A, the only bonus to learning for those skills would be from the bonus all new players would get based on the highest skill level achieved in the game. However, if Character B decides to take up learning climbing, suddenly that character would gain an additional bonus from Character A, allowing climbing to outpace all of their other skills, leading to a lopsided character.

I guess the question would be, just how undesirable is this? In such a system, would it be desirable to have a secondary characters bonus to any skill be based on the highest skill achieved by the primary character, instead of a skill-by-skill basis? Ignoring the learning rate, it seems like that would provide for a more 'realistic' character.

This only follows if you don't organize your level strengths and times accordingly. Let's take a rather extreme example case where each subsequent level takes exactly as long as all previous levels combined. Will a new player ever catch up to an old player? Well, no, not technically, but a new player with half the play time of an old player will be only one level behind. That's an effective catchup without any "tweaks". The question merely becomes then:
What is the effect of gaining a particular level in a skill?

Taking longer to gain the next level is itself already a diminishing return.

You bring up some good points here. I toyed with something closer to what you describe, the extreme example, but decided it would get to a point way too quickly where it would take too long to advance. That line of thinking, combined with personal preference/experience, led me towards a system where the curve was much more gradual. I also decided that while the training time should become longer and longer the higher up you got, at some point it should reach a cap as to how long it should take. This is something I completely forgot to add to the original post and will update it accordingly.

The reason for this is that once the time to advance even one level starts to get way too long, it starts to get a little discouraging, at least from my own experience. Once it goes from "Whew! I had to work two weeks to advance!" to "I had to work on this damn skill for six months to get it to move one level" I think it would just discourage a decent portion of the playerbase, but I could be wrong.

If a maximum cap is put in place as to how long a skill takes to level up, I would run into the issue where there really are no diminishing returns to speak of once that cap is hit. That is why I would want to create an algorithm which, the higher up you got, the less each individual skill level mattered. The whole thing would obviously need to be tweaked/balanced properly to achieve the desired effect but I think it could work.

I had a discussion with my wife on the way home tonight about such a tweak to the system; giving bonuses to a character based on another character on an account. She convinced me, using economics actually, that this would actually be a pretty bad idea. So I guess the answer to my own question is, not only would one of these alternatives be undesirable but that both of them are. I can see, and logically argue, for giving a same-character bonus to some skills based on others though. Having a high skill in two-handed weapons giving a player a bonus in using medium or large weapons since some basic swordsmanship would apply across most blades.

The danger here is that you're effectively rewarding players for not playing. Most newbies will immediately quit if they see an empty mud, so (particularly when the mud is just getting started) there's a huge benefit in encouraging your established players to stay online.


It's a problem of open-ended advancement, yes - in fact I've mentioned it before: http://www.mudbytes.net/topic-2506

However in a typical mud, if you play for one hour per day for a month, then I start playing two hours per day, after another month I'll overtake you. With the bucket proposal, if you start before me I can never quite catch up with you (let alone overtake you), unless you stop playing for a while.

It's not quite as bad as I thought, now that you've clarified your proposal is per-skill. However the same issue would still apply to the character's primary skill/s.


The problem is that newbies won't view it as a bonus, but as a penalty; the closer they get to catching up, the slower they'll advance. Likewise, I feel it could be frustrating for the oldbies, who'll see those newbies flying up the skill ranks that they themselves had to crawl up.

With my "varied hole-size" proposal, the newbie could still advance at (for example) twice the speed as the oldbie, but to do so they'd have to earn twice as much experience. The oldbie would see that the newbie was literally putting in twice the effort, and the newbie would feel their efforts are always equally rewarded - although they'd need to put in less and less time the more they advanced (which could be another problem, although in my experience this actually matches the way many people play anyway).


Players eat through content far faster than developers can create it, and will often get bored and quit once they've finished everything - but they (hopefully) hang around long enough to draw in the next generation, who then make their own way through that same content. Over time the content grows, and therefore the players stay longer.

