Worlds of Design: Always Tell Me the Odds

aramis erak · May 27, 2020

clearstream said:
It seems like the purpose of dice pools might be exactly contrary to the assumptions of the OP, whose analysis appears built upon a notion that players should, and need to, know their odds. The question it makes me ask is, why? Why does it really matter that players know their odds? I could come back with a notion that they should not know their odds, but only when their chances are improved or worsened, and when they are stronger or weaker. I could even think that knowing the odds in any exact sense is a chimera.

The first few dicepool-count-successes I encountered (Warhammer Fantasy Battle, Warhammer 40K Rogue Trader, Space 1889, Shadowrun, VTM 1E) were all using them as skill dice vs difficulty number. WH40K:RT and WHFB used tables to hide the difficulty calc, but many players intuited it. Space 1889 was co-released with a pair of minis games (and a hexmap version of one of them), and the combat mechanics are pretty close between them. Shadowrun and VTM are just a bit after as S:1889, and I know the guys at FASA knew Frank Chadwick of GDW... because we see the FASA guys in the credits of many Traveller products.

While the probabilities are a bit opaque, the intent stated by Frank (in Space: 1889) was to keep it compatible with the minis games and board game. It's a standard of minis games to roll a pool vs a fixed TN by die to resolve a unit-on-unit action, and so the combat systems in the RPG are exactly the same as in the other games.

(This same rationale is apparently why the various mechwarrior games do math to make the RPG rolls 2d6 roll high against a TN by skill, where the skill goes down as the character improves... it's because BattleTech does that.)

It's worth noting as well: Chainmail also used a pool of dice vs a TN matrix of Attacking troop type by Defending troop type.... indicating both how many men per die, and the individual results on the d6 which are needed to kill one target. This is fairly standard stuff in minis games. It reentering in the late 80's after TSR abandoned it in 1977... almost an inevitability. And the D&D minis rules in Dragon E-Zine use it, as well...

clearstream said:
The OP offers a one-dimensional analysis of the subject, notwithstanding that I appreciate the thought and effort in writing, and that it has prompted an interesting discussion.

I don't agree that there's evidence of effort in writing. I do agree that the resulting discussions are worthwhile. I'd concede that some effort may have been made to find a controversial subject.

I also agree that knowing the odds really isn't always a good thing from a story standpoint. There are times when it's great, and times when it's not so great.

clearstream · May 27, 2020

Hussar said:
But, again, if we're now adding negative consequences in addition to increased chances of failure, then the reward also needs to be increased as well. Say, sure, it's only one Acrobatics check to swing from the chandelier, but, if you fail, you land prone, then how much, in addition to allowing the attack, do we have to reward the player in order to entice them to actually do it?

I think a DM should include and permit players to make bad wagers. Imagine four worlds

a) Using chandeliers pays out better. Outcome = always use them. There is no interesting choice here, and narratively it will become repetitive.

b) Using chandeliers pays out fairly. Outcome = it is a matter of style, but you might as well always use them. The player takes no real risk and gains cachet.

c) Using chandeliers pays out differently. Outcome = it is a matter of circumstance. In the right circumstance, always use them.

d) Using chandeliers pays out badly. Outcome should be = avoid using them. There are a few ways it could still be desirable. Say it is crucial that the BBEG goes down this round. It can be worth the penalties just to have whatever unfairly meagre bonus the DM has put on offer. Or say a player just wants to look flashy. As hinted at in b), looking flashy has worth on the social and narrative dimensions of play. Just because that worth isn't measurable in game doesn't mean that it isn't on offer.

