Hi all,

This is my first message here at the RPG site.

I'm trying to design an RPG system, and I've been thinking about what I consider the most fundamental aspect of any game: the success/failure ratio. (I'm not saying that's the ONLY thing that keeps players interested, I'm just saying it's an important aspect.)

Now, this may be stupid, but I believe there is a range of success/failure ratios that keeps players invested in the game.

You succeed all the time? Game becomes boring. You fail all the time? Game becomes boring. You succeed/fail exactly at 50/50? Boring. (You might as well toss a coin.)

So, is the "interesting" success/failure ratio more towards the 30-40% range, or the 60-70% range?

Some games, such as Old School DnD, seem to put low-level characters at the lower range of the spectrum and then advance them towards the upper range. But that, in my experience, makes playing low-level characters a chore.

But in a game where characters start out at the 60-70% end, character development is rather pointless.

Any thoughts on this?

Cheers,
Jerzy

Yeah the difficulty of the task dictates the chances of success in combination with the skill of the character making the attempt. A flat percentage chance doesn't work, that needs to be modified by the task. So more skilled characters have a better chance of success than less skilled characters at difficult challenges.

My experience has been that players will stick with it if challenges require them to use their heads beyond just rolling a dice (ie wearing soft wool footpads rather than boots to help sneaking, hence the term footpad), so make sure the system allows for adjustments to rolls on that basis.

Welcome to the RPG site by the way, anything goes here but don't bother starting any setting riffs, the elders don't have no truck with no creativity.

I've read that big  game companies (board games, etc) have done research on this, and determined that the magic number is about 66%.

It feels a bit too generous to me, but it might be a good rule of thumb.

Quote from: jerzyab;687566You succeed all the time? Game becomes boring. You fail all the time? Game becomes boring. You succeed/fail exactly at 50/50? Boring. (You might as well toss a coin.)

So, is the "interesting" success/failure ratio more towards the 30-40% range, or the 60-70% range?

I think it's worth differentiating between your chance of success averaged over all your die rolls, and your chance of success on any one die roll.

It could be, for example, that you have wildly a different success chance for each task, but the average is exactly 50/50.

In some games, you try to marshal as many advantages as possible to maximize your chance of success.

Amber and Dungeon World are interesting counterpoints, because in both of these games the chance of success is fixed (Amber at 100%/0%, depending on your stat and the balance of advantages). In Dungeon World advantages are enormously important, but these are used to influence what success and failure mean, so the chance of success (whatever success for that situation) remains identical from task to task.

My offhand answer would be 'about 70%' I guess.
However, I think its handy to have extra degrees of success to leave a margin for character growth.

What does margin of player growth mean exactly here? When you level up, your numbers get bigger, but so do the challenges. If you can jump further and hit harder, you need bigger cliffs and things that can hit harder back. Scaling is fine as long as the thematic difficulty is scaling as well. Why would the challenges get easier as players levelled? The challenges should be advancing with them.

Quote from: Doccit;687650What does margin of player growth mean exactly here? When you level up, your numbers get bigger, but so do the challenges. If you can jump further and hit harder, you need bigger cliffs and things that can hit harder back. Scaling is fine as long as the thematic difficulty is scaling as well. Why would the challenges get easier as players levelled? The challenges should be advancing with them.

Challenges are what they are in my games. If 1st level characters decide to track down an ancient red dragon and fight it, the dragon is most likely going to butcher them. On the other hand, if a party of 12th level characters attack a small group of kobolds, the kobolds are likely going to be butchered. The level of the party does not change the opposition. The ancient red dragon isn't suddenly weaker because a first level party stupidly decided to confront it nor are the kobolds going to get more powerful just because a group of high level PCs encounter them.

I run a sandbox world and what PCs encounter depends on where they go and what they do -- the world doesn't care about their level.

Quote from: RandallS;687659Challenges are what they are in my games. If 1st level characters decide to track down an ancient red dragon and fight it, the dragon is most likely going to butcher them. On the other hand, if a party of 12th level characters attack a small group of kobolds, the kobolds are likely going to be butchered. The level of the party does not change the opposition. The ancient red dragon isn't suddenly weaker because a first level party stupidly decided to confront it nor are the kobolds going to get more powerful just because a group of high level PCs encounter them.

I run a sandbox world and what PCs encounter depends on where they go and what they do -- the world doesn't care about their level.

I'm with you there, but it can still be useful to a GM to know what level the characters (Edit: PCs or NPCs) "should" be to face a threat, or what rank they should have in a given skill to have a fair shot at success. It doesn't mean that you have to run a game catered to the player's levels.

It's useful for setting coherence; if you say that three red dragons have lived in an area for 100 years, and there's a village that has somehow survived there, you know that the people who defend that village are likely comparable in power level to the threat, and it has resulted in a stalemate.