With open-ended advancement you'll need to constantly keep adding new content, and with your bonus scheme your older content will be used less and less, as new players hurtle through it at increasing speed. This will make it more difficult to build up a playerbase.

You could have the content automatically scale in power as well, but then the top players would find themselves constantly facing the same content to advance, and that really would feel like a grind.


With open-ended advancement the grind is always going to be endless, but an account system is a decent way of discouraging multiplaying - players who put their characters on the same account would advance faster, but they wouldn't be able to play more than one at a time. Players who created multiple accounts would be able to multiplay their characters, but they'd advance slower.


Yes, I think that would be more desirable. Not because of realism, but because it encourages diversity. If each skill is advanced in isolation relative to the top player in that skill, you'll rapidly reach a point where every character has pretty much the same skills, simply because the lower skills advance faster.

Part 1 of 2: Post is just too long to do in one.



You are absolutely correct here. This is one of the things that has been buzzing around in the back of my mind the whole time. It really all comes down to a balance issue. How much time should players have to work at advancing vs performing other activities? Of course this ties into a suggestion I believe you made in your first response, namely that if I wanted to avoid an endless grind to allow players to max out within a short time.

In both cases the same problem arises, just to different degrees, how does one keep players in the game if grinding isn't all there is to it? I've got several ideas but they can really be broken down into a couple primary categories: Social interaction (roleplay amongst themselves, GM driven storyline etc...), exploration (which is limited unless you come up with a truly dynamic, unending world), economy (player shops, crafting etc...), dynamic quest system (I know, I know...dynamic doesn't actually explain anything but a longer explination would be off topic here). What it really all comes down to is creating a community. If you can foster that, then the players themselves will often find ways to, essentially, create more 'content' for themselves. This could be in the form of tournaments, dances, plays and other performances and so on. The key is to not get in their way and give them the tools with which to develop the community further.

I read an article not long ago, which was actually railing against George Lucas, which brought up an interesting concept for me. It basically said that an artist can't understand, and really therefor can't judge, their creation because it is the audience which imparts meaning onto their work. The audience makes the work theirs once it has been put out there and in that way they actually share ownership of this creation. If looked at in that light the goal of the MUD isn't, necessarily, to create a complete experience for the players to just navigate it is to create an environment where the players can come and create their own experience.



This is definately an issue and this is a trade off I've made based on a conscious choice, let me see if I can try and explain my reasoning. Let's take a look at two different players:

Player 1 is a college student who might have 40 hours to kill in a week and they devote all of that to the MUD.

Player 2 is a 40+ year individual, married with kids and has a full time job. They have maybe 5 hours per week to devote to gaming.

Let's assume both of these people start playing at exactly the same time. In a more typical MUD environment Player 1 would be able to advance very very fast while Player 2 would advance a hell of a lot slower. Fast forward 2 years. Assuming all things being equal, Player 1 will be 8 times further along then Player 2 and when viewed from an in-game perspective this makes perfect sense, Player 1 has dedicated a lot more time. I look at this situation from a different view though, both players have dedicated as much of their free time as they possibly can over the same span of real-life time. In addition there is the creation of content for, what is most likely, a much smaller portion of the overall population that the rest of the population may never see.

As we have discussed there are multiple ways to handle this last issue and I'm not saying that my solution to it is the best way to overcome it, but it is these types of concerns which led me to the creation of this system. The training times, bonuses, size of bucket, draining speed of the bucket and so on can obviously be tweaked and tailored based on the population of the game and the amount of time most players tend to dedicate to playing in order to make it as 'fair' as possible.

Post 2 of 2:



The first point is a good one, I hadn't really thought of new players viewing it as a penalty. Of course if given the choice of no bonus at all, and thus longer training times, or gaining some bonus do you think most newbies would still complain?



This is an innovative solution. I would have to think about this some more and the overall impact but I must admit I am intrigued. I think a solution along these lines could have some extra benifits as well such as forcing newer players to specialize, unless they have a ton of time and can do every skill they have access too, which would help create some diversity.



