My Game, Can I Show You It?

[Red_Rob]

Everything has its own dice pool. The task resolution mechanic is stat + skill. Lockpicking is keyed off of Dexterity, so: if a character has a d6 Dex and a d8 Lockpicking, their dice pool for using that skill is d6+d8.

Any time dice are rolled, if you roll higher than the average, you get 1 XP towards the skill;

I forgot to mention, this sounds kind of awkward at the table. You have multiple dice types combined in differing amounts on the fly, and yet any time you pick up the dice you have to first work out the average and then in addition to checking for success or failure / success margin / critical success or failure, or whatever else comes from the roll, also remember to check whether you beat the average. I can see the "average" checking getting missed during hectic scenes, which could cause some butthurt.

It just sounds kind of cumbersome to me.

[Avoraciopoctules]

Perhaps if we use some kind of CR equivalent to handle it. Like... say a random goblin is CR 3. If your average for a given roll is higher than 3, you can't gain XP for it from that goblin, it just doesn't provide enough of a challenge. Obviously we'd have to figure out a way to generate that number, and preferably in a way that doesn't immediately make it obvious to players what the CR of things they're fighting is, but it sounds - off the cuff - like it might work.

You are adding a lot of complexity to your game's XP system, and gamers have a limited tolerance for complexity. Are you sure you wouldn't be better served by a simple "XP for quest completion" analogue? Then you could channel that extra complexity into equipment or setting, something that could excite interest from prospective players instead of presenting a bigger learning curve to prospective Mister Caverns.

Figure out the stories you want your game to tell. Then figure out how you can mechanically incentivize them. Personally, if I was doing a caravan game I'd probably award build points at plot milestones and for successful journeys from one hub of civilization to another.

The "caravan" term was something I just came up with in response to the realization that "party" may not be the best term, but I'd been using "party" as a default because I come from a primarily-D&D background. It wasn't intended to imply that it is a game about a caravan in the real-world sense of the word.

That said, the reason I went with it is because part of my vision for this game was inspired by FF Crystal Chronicles. I like the notion of a party using a wagon or small wagon train as a base of operations.

Well, if you want to run a game where the party acquires Plot Juice by hook or by crook to save their home base, giving advancements for every drop added to the Macguffin back home would probably be a good starting point.

[GnomeWorks]

I forgot to mention, this sounds kind of awkward at the table. You have multiple dice types combined in differing amounts on the fly, and yet any time you pick up the dice you have to first work out the average and then in addition to checking for success or failure / success margin / critical success or failure, or whatever else comes from the roll, also remember to check whether you beat the average.

I realize this is just anecdotal, but it works out really well. We've had zero problems at the table with it, and had quite fast task resolution (3 on 3 combats last around 10 minutes).

It's just a matter of knowing what you need to track on the character sheet. In d20, you note the modifier for each ability score; in my game, you note the average value for your skills.

You are adding a lot of complexity to your game's XP system, and gamers have a limited tolerance for complexity.

I know, I thought of that when I suggested it. The extra look-up time for the compare between the CR of a monster and your average roll for the skill... you'll be doing a look-up for average anyway, for XP purposes, but it is another number for the MC to remember, which makes the game run slower.

I'm sure that some combination of these ideas would work out okay, though. I think we've got some wiggle room, time-wise, before things start feeling clunky.

Are you sure you wouldn't be better served by a simple "XP for quest completion" analogue? Then you could channel that extra complexity into equipment or setting, something that could excite interest from prospective players instead of presenting a bigger learning curve to prospective Mister Caverns.

I've gotten very attached to the notion of gaining XP on skill use: it makes a lot of sense to me, and to my playtesters. It feels most associated to me; honestly, going to anything else would feel like disassociating XP gains from what caused them, and that doesn't sit well with me.

Well, if you want to run a game where the party acquires Plot Juice by hook or by crook to save their home base, giving advancements for every drop added to the Macguffin back home would probably be a good starting point.

In a particular instance of the game, that wouldn't be a bad call.

I am not looking to tie the game down in that much specificity, though.

[Avoraciopoctules]

Thing is, some mechanics just aren't suited to a cooperative storytelling environment. With this XP system, characters will diverge with time. You could face long-term balance problems if one person doesn't make it to a session. Unless, of course, it takes a lot of advancement points to make a meaningful difference in character effectiveness. But that itself could be considered a problem.

If you are worried about disassociating XP gains from what caused them, I'd like to point out that a lot of learning comes from wanting to learn something. Lots of people go through public school learning barely enough for the next test and then discarding the knowledge right after graduation. Should someone completely uninterested in learning to fight better improve because a couple times they flailed ineffectually at assassins before their bodyguards closed in? What if your explorer's player complains that these added dots in social skills they've been given ruin the gruff ranger archetype they wanted to play?

[GnomeWorks]

Thing is, some mechanics just aren't suited to a cooperative storytelling environment. With this XP system, characters will diverge with time. You could face long-term balance problems if one person doesn't make it to a session.