Of course the dragon isn't going to get any weaker. I'm not saying it should, but I'm assuming that the player's aren't going to be asked to kill a dragon at levle one. I'm not really sure what makes something a sandbox game in the context of a table top, because if it means "the players can choose to do anything they want", then I don't think I've heard of someone running a game that wasn't a sandbox.

I don't think you're asking how often players should be able to succeed at any random task chosen among everything conceivable. Level one characters should be able to do level one tasks x% of the time, and level two characters should be able to do level two tasks x% of the time. If a level one player tries to succeed at a level two task, they'll be successful less often.

Are you asking how often a level one character should succeed at a level one task compared to a level two character? That all depends on how big thematically you want the difference between your levels to be. In a game where you start off as a peasant and turn into a god in ten levels, at task that you've got a 60% chance at at level one you might have a 100% chance at at level two. In a game where you start as a peasant and turn into a knight in ten levels, perhaps 65%.

Quote from: Doccit;687650What does margin of player growth mean exactly here? When you level up, your numbers get bigger, but so do the challenges. If you can jump further and hit harder, you need bigger cliffs and things that can hit harder back. Scaling is fine as long as the thematic difficulty is scaling as well. Why would the challenges get easier as players levelled? The challenges should be advancing with them.

I find having to make everything 'level appropriate' is something of a burden on the GM - a metagame expectation I don't really want to design adventures around.

In part I guess a higher level PC faces some higher-level challenges (things that a lesser PC will avoid because it would likely kill them) but there are still going to be 'lesser' challenges because they happen to find some orcs in the forest, or need to climb a tree for some reason.
Having degrees of success then is good since the PC can attempt a lower level task and still get extra benefit from skill increases - so they might do a task faster.
Similarly, its nice if the game lets a PC take a difficulty penalty to do something more impressive; its an incentive for the PC to adjust their numbers backward a little toward the 'level appropriate' chance of failure.

Quote from: Doccit;687650What does margin of player growth mean exactly here? When you level up, your numbers get bigger, but so do the challenges. If you can jump further and hit harder, you need bigger cliffs and things that can hit harder back. Scaling is fine as long as the thematic difficulty is scaling as well. Why would the challenges get easier as players levelled? The challenges should be advancing with them.

If you are only building encounters based on the PCs levels, in order to have appropriately challenging encounters, you're not doing a sandbox.

In a sandbox scaling happens because of the choices the PCs make, not the choices the DM makes. If the DM places a lich in hex 7, then there is a lich in hex 7, and 1st level PCs may encounter it if they wander around hex 7*. A successful resolution for them would be getting away alive. At 10th level they might deliberately return, because they think they are now ready to take it on.  The same goes for the band of goblins in hex 10, a level appropriate challenge they might seek out at 1st level, but they can also encounter them at 10th level, in which case it's the goblins that get TPK'd.

*If they are wandering around blind, i.e. they haven't checked with locals for rumors of threats, and they get TPK'd, well, that's what taverns, rumors, and Charisma for that matter, are for.

Quote from: Monster Manuel;687662It's useful for setting coherence; if you say that three red dragons have lived in an area for 100 years, and there's a village that has somehow survived there, you know that the people who defend that village are likely comparable in power level to the threat, and it has resulted in a stalemate.

That's certainly one possibility for why the village is still there, but a fairly unlikely one for my worlds (as the dragons probably would move rather than remain in a long term threatened area).  Other possibilities include the village owns fealty to the dragons, the village makes regular sacrifices to the dragons, the village acts as an early warning system for the dragons by sending them a quiet message when they see adventurers heading for the dragon lair, the dragons have been asleep for many years and the village grew up there not even knowing the dragons were still around, etc.

On the original subject of this thread, I find the concept of some type of success/failure rate one can design around alien. Success in my campaigns is based as much (or more) on what the players have their characters do as what is written on the character sheets. According to their character sheet, a low level thief might only have a 20% chance of walking up to someone and picking their pocket.

However, that only means the character has an 80% chance of failure if that's all the character does: walk up and try to pick the victim's pocket. If the player talks other characters into distracting the victim, the chance of success will go up. If the character makes the attempt in a crowd where everyone is getting jostled, the chance will be higher, if there is that large crowd and  other PCs distract the victim, the chance of success will go up even more, etc.

In games were player skill is intended to be as or more important than the exact chances of success on the character sheet, the high starting values suggested by game theory (such as the 66% chance mentioned earlier in this thread) would basically mean auto-success if the players used player skill in the way I mentioned in the previous paragraph.

However, my main objection to this idea is that it often leads to "always fighting orcs" if the campaign is set up around level-based encounters (which games designed like this often are today). Level advancement really does not matter because no matter what level the characters are, they will generally have the same 66% (or whatever) chance of success because adventure designers are designing the challenges they face to their level.