The reason why swinging on a chandelier is exciting is because it is flashy and inefficient. Taking Suits' definition as admonishment, it is a "voluntary attempt to overcome unnecessary obstacles" where less efficient rather than more efficient means are adopted. It is the very heart of play! Games are played for stakes, and the most significant stakes are those outside the circle of play: in the stories we create. The sweet spot might well be exactly where the odds are rationally dislikeable... if I can pull that off I'm going to feel something that I could never feel otherwise!

clearstream · May 27, 2020

aramis erak said:
The first few dicepool-count-successes I encountered (Warhammer Fantasy Battle, Warhammer 40K Rogue Trader, Space 1889, Shadowrun, VTM 1E) were all using them as skill dice vs difficulty number. WH40K:RT and WHFB used tables to hide the difficulty calc, but many players intuited it. Space 1889 was co-released with a pair of minis games (and a hexmap version of one of them), and the combat mechanics are pretty close between them. Shadowrun and VTM are just a bit after as S:1889, and I know the guys at FASA knew Frank Chadwick of GDW... because we see the FASA guys in the credits of many Traveller products.

Now that you draw my attention to it, I find it plausible that dice pools have such origins. As you suggest, to see games like 40K with their handfuls of dice generated by figures provides us with some solid examples.

aramis erak said:
I don't agree that there's evidence of effort in writing. I do agree that the resulting discussions are worthwhile. I'd concede that some effort may have been made to find a controversial subject.

I have so far found this series pretty surprising in that treatment of the material seems very light. Where success is required on every trial a DM can simply multiply, but a DM is seldom in that world. A more common case - combat - sees players looking for X successes in less than Y trials (where at Y they will be dead). A take-away from compounding odds applies to some instances of skill use, and could help DMs better understand why stealth rolls should be let ride until broken. So there's that.

Maybe finding the subject has been the key thing. And FTM maybe the depth explored is chosen for an audience? I mean, why go into deeper discussion of game design when that might not be of interest to everyone?

Hussar · May 28, 2020

billd91 said:
As long as you’re not stacking on lots of chances for it to fail by requiring too many skill checks, I’m not sure the risk:reward calculation needs to be too big a deal, particularly if you think relatively small and simple. Wanna swing from the chandelier to attack? Give me a DC 15 Dex (Acrobatics) check - if you succeed, you get advantage on the attack (which for a rogue implies SO much more), fail and you attack with disadvantage.
Life‘s easier if you just keep things simple.

Ok, let's presume, for a moment, that we're talking around a 5th-8th level game. Fairly bog standard as things go. So, barring the rogue with expertise or some sort of magic item or whatnot, by and large, a character is going to have about a +7 (ish) Acrobatics check. (Yeah, 20 Dex and proficiency gives us, what, +9? +10?, so, let's benchmark it at +7) So, a DC 15 check means that I have a 60% chance of success. And success effectively gives me +5 on my attack. However, I still fail 40% of the time and that gives me a -5 (effectively) on my attack. Is this actually a balanced risk:reward? Honestly, yeah, it's probably close enough that it's not too bad.

OTOH, with a DC of 15, that basically means that any character not proficient shouldn't even consider trying. So, where do we set the bar? See, 5e describes DC 15 as Moderate. Is that where we want to benchmark? Or, do we want everyone to give it a go, but, those with proficiency are actually better at it? That's obviously going to depend a lot on what kind of campaign we're going for. In my Ghosts of Saltmarsh game, which is supposed to be swashbuckling, rollicking adventure, then, probably, the DC would be better set at about 10 so, everyone gets a shot at trying this but, those that are acrobaticly leaning can try it with a fair certainty (although, still not automatic) of success.

I guess, the point I'm trying to make here is that context matters. A lot. Many DM's, IMO are going to set DC's higher than maybe they should. I've seen people set this DC as high as 20 in other conversations. Or, increase the penalty for failure- failed check means no attack at all and you fall prone. Things like that. @Bill91's example is probably pretty good, honestly, depending on what feel you are going for. But, IMO, it's better to err on the side of easier, than harder.

Hussar · May 28, 2020

Now, here I do disagree somewhat.

clearstream said:
I think a DM should include and permit players to make bad wagers. Imagine four worlds

a) Using chandeliers pays out better. Outcome = always use them. There is no interesting choice here, and narratively it will become repetitive.

