5E Consequences of Failure

[ClaytonCross]

My definition is that there must be sufficient risk that a player has to weigh it against the benefit. Failure should come with at least a twinge of regret for having rolled.

“Failure to earn a reward” wouldn’t seem to qualify.

understood. That's why I asked. If that is the case, there is the alternative of adding a default of assumed guilt and hostility to failures. Might be hard to come up with reasoning for that on the spot every time and if enemies would attack you on sight already that doesn't function. So at that point it seems like a GM best effort based on the situation and what you can come up with on the spot. Its kind of like you mentioned in original post, it might be a good rule of thumb, but its a tool you have to learn to use over your past habits and being called on the spot. ... It might be better considered more of guideline than an actual rule, where you try to implement at best effort on the spot to establish a new habit, but trying to force it might be an impediment in actual implementation. In hind sight with all the time in the world you will get a lot of great advice on a per case basis. but in the moment you may have to bail out of the attempt from time to time. As long as your practicing with the attempt you will likely get better at finding meaningful consequences or discarding the test.

… not really useful advice, but if people add suggestions for specific encounters you can add them to the tool box of your mind and hope you can grab a decent one when you need it. For you example this time, considering attempts of stealth acts of guilt by default would be useful sometimes.

 

[iserith]

I'll say this:

If your Apocalypse World games play like your Dungeons and Dragons game it is likely because you are ignoring a vast swathe of the game. The majority of what makes Apocalypse World Apocalypse World is GM facing. It does not have GM advice. It has GM instructions.

It tells you what your goals are. It tells you what your principles are. It tells you how to live up to those principles and goals.

When it tells you not to preplan a story, it reiterates with language not fit for this board. This is not unlike when other players are instructed to play their characters with integrity. Vincent Baker means it.

Tom Moldvay also makes no suggestions. He tells you how to be a referee.

These form a set of expectations that GMs are expected to uphold.

Yep, when people get excited about running, say, Dungeon World, and talk about it like it's D&D Light I'm like "Nuh uhhhhh." It's nothing like that at all outside of having some of the same fantasy tropes. The GM is greatly constrained by comparison and you can't be trying to run your plot-based games with it. It's a good deal harder to stick to the agenda and principles of that game than to just run some halfway decent D&D.

 

[Campbell]

Yep, when people get excited about running, say, Dungeon World, and talk about it like it's D&D Light I'm like "Nuh uhhhhh." It's nothing like that at all outside of having some of the same fantasy tropes. The GM is greatly constrained by comparison and you can't be trying to run your plot-based games with it. It's a good deal harder to stick to the agenda and principles of that game than to just run some halfway decent D&D.

Just from the tempo alone I'm generally more exhausted after a 3 hour session of Apocalypse World than a 5 hour session of Fifth Edition. When I first started I could not go longer than 2 hours.

I have a broadly similar reaction when someone tells me they want to try running Classic Dungeons and Dragons because they want to go back to when things were simpler. Then I tell them about gold as XP, reaction rolls, morale, initiative and casting rules, 10 minute turns, and how you have your dungeon map and players have theirs. Running B/X is significantly harder than running the modern game.

 

[iserith]

Just from the tempo alone I'm generally more exhausted after a 3 hour session of Apocalypse World than a 5 hour session of Fifth Edition. When I first started I could not go longer than 2 hours.

I have a broadly similar reaction when someone tells me they want to try running Classic Dungeons and Dragons because they want to go back to when things were simpler. Then I tell them about gold as XP, reaction rolls, morale, initiative and casting rules, 10 minute turns, and how you have your dungeon map and players have theirs. Running B/X is significantly harder than running the modern game.

I'd put my 4-hour D&D 5e game up against Apoc or Dungeon World of similar length for level of exhaustion by the end, and not just because of all the booze. We get stuff done and it's hard. The players earn their victories in these games and they keep me on my toes. At least with Apocalypse World and Dungeon World, I can outsource things to the players more than I can in D&D 5e. I could do that in D&D 5e and do with some things sometimes, but it's just not the right game for really leaning into that.

 

[ClaytonCross]

Interestingly, in your example of stealth, you move failure from just not being able to get surprise to the opposition becoming hostile. This isn't just denial of a reward, it's a change in fiction away from the intent of the action.