Yes there will be different levels of challenges if players are just wandering around aimlessly, but when the players go into the tavern the GM is directing them through the taverners to what is level appropriate, right?

QuoteIf you are only building encounters based on the PCs levels

Do you mean building them in advance, or building them period? It only makes sense to plan for everything to go as planned (IE, the players are directed to this particular dungeon by the guy in the tavern, so you build that dungeon). It doesn't make sense to power up goblins or power down lichs if the players haphazardly run up to them, but the players fighting the lich isn't part of the plan. That is the sort of bridge I'd cross when I came to it.

And it isn't a sandbox if the lich isn't pre-built? What if the players decide to kill the quest giver? Does he need to have combat stats ready for that possibility? What if they want to burn down town hall? Should you have that building's HP pre-calculated?

Its a sandbox if the players can choose to fight the lich (given that it is there), but I don't think the GM needs to build it in advance of the players wanting to hit it. Pre building mostly the level appropriate ones only makes sense, because that is what the NPCs will be sending them to.

Quote from: Doccit;687768Its a sandbox if the players can choose to fight the lich (given that it is there), but I don't think the GM needs to build it in advance of the players wanting to hit it.

In my sandbox campaigns, my writeup of the area might say "Lich - 12th lvl MU" then  would then list its HP, saves, AC, spells, etc. in a few lines of text. Basically what I would need to run the Lich if the PCs encounter it. In TSR D&D, that's not a lot of info.

Quote from: Doccit;687768Yes there will be different levels of challenges if players are just wandering around aimlessly, but when the players go into the tavern the GM is directing them through the taverners to what is level appropriate, right?

Wandering around can be fun. But what I meant by rumors in the tavern isn't the GM directing them, but the players choosing based on the rumors. The tavern (or whatever) would have rumors of level appropriate and level inappropriate jobs/tasks/challenges/etc., the players choose which ones to take on.  

QuoteDo you mean building them in advance, or building them period? It only makes sense to plan for everything to go as planned (IE, the players are directed to this particular dungeon by the guy in the tavern, so you build that dungeon). It doesn't make sense to power up goblins or power down lichs if the players haphazardly run up to them, but the players fighting the lich isn't part of the plan. That is the sort of bridge I'd cross when I came to it.

The key word there is "only", which is why I italicized it. If all the encounters are level appropriate, it isn't a sandbox. And no, you don't need to pre-build anything; winging it is part of the fun of sandbox GMing.

QuoteAnd it isn't a sandbox if the lich isn't pre-built? What if the players decide to kill the quest giver? Does he need to have combat stats ready for that possibility? What if they want to burn down town hall? Should you have that building's HP pre-calculated?

I'll take this seriously, so the answer is if the players decide to do any of that, just wing it. You don't need to pre-stat anything, let alone everything. If you need to stat something up, take a couple seconds (or even a couple minutes) and throw something together. Once more for emphasis, JUST WING IT*.

That said, having a stable of stock characters that can be repurposed to the needs of the moment is helpful too. It makes winging it easier, unless they are difficult to look up, in which case just wing it.


*Not everyone is comfortable doing this, or likes it, so it isn't advice that fits every personality. RandallS's suggestion works as well. I'd probably just have lich, level 12, and make it up if I need to from there. My guess is RandallS would also wing burning down the town hall.

QuoteIts a sandbox if the players can choose to fight the lich (given that it is there), but I don't think the GM needs to build it in advance of the players wanting to hit it. Pre building mostly the level appropriate ones only makes sense, because that is what the NPCs will be sending them to.

You make this sound like the GM is speaking through the NPCs to send the PCs to level appropriate encounters. This needn't be the case.

Take the lich example. A tavern encounter could be with Jennifer Connelly's character from Labyrinth, who is looking for someone to help her rescue her baby brother from the lich. What do the 3rd level PCs do? Ignore her (could you ignore her? You can make it more difficult for them by showing them a picture of her as well.). Take the job and likely get killed? Go with her and try to protect her while keeping her way from the lich? Try and sneak in and sneak out with the brother when the lich is away (and how do they get their hands on the lich's appointment book anyway)? Turn her down, and watch her go off on her own, only to encounter her as a vampire minion of the lich 10 levels later? Or maybe she goes off on her own and succeeds, with a bard writing a song about it making the PCs laughingstocks in the area.

Or say it is the goblin king who has kidnapped her brother; a 3hd goblin with a band of 30 normal goblins, and she begs the 12th level party for help. Do they ignore her because goblins are small potatoes now, or do you spend a session hacking up screaming and fleeing goblins? Bear in mind that even though the goblins would be a walk-over of an encounter, getting the brother killed means the mission fails, so a head on approach wouldn't be a good idea anyway. (On the other hand, a big enough sleep spell wouldn't even be an encounter.)