Maybe. Depends on the penalty. Players are often very, very risk averse and what you might think of as a better payout, still carries the risk of failure and thus the penalty. Unless, of course, we're talking about a better pay out that is SO good that it would be stupid not to take it. Obviously, that's bad too. Generally speaking though, it's very rare that this happens.

b) Using chandeliers pays out fairly. Outcome = it is a matter of style, but you might as well always use them. The player takes no real risk and gains cachet.

c) Using chandeliers pays out differently. Outcome = it is a matter of circumstance. In the right circumstance, always use them.

"Fairly" is a very nebulous idea. And, what is "fair" will vary a lot from table to table and game to game. Honestly, for me, so long as it's in this ballpark though, we're probably doing the right thing as a DM.

d) Using chandeliers pays out badly. Outcome should be = avoid using them. There are a few ways it could still be desirable. Say it is crucial that the BBEG goes down this round. It can be worth the penalties just to have whatever unfairly meagre bonus the DM has put on offer. Or say a player just wants to look flashy. As hinted at in b), looking flashy has worth on the social and narrative dimensions of play. Just because that worth isn't measurable in game doesn't mean that it isn't on offer.

To me, though, this is just bad game design. It's not rewarding smart play. The smart play would be to not do this since the payout isn't worth the risk. If it does succeed, well, it's just luck, not skill or particularly good play on the part of the player. Yay, you rolled a high number!! And, since the player will almost certainly fail, and failure in this case is probably catastrophic - as in whatever the player was trying to do is no longer possible anymore (the guards are alerted, the bad guy escapes, whatever) it's pretty much a suckers bet. Really, it's the DM forcing outcomes on the table. "I don't want the players to do X, but, if I outright say no, then I'm a railroading DM and that's bad. So, instead, I'll give them a slim chance of success and when that fails, everyone's happy."

I really, really don't like this approach.

The reason why swinging on a chandelier is exciting is because it is flashy and inefficient. Taking Suits' definition as admonishment, it is a "voluntary attempt to overcome unnecessary obstacles" where less efficient rather than more efficient means are adopted. It is the very heart of play! Games are played for stakes, and the most significant stakes are those outside the circle of play: in the stories we create. The sweet spot might well be exactly where the odds are rationally dislikeable... if I can pull that off I'm going to feel something that I could never feel otherwise!

However, "Pull that off" doesn't require the odds to be so bad that failure is the most likely result. There's a reason 5e has gone with bounded accuracy. It's a lot more fun to hit things than have multiple rounds of whiffing. Setting the PC up to whiff the attempt typically doesn't lead to interesting stories. It leads to player frustration and then the player never attempting anything that's not rules defined afterward. If outside of rules defined actions fail more often than rules defined actions, then it's perfectly rational NOT to attempt actions outside of rules definitions.

Heck, there's a reason you see players who would rather start dropping spells rather than attempt anything with skills. Spells have concrete, defined effects. Skills are timey wimey vague bundles of possibilities. Why bother with a disguise when a Change Self spell is so much better? Why bother trying to sneak when Pass Without a Trace means the entire party sneaks pretty much automatically? That sort of thing.

clearstream · May 28, 2020

Hussar said:
Maybe. Depends on the penalty. Players are often very, very risk averse and what you might think of as a better payout, still carries the risk of failure and thus the penalty. Unless, of course, we're talking about a better pay out that is SO good that it would be stupid not to take it. Obviously, that's bad too. Generally speaking though, it's very rare that this happens.

I mean that it pays out better. There is no risk. A bonus without any penalties, for example.

Hussar said:
"Fairly" is a very nebulous idea. And, what is "fair" will vary a lot from table to table and game to game. Honestly, for me, so long as it's in this ballpark though, we're probably doing the right thing as a DM.

This seems to quibble definitions, which isn't what we are about here. Speaking to your earlier example, fair would be where any penalty is balanced by gains in proportion. The expectation over time is undifferentiated for using chandeliers.