I completely agree on the content issue. I don't honestly expect to be able to keep up with the players but I think knowing exactly how fast they can advance and having a cap on how fast they can advance would give a little bit of breathing room, not much but every little bit helps. Of course trying to foster an environment where the players are actually able to, in some ways, create content/experiences for themselves would go a long way to retaining some of the older players, as I stated above.



The slower advancement on multiplaying was a benefit that slipped past my radar. One of the things I'm worried about under such a scheme is a new player who comes in and is watching other new characters around them advance faster, not based on bonuses to that character but on the fact that the other player has yet another character who is higher up.



I see your point here and wholeheartedly agree that diversity amongst characters is a highly desirable trait.

Actually some MMO games do just that. For example, WoW gives you a "rested bonus" where you get more XP (for a while) if you haven't played for a couple of days.

The idea is to try to balance out things a bit so that people who can afford to play every day (or all day every day, even) don't get too far ahead of those that can't.

Indeed, knowing you will have a bonus after not playing all week, actually gives you an incentive to log on, during the weekends.

---

An interesting thing to consider here is that the player at the keyboard may have skills that the avatar doesn't. Consider two players who both make a level 1 character. One has played this game before, many times. They know where things are, how the game mechanics work, when to fight, when to run away, how to most effectively skill-up etc. The other one is genuinely a newbie. But initially the server can't tell the difference between the two.

One way of slowing down the skilled "real" player (if you want to do that) is to randomize things a bit. For example, quests may appear in different locations, or get you to do different things.

Or maybe the game could detect that player A is making fast progress and quietly ramp up the difficulty a bit (eg. make mobs harder to kill).

One of the most extreme examples of this is EVE-Online. A wonderful game I loved to death, I just couldn't keep playing because it was a second job. All learning is time based. A skill has five levels, each level increasing in time to train, with the fifth level taking more than the other four levels combined. They even put in a skill queue where if your training will end in the next 24 hours you can automatically set the next skill to training. They haven't lost players overall though, in fact nine years after release they continue to set records for numbers of people online at any given time. The key to the success is the sheer amount of OTHER content that players have to work with, most of it centered around the players working with or against each other. That game is really what got me thinking along the lines of a week long skill absorbtion.



That would be an interesting thing to set up for sure, I imagine it would have to be done by trying to figure out what the 'typical' new player advancement rate is and anyone who is an outlier would have their difficulty bumped up. I am sure you could set an automated system like this up with a little thought, along with plenty of trial and error. Of course you would need a large enough sample to work with to make the algorithm viable.

I am all for randomization and really do want to set up a quest system that has several variables to consider each time it generates a quest, ending up with millions/billions of potential quests. Such variables could be things like primary/secondary skills of the player, location, how any quests have been completed in the same area, how high the population is in an area, if certain types of creatures are being over or under hunted and so on. With a little planning I think you could come up with something to keep players busy for a very long time.

This got me thinking back to my experience with EVE more and there is an interesting psychological effect that came out of their particular training method. Despite the fact that logging in and out had no effect on how fast a person trained, people tended to become almost obsessed with the skill training times, logging in so they wouldn't miss and instant of skill training time when one skill was about to tick over, and constantly checking their skills even if they knew they had 2 days left. I think this comes from the fact that a player knows that, literally, they can't make up for lost time as it is just impossible due to the game rules. So in a system like the one I am describing there might be the same sort of drive, even if not to the same degree. Players will know that if they don't log in this weekend, or this weekday or whatever, they simply can't make up that training time at a later date. It would be an incentive to log in fairly regularly, if even just for a short time to top off a skill or two.

But the flip side is, it is a disincentive to keep playing if there is a penalty for not logging in.

Say you go on a one-month holiday. If there is a penalty for not playing (eg. the skills drain away) you might think "oh no, I've been away too long, not much point in logging in again".

From what I recall in WoW you very rarely actually go backwards. You might not go forwards (eg, if you can't find any mobs to kill or quests to do) but you don't go backwards.