My thinking is that if a player doesn't show up, provided the game didn't stop in the middle of something where their disappearance doesn't make sense (middle of a dungeon, etc), the MC can track how much time passes during that session, basically tell that player "you weren't here, so your character was training; you have X time-units to work with."

It's not perfect, but I don't think any game system perfectly deals with a player's absence in terms of advancement.

Unless, of course, it takes a lot of advancement points to make a meaningful difference in character effectiveness. But that itself could be considered a problem.

It depends. XP requirements are multiplicative, so the higher you go, the longer it takes to get better.

If you are worried about disassociating XP gains from what caused them, I'd like to point out that a lot of learning comes from wanting to learn something. Lots of people go through public school learning barely enough for the next test and then discarding the knowledge right after graduation.

At the same time, though, there are things you pick up without really realizing you've picked up on them.

Should someone completely uninterested in learning to fight better improve because a couple times they flailed ineffectually at assassins before their bodyguards closed in?

In this case, I would argue that - in that moment - the answer here is probably a very strong yes. After the fact, probably not, but that's where decay comes in.

What if your explorer's player complains that these added dots in social skills they've been given ruin the gruff ranger archetype they wanted to play?

There is a point at which the ability to model things meets the wall of mechanical complexity - diminishing returns, in terms of additional mechanics to model things in greater detail.

I think I am willing to sacrifice precise "sanctity" of a character concept for our advancement system. Yes, you may wind up with things that you didn't originally envision your character doing; but that feels like a small price to pay, here.

[tussock]

Hehe. RPG players tend to try to stay in character, but if they don't ...

In UFO: Enemy Unknown, an ancient-but-famous computer game from before time, every stat was raised by using that stat, and the harder you pushed it the better.

So the sniper got better by sniping holes in distant walls every turn. The throwers threw all their grenades far away from the aliens to improve throwing. The strong people weighed themselves down so heavily they could barely move and just danced in place. And all your reactions/breach guys at the end stood outside the crashed alien spacecraft and relied on mass reaction fire to shoot anything that stepped out. Anyone who needed speed or endurance randomly ran from end to end of the map on a clear part as that went on..

You are trying to be cute, but you are full of shit. There is a long history of games with mechanics of use, and people grinded in all of them. In Xcom, you did run people around to gain X, or drag out pointlessly easy fights to farm up your people.

In Morrowind you do travel by jumping. In Oblivion you do Sneak walk into a corner and leave the room. Those are all things people actually do, because having a better character is better than having a worse character, and if all you have to do is not solve the problem immediately, but just bring it really close to being solved, and then putz around to grind stats, then that is what people will fucking do.

I don't understand. You said I was full of shit, and then you copied what I'd said to correct me. Are you trying to be cute? Is that what that means?

I'd like to say you're just being a reactionary idiot, but it'd be funnier if I'd missed the point again.




Anyway, my point is people don't merely like rewards, they seek them out in fine detail really aggressively, and will fucking hate you for frowning on that. XP for treasure makes people into sneaky treasure-seekers, XP for furniture removal makes people Greyhawk everything. XP only for monsters makes people kill everything that doesn't have a "DM's pet, not for killing" sign on it, and even then sometimes.

XP for "having a good time" makes people plaster fake smiles on their faces even when they hate the game. Which is a thing which makes people genuinely happy and helps them have a good time, but is also super-manipulative and disruptive in the long term because people can't complain about real problems they're having.

It's a powerful tool, is what I'm saying, people will pretty much do anything for it. Well, no, because they can leave instead if that's more comfortable, but anything socially appropriate. Items and character options and all sorts work the same way, if you've got to chat to NPCs to get skills, you chat to NPCs (and work to keep them friendly and alive and well served with their quest demands and so on). If you've got to hit dungeons for important items tucked away, dungeons will be hit and cleared in detail.

3e D&D is finding the best options and gear in the books, and "defeating" monsters with it. That is all it is in the end because nothing else is rewarded. Even getting more money in game just means you get less money later, in the 3.5 rules at least. That's where it tends toward the linear story structure and always-hard fights surrounded by an empty void of no-player-access found in 4e.

[fectin]

Regardless of how your experience has gone, the obvious way to optimize in your advancement system is to roll for things all the time.

Preferably with as high dice as possible, because 1dX + 1dY gets slightly more likely to roll over the average as the higher of X or Y grows. In fact, lets break that down real quick: only the largest die affects the likelihood of gaining XP, so your probability by largest die size is:
d4: 37.50%
d6: 41.67%
d8: 43.75
d10: 45.00%
d12: 45.83%
d20: 47.50%

Your advancement then, is primarily based on how many actions you can cram into a given session, but also based on cherry-picking actions from an arbitrary list of things you've already done a lot of. That's seriously encouraging crazy behavior, where every character is OCD about their one thing, and does it at every opportunity.

[GnomeWorks]

Regardless of how your experience has gone, the obvious way to optimize in your advancement system is to roll for things all the time.

Yes.

This goes back to the sandbagging discussion earlier in the thread, I think. We hadn't come across that before, but it is a legitimate concern. So obviously there needs to be a mechanical disincentive in place to prevent people from doing one thing repeatedly over and over for XP gains.