Hussar said:
To me, though, this is just bad game design. It's not rewarding smart play. The smart play would be to not do this since the payout isn't worth the risk. If it does succeed, well, it's just luck, not skill or particularly good play on the part of the player. Yay, you rolled a high number!! And, since the player will almost certainly fail, and failure in this case is probably catastrophic - as in whatever the player was trying to do is no longer possible anymore (the guards are alerted, the bad guy escapes, whatever) it's pretty much a suckers bet. Really, it's the DM forcing outcomes on the table. "I don't want the players to do X, but, if I outright say no, then I'm a railroading DM and that's bad. So, instead, I'll give them a slim chance of success and when that fails, everyone's happy."

Were it the only option available to players then it would be railroading, but where do we say that? In any circumstance there are a number of - let's call them - offers that a DM expresses to players or will express if asked. So players have before them an array of offers. The least number in your example was three: use the chandelier, use a normal attack action, do nothing. It's up to them which offer they avail themselves of. Remember though that the rational player you are advocating for must always accept the best offer: if that is chandelier, then it is always chandelier.

There's no reason why chandelier should be more or equally efficient to a normal attack. I think there are good reasons it should not be (else, if it is mechanically superior, and efficiency is the player's concern, then it should always be used). What I am saying is that players should not be focused on efficiency, and your concern overlooks the flipside of the OP's observations. The players will get multiple attempts. This won't be the last chandelier they ever see (well, unless it is literally the last chandelier they ever see... that can happen!) It might usually be a bad idea to chandelier, but sometimes circumstances demand risks be taken: the other characters are all down, one solid hit will drop the BBEG, one more round will see our heroine disintegrated.

Failing this time sets up next time to be that much more exciting. I think many DMs overlook this, but ask yourself a super-simple question: is a game more fun when players always succeed? No chance of failure? No real risks? No losses?

Hussar said:
Heck, there's a reason you see players who would rather start dropping spells rather than attempt anything with skills. Spells have concrete, defined effects. Skills are timey wimey vague bundles of possibilities. Why bother with a disguise when a Change Self spell is so much better? Why bother trying to sneak when Pass Without a Trace means the entire party sneaks pretty much automatically? That sort of thing.

Are we are in a world where players must have certainty over outcome? Notionally, skill use is indefinitely repeatable and spell use is not. One consequence of 5e RAW at many tables is that spells are readily replenished. The issue there is not risks, it is that an intended cost turns out not to be a cost.

CodeFlayer · May 28, 2020

CapnZapp said:
Sure.

As long as you aren't saying this to argue games don't have to have sound math, okay. Edit: which does not have to be true, either for you or the other poster.

In most games, if there is a chandelier object, there are specific rules and appropriate math as needed, all covered by the rulebook. They are bounded environments. RPGs are something very different, open ended, by involving things and situations that cannot be anticipated with any certainty(unless it is a large red button that says 'press me'), and makes for much of the intrinsic fascination and enjoyment that RPGs give. Thus, I see it as the GMs role to craft something interesting in an exchange or two with the player for something like the chandelier gambit, and then accept how the dice fall (as outlined by another poster). When I wrote of taking control of the simulation, I was referring to things like death mechanics so that, with the tables permission before the fact, we can explore as group what we want to happen next. The player may want to move on to another archetype, or the table may want to try a different campaign.

I would opine that if a GM were to decide to fudge a die roll, then, they should never speak of it. It breaks the magic.

Fenris-77 · May 28, 2020

Managing things like chandeliers in D&D is a little more complicated than in some other games (notably more fiction-first games) because of the D&D combat system. Well, to be more specific, the way the D&D combat system is usually implemented at the table. The underlying mechanics of D&D don't really differentiate between combat and non-combat as much as people think. Both are based on d20 rolls with bonuses for skill and rising difficulty targets. Combat obviously has a much more robust set of mechanisms for adjudicating the outcome of actions, but the basic principles are the same. The main difference is the initiative roll. Initiative serves to somewhat arbitrarily divide the set of all possible actions and I think this quite often restricts how both players and DMs think about what possible and how to adjudicate the outcome.