Even dying has a fairly minimal penalty (no XP loss, minor equipment damage, and a bit of time wasted while you "run back to" your corpse).

I completely agree with this and have never liked the idea of penalizing someone for not playing. There is simply too many unknowns in life to expect someone to play X amount every single day or every so often, it's a game and life should come before that. I would never play a game that penalized you in such a way. Not to mention from a developer perspective I can't imagine such a system doing anything but keeping the player base insanely small or driving the players away completely.

Actually as I wrote that last sentence I realized that there is one way such a game could probably work...it would need to be graphical and would need to be geared towards the Asian market.

Was thinking about this whole thing last night and this morning and came up with a slight tweak to the system and wanted to see what you guys thought.

Forget everything about bonuses and just get back to the basic system of having a bucket that drains at a specific rate for players. Now, we have established that different players--whether due to playstyle, real life obligations or they have just been in the game a while and are playing less now--have different amounts of time to dedicate to the game. What if a 'slider' was introduced to the skill system that would create the following effects:

At one end of the spectrum (let's call it the right side for illustration purposes) the bucket size would be smaller but would drain at a faster rate, let's say an hour or less, much closer to what a more 'traditional' skill learning system might be. In this situation a player would be doing a lot more grinding and it would suit a player who could dedicate 30+ hours to the game per week and is willing to sit there and just train, train, train.

At the complete opposite end of the spectrum (the left side) you have a very large bucket that drains a bit slower, basically the system I originally described with it taking a week to drain. This would be ideal for players who could only play for a few hours a week, or those people who want to spend much more of their time roleplaying without being completely left out of the advancement loop.

There could be a couple steps between those two for people who are, obviously, more in the middle.

The amount of time it would take to completely fill up the bucket would need to be examined, obviously, and tweaked to work with the range of settings now available.

This slider could only be moved IF the character currently has zero experience in their bucket, helping to prevent the system being gamed. To support this players would have to be able to turn off their learning while still being able to perform any/all normal activities in case they would want to switch the slider.

Now, to help make this system a bit more fair to those players willing to put in the extra time and effort (those all the way to the right), anyone with their slider set more to the left will learn at a slower rate. By slower rate I mean in terms of absolute skill levels learned in any given time period, not how much they learn from a successful activity. If we assume that someone with their slider all the way to the right learns at 100% of maximum speed, someone with theirs all the way to the left might learn at 60%, or 70% speed; I would need to do some number crunching and testing to see what the optimal difference would be.

From an implementation standpoint it would be incredibly simple to do, just a matter of changing something like three numbers/variables in the exact same algorithm depending on where a slider is set.

The potential benefits, as I see it, are numerous. You keep players who can only play for short periods of time interested, those players obsessed with the maximum advancement/grind would get to do their thing, roleplayers would get to roleplay without being left in the dust and more people would get to enjoy mid-higher level content that the GM's/Dev's spend the time making.

I would still want to give same-character bonuses to related skills--a high level large sword skill giving a bonus to learning a low level medium sword skill--but I am not sure if any other bonuses such as what I was describing in previous posts would be desirable. I'm still thinking through the implications with this new tweak.

Yes, because they won't see it as a bonus, they'll see it as the "normal" training time.

In my mud, I introduced a "boost" system that allowed players to get 2.5 times the normal primal (effectively exp) for their first couple of dozen kills each day. It was part of a mechanic I introduced to discourage botting, but the players were initially happy with the faster advancement.

A few months later, the next generation of players complained that it wasn't even worth playing the mud once they ran out of boost points, because the unboosted "penalty" was so severe. They said they might as well go and play a different game until their boost points had recharged.

I later reworked the system to something a little reminiscent of your slider, introducing casual, hardcore and skill boosts, and ensuring that the more time-consuming activities also give the largest rewards.

Human psychology is an interesting thing sometimes. I can see how the new players would view such a system differently from those people most familiar with the pre-boost system. Have you been satisfied with the results of your new boosting system?