Personally, I think that evaluation of the goal in goal and approach is important for failure consequences. A success moves towards the goal, a failure away. This simple framework does a good job of removing questions and firmly rooting decisions in the fiction of the moment. It's what you've done with your stealth example, assuming the goal of the sneaking is to get the drop on the opposition. You succeed, you get the drop. You fail, they not only notice you, but immediately move into hostilities without you having the drop. Towards, away.

If you can't think of how to move away from the goal of the action, then don't call for a roll. I think it's okay to involve meta-goals, here, though. Moving away from a meta-goal can be sufficient punishment, like, say, losing time in a race against the clock, or having a fight when you're trying to conserve resources. Goals can be layers, but a failure should move away from a player stated goal.

Er…. you get that the "change" was a second consideration on how to handle the same thing, in case the OP doesn't consider the first way valid.....

"I am not sure "consequence of failure" trade of thought includes "benefit for success" by your definition. If it does than the question needs to be "Does failure have meaningful consequences and/or success have meaningful benefit over not attempting the action?" If the answer is yes, it seems reasonable to call for a roll since, failure to gain a benefit is also a consequence. Though, I can defiantly see scenarios of "and" being more interesting than ether/or."

Example 1. was a "denial of benefit" equals a "consequence of failure", do you need more?
Example 2, was if you do not consider "denial of benefit" as a consequence justification for role by your definition, perhaps having NPCs react to the action of the check on failure would be a way to grant players the role by creating and for warning the consequences.

Example 2 was stated as a possibly ALTERNATIVE, incase me and the OP disagreed on what can constitute a consequence. This lets me address the intended scenario in question in a single post regardless of the OPs interpretation of "consequence of failure". It does not "change" my original response, however, I felt I should respond to both trains of thought so its clear that my question was sincere and I was not detailing the OPs method as "wrong" but just trying to make sure I was clear of the intent of the question.

I don't see a problem with "A success moves towards the goal, a failure away." as general guidline however, I don't really feel the need to hold it as a hard rule. Sometimes the success is just not failing and failure is just not success, however in order for a role to have meaning at least one of them must be significant to the progress of a goal. For example, with jump over a deadly trap pit trap, if they succeed the jump and can clear the trap, they just continue on their way having avoided danger. If they fail, then they fall in the trap they die. They don't get set further from danger because that was make them safe. Success here is simply not failure, but the consequence of failure is sufficient enough that the role is meaningful. Could they look for another way around? Sure, but if they choose to jump and except that risk, I am not going to stop them from jumping and will give them the role. I am not going to give them a bonus for not dying if they making it though unless I told them in advance there was some benefit to jumping the whole. Sometimes players just want there characters to go the scarier path because they want there PC to look death in the face and laugh.... some times death says "welcome to the after life" to the PC when they do. Alternatively, a PC be short on time but see a valuable gem stuck in the wall and want to try and take a hammer and knock the gem out, grab it and keep running. Success means they swing knock them gem out and pick it up. Failure means the don't get them gem (ether failing to knock it out, or knocking out but not being able to find it and keep running). It doesn't mean there attempt to gain wealth while running out causes them to mysteriously drop their coin pouch.

…. Now... If players take the jump because I told them it was a short cut then they have two objectives, saving time and not dying. Perhaps they are willing to risk a jump because they are athletics but not a fight with pursuers because they vastly out numbered. The pit of death might be better odds of escape than other options and their pursuers might be duergar in heavy plate unlikely to attempt or make the jump. Then success helps them escape from death and failure takes them strait to it. The thing is players don't always do things to achieve reasonable goal. Sometimes a players does something silly and you blink twice and say "are you sure... are you really really sure.... you could die...." and they are totally, "Yep, someone hold my bear and watch this!"

… Critical Role example, Kealith a jumps off a cliff into the ocean AFTER the battle is over, turns into fish and attempts to avoid the jagged rocks just described at the bottom. Goal.... to look cool, if successful she gets style points and if she fails she dies a horrible death. The GM makes this clear... the player "oh well I am going for it, its in character" ….. This happens all the time at tables I have been at.