I'll use a matched pair of examples to illustrate what I'm trying to say, and I'll stick with the chandelier swinging for both. So, example one, the character wants to leap from the balcony, swing from the chandelier over a group of the Duke's party guests, and knock over the burning brazier before the evil chancellor can drop poison into it and kill everyone at the ball with poisoned smoke. Fancy right? There are a ton of ways to adjudicate success and failure here, but the majority of them are going to involve one, or maybe two skill rolls. A swinging roll and a booting roll, or maybe even just a single roll for both. I don't think it's controversial to say this isn't hard for a DM to manage. Sometimes this shizz just writes itself. The action is cool, it moves the story forward, and it provides a wealth of hooks to complicate things in the event of failure - it's a pretty ideal role playing action. If I were adjudicating this, I probably wouldn't even make it all that hard. It's pretty fancy, so obviously there needs to be some chance for failure, but I wouldn't set the bar super high as I don't want to disincline the players from trying this stuff in the first place. Some GMs might opt for more rolls and higher DCs than I would, but I think those DMs are doing both their game and their player a disservice. On to example two.

So, in example two we're going to keep most of our moving parts, but change a couple of key things and take a look at how that changes adjudication. In this example, the character is trying to escape the duke's manor after being discovered ransacking the evil chancellor's office (looking for evidence of malfeasance, no doubt). The PC races out onto the same balcony, overlooking the same group of party guests, the burning brazier, and the evil chancellor twirling his villain mustache. The PC has discovered that the Chancellor's secretary is bringing the poison to the ball, and is in fact due to arrive any moment. And lo and behold, the Secretary appears at the top of the stairs leading up from the dungeon. The secretary cannot reach the brazier or chancellor or bad things will happen! Our swashbuckling PC announces that he's going to leap from the balcony, swing from the chandelier, and boot the secretary back down the stairs, preventing him from delivering his vile burden. Sweet! This is some real Dumas stuff, and as a DM I'm still excited, but this is where it start to fall apart in some games. The culprit is the initiative roll.

In our second example, the action has reached a decision point for the DM. The most common answer is this: What are we doing here? Well, we're booting the secretary down the stairs, so that's combat, so the answer is roll initiative, right? I'd argue that this is the wrong choice. Moving from narrative play to what I'll call combat play introduces a strict order of operations, and a bunch of mechanics, and that's the first stumbling block here. Now we have to figure out if the secretary is surprised, or not, and then we have to roll to see who goes first. If the secretary goes first the whole plan is in the cacky because he'll obviously move away from the top of the stairs. Here's my question, why ruin a perfectly good plan by adding an additional role that can only serve to add additional chances of failure? There are some additional problems as well, also introduced by the decision to roll initiative. Because we are now in combat mode, most DMs look in a different toolbox to decide how things go. Now we're talking about attack type, armor class, damage, and a much stricter set of rules about applying conditions like knock back. Maybe the secretary should get a Dex save to avoid the boot? Holy crizzap, entities are multiplying like coat hangers in a closet here. Now we're looking at a minimum of three rolls, initiative, one to swing and an attack roll, never mind a possible saving throw. Regardless of difficulties, this whole idea just got a lot more complicated and the chances of success went way down and the number of moving parts the DM has to consider went up.

Obviously you can adjudicate the chandelier swing just fine within the combat rules, but you can't escape the extra failure state that comes with the initiative roll. A snazzy DM can probably navigate the additional rules layers without too much holdup or thought, but a new DM could get bogged down a little because the action in question doesn't really fit neatly into the combat rules, so there's also some potential lag while that DM crunches numbers in his head. Hitting the pause button before you call for the initiative roll creates a moment in the narrative, a moment within which you can still work outside the combat rules, and you can keep the narrative flowing. The PCs get a moment to opt in or out of combat, and a moment to act while they are still in charge of what happens next. That moment is a valuable tool, and a lot of DMs regularly miss it because they shout Roll for initiaive! a beat too quickly.