In fact, lets break that down real quick: only the largest die affects the likelihood of gaining XP, so your probability by largest die size is:

I'm looking at your math, and... not really seeing it. I ran some example dice pools through anydice, and got results that seem counter to yours (as you increase the size of the larger die, the chance to roll higher than average seems to decrease slightly).

I'm not necessarily saying you're wrong, just not seeing how you're coming at the problem. Please feel free to explain further.

Also, feel free to not include d20's in your rebuttal, we don't go there.

Your advancement then, is primarily based on how many actions you can cram into a given session, but also based on cherry-picking actions from an arbitrary list of things you've already done a lot of. That's seriously encouraging crazy behavior, where every character is OCD about their one thing, and does it at every opportunity.

Sure, I can see that.

However, I think implementing some kind of system wherein you can't gain XP from an action if that action has some "difficulty rating" or something that is lower than your "take average," would go a long way to solving this.

You're right - that will cause people to seek out things to challenge them so they can get more XP for their good skills. Sounds kinda like how it works in the real world, IMO: mountain climbers keep searching for bigger and bigger mountains to climb. Seems fairly sensible to me, though if you see a problem I'm missing, feel free to point it out.

[Dean]

Almost any time you put yourself in a design space where you need to figure out creative ways to prevent or punish people from taking the actions that you have decided to reward you have gone awry. The results of your difficulty rating adjustment won't make mountain climbers seek out higher mountains it will make them smear themselves in jelly and climb trees because that turns out to be the exact difficulty they want. It is a better space to be in where you incentivize people to do what you want than to incentivize people to do something you don't want and then try to outsmart them from doing it.

[GnomeWorks]

The results of your difficulty rating adjustment won't make mountain climbers seek out higher mountains it will make them smear themselves in jelly and climb trees because that turns out to be the exact difficulty they want.

I think that would depend on how it's implemented.

It is a better space to be in where you incentivize people to do what you want than to incentivize people to do something you don't want and then try to outsmart them from doing it.

Fair.

However, I don't think that modifying the system in this manner - adding an upper bound to encounters and obstacles that says, "any skill rated higher than this, no XP for using it here" - is particularly egregious.

I want to make this model work. Adding a bit more complexity to make it work (1) does not seem that difficult, thus far, and (2) still looks better to me than changing to a different system entirely.

[fectin]

So, what do you want people in your game to do?

(on the likelihood of rolling over, you're probably calculating average wrong. Remember, for 1d10 the average roll is 5.5, not 5. Adding two dice together leaves you rolling the exact average some amount of the time, and for smaller dice that comes up more often.)

[GnomeWorks]

(on the likelihood of rolling over, you're probably calculating average wrong. Remember, for 1d10 the average roll is 5.5, not 5. Adding two dice together leaves you rolling the exact average some amount of the time, and for smaller dice that comes up more often.)

...playing around with the statistics again, it would seem that the chance of rolling above average is dependent on the size of the dice. That is odd that that had escaped my notice; I'm intrigued by the mathematical reasoning behind that.

I would have assumed that the chance for rolling above average for d8+d2 was identical to the chance for d6+d4, but that's not the case. I suppose it's probably due to what you said: for smaller dice, the chance of hitting exactly the average decreases, leaving more room for over/under.

So, what do you want people in your game to do?

I just finished watching The Grey, and found myself thinking the entire time that that entire movie should be reproducible in my game. And not just with MTP-style hand-waving, I mean that everything in the film should be reproducible with our mechanics, to the point of thinking about writing it up as a playtest scenario.

Ultimately, I guess I'm looking for something akin to D&D, but significantly lower power. This game began as a hack to 3.5, but then I got tired of dealing with a lot of 3.5's issues, and started working on something from the ground up.

More than that, though, I also want to encourage a particular style of play. I want a West Marches-style sandbox to be the norm, as opposed to Paizo's adventure paths.

Does that sufficiently answer the question?

[Grek]

I'm intrigued by the mathematical reasoning behind that.

The outcome table for d8+d2 is as follows:

Roll	1	2	3	4	5	6	7	8
+1	2	3	4	5	6	7	8	9
+2	3	4	5	6	7	8	9	10

while the d6+d4 table looks like:

Roll	1	2	3	4	5	6
+1	2	3	4	5	6	7
+2	3	4	5	6	7	8
+3	4	5	6	7	8	9
+4	5	6	7	8	9	10

With above average results underlined.

The average of 1dX + 1d(N-X) approaches a maximum as X approaches 0.5N for much the same reason that x*(n-x) approaches its maximum as x approaches 0.5n

[fectin]

More than that, though, I also want to encourage a particular style of play. I want a West Marches-style sandbox to be the norm, as opposed to Paizo's adventure paths.

Does that sufficiently answer the question?

Ultimately, people will do the thing that makes you give them xp (or whatever other character advancement).
- If you want characters to run about stabbing things, you give them xp for stabbing things.
- If you want them to explore, you give them xp for exploring.
- If you want them to create their own goals and then accomplish them, you give them xp for that.
So, to figure out what you should award xp for, figure out what you want your players to do.