billd91 · May 28, 2020

Fenris-77 said:
So, in example two we're going to keep most of our moving parts, but change a couple of key things and take a look at how that changes adjudication. In this example, the character is trying to escape the duke's manor after being discovered ransacking the evil chancellor's office (looking for evidence of malfeasance, no doubt). The PC races out onto the same balcony, overlooking the same group of party guests, the burning brazier, and the evil chancellor twirling his villain mustache. The PC has discovered that the Chancellor's secretary is bringing the poison to the ball, and is in fact due to arrive any moment. And lo and behold, the Secretary appears at the top of the stairs leading up from the dungeon. The secretary cannot reach the brazier or chancellor or bad things will happen! Our swashbuckling PC announces that he's going to leap from the balcony, swing from the chandelier, and boot the secretary back down the stairs, preventing him from delivering his vile burden. Sweet! This is some real Dumas stuff, and as a DM I'm still excited, but this is where it start to fall apart in some games. The culprit is the initiative roll.

In our second example, the action has reached a decision point for the DM. The most common answer is this: What are we doing here? Well, we're booting the secretary down the stairs, so that's combat, so the answer is roll initiative, right? I'd argue that this is the wrong choice. Moving from narrative play to what I'll call combat play introduces a strict order of operations, and a bunch of mechanics, and that's the first stumbling block here. Now we have to figure out if the secretary is surprised, or not, and then we have to roll to see who goes first. If the secretary goes first the whole plan is in the cacky because he'll obviously move away from the top of the stairs. Here's my question, why ruin a perfectly good plan by adding an additional role that can only serve to add additional chances of failure? There are some additional problems as well, also introduced by the decision to roll initiative. Because we are now in combat mode, most DMs look in a different toolbox to decide how things go. Now we're talking about attack type, armor class, damage, and a much stricter set of rules about applying conditions like knock back. Maybe the secretary should get a Dex save to avoid the boot? Holy crizzap, entities are multiplying like coat hangers in a closet here. Now we're looking at a minimum of three rolls, initiative, one to swing and an attack roll, never mind a possible saving throw. Regardless of difficulties, this whole idea just got a lot more complicated and the chances of success went way down and the number of moving parts the DM has to consider went up.

<snip>That moment is a valuable tool, and a lot of DMs regularly miss it because they shout Roll for initiaive! a beat too quickly.

Honestly, I think you're fretting the initiative roll too much. Yes, there are all sorts of ways rolls can complicate things. But just because the roll indicates the secretary may go before the hero, that doesn't mean the hero's plan is squashed. After all, the whole point of swinging down from the balcony is intended to be a surprise, right? So that secretary is going to assume the hero will come down the normal way. He may spend his action, assuming he even sees the hero (which we'll assume he does otherwise we'd just be giving the hero the drop on him because of the surprise since the secretary is trying to look unrushed and nonchalant - all the better to poison a bunch of unsuspecting party-goers), catching the eye of any guards on the party level to send them up to stop the hero from using the stairs.

If you turn to initiative to help time out the adjudication of actions, you don't have to play it like everyone gets everything they want accomplished on their turns, particularly when the players are angling to do something surprising like swinging on a chandelier.
And that's one reason I think you don't want such tactics to be too common, even in a swashbuckling game. If the PCs are doing it all the time, it's no longer a surprise. And when it's no longer a surprise, why aren't people expecting it and countering it? By all means, if a PC is fishing around for something in desperation like "Oh, my god! The secretary is coming into the ballroom with the poison. I gotta get down their fast!" let them know "Well, there's a chandelier hanging about midway between you and him..." But if the PC is known as a chandelier swinger, time to shake him out of his comfort zone and do something more creative.

Ulfgeir · May 28, 2020

Another point is when you do skills. You might not want or need the exact value, but you should have a sense of degree of whether or not that is something you even have a remote chance of succeeding at. Especially if it is something that is part of your occupation.

In D&D and similar games it will often be that unless you have maxed out your stat and skill-ranks for that skill, and you have a ton of magical items to help you, you have no chance of succeeding because you are too low level (unless you roll a natural 20, sometimes not even then), or some stuff that you cannot fail once you get your stats high enough.

In D&D or BRP-based games it is relatively easy to have a sanity-check on your chances (That does not mean that the value you need to hit is anywhere near sane though). In a game like the FFG Star Wars or other games with pools, it is much more difficult.