Some version of "oh well I am going for it, its in character" usually comes right before players attempt a goal that will result a test of which ether the success or the failure is not really relevant but a test is needed. Its usually a "hold my bear" moment and they are usually character defining moments the players will remember for years. I will let players call that role even if it lacks consequence or benefit but its amusing to let fate decide the outcome.

 

[Charlaquin]

It's an informed response (ie: how the characters have acted, what sort of behavior they are likely to be engaged in, what sort of conditions they're under, etc), and partly supported by the players at the table to shape it into the final result. It was a cooperative effort between the GM and the players, not just the GM telling the players how they acted.

I did try to simplify the description of events for the sake of posting here. I didn't expect to have to explain that.

Sure, but even as a collaborative effort, it still involves the DM overstepping what I would consider to be their role. If your group enjoys it, that’s awesome, but it would not be for me.

Meta-wise, I knew because of the situation and events that have happened in the game. It was specific to that particular situation. It's not a conclusion that I would always reach.

Ahh, ok, that makes sense.

And as I said, I didn't act on that directly because I wanted to remain IC.

Which is all well and good if that’s a decision you made for your own role playing satisfaction. But I don’t like making this kind of play an expected part of the social contract. If “the DM shouldn’t describe actions taken by PCs” is the first guiding principle of my personal DMing style, then “no one should police anyone else’s roleplaying decisions on the basis of ‘metagaming’” is the second.

You're conflating two separate aspects of the argument. It's "lacking" because all it is is, "Enemies attack; roll initiative".

If the narrative justification for the failed rolls isn’t what’s lacking, then what is? The rolls themselves? Again, I think the game is better off for the lack of rolls that don’t accomplish anything.

It's pure game, almost no story.

On the contrary, it focuses more on narrative resolution than mechanical resolution. If reading your method as “slapstick” is on me, reading mine as “all game, no story” is on you. For the record, a similar scene played out at my table wouldn’t just go straight to “everyone roll Initiative.” It would involve just as much of me describing the environment and the players describing their character’s actions. It would just involve fewer checks to resolve actions that are of (what I would consider to be) little consequence.

And as I noted, "Only in hindsight". I'm not saying it's lacking in "slapstick"; that's entirely on you. It's lacking in ways to explore and expound on characters as real living beings in serendipitous ways — chance events that allow characters to develop or express themselves outside of the rigid structure of pure Exploratory play, or fixed modules that have absolutely no concept of who the characters are.

There’s that “exploratory play” thing again. I don’t know what to tell you, but if you think that characters can’t develop or express themselves through chance events in “exploratory play,” then my game must not be whatever you think “exploratory play” is.

And even if it is slapstick, that doesn't make it bad. If that's what makes a game fun for a group, they should certainly be allowed opportunities to expand on that.

I don’t disagree. If you though my comment was meant as an admonition against slapstick, then you have misunderstood me, and I accept that it’s probably on me for not being clearer in my meaning. I like my games to be funny. I’m just not a fan of using comedy to handwave away failures that occur in circumstances where success should be certain.

 

[5ekyu]

Help me understand what you're saying here. @Ovinomancer's point I think is pretty clear: if the player says, "Do I know this creature's vulnerabilities?" and rolls a 1, and the GM says, "Yes, you recall that it's vulnerable to fire...." isn't the player pretty darned sure that the creature is NOT vulnerable to fire? But, given that this is metagame knowledge, is the player expected to try a fire spell and learn his mistake or is it ok to avoid fire spells, because you know that the roll failed, and thus you were given bad information? Because both of those results are pretty unsatisfying.

Is there an alternative result I'm missing?

"You character has heard a lot about creatures like this but the rumors have been mixed, often contradictory. Some have listed it as vulnerable to lightning, others fire and some have listed it as resistant to non-silvered weapons and another said it was difficult to affect with spells - some kind of magic resistance."

Now, which of those is true, which is false? Which resistances or other traits were not mentioned?

You the player and your character have knowledge that you have little confidence in, cited from the in-game sources.

Now which of these progress with setback nuggets of wisdom do you choose to try and risk?

 

[Charlaquin]

For clarification, does "meaningful consequence of failure" include or not include "gaining advantage on success"?

I get the impression this question is directed at Elfcrusher, but my two cents for anyone who might be interested is, yes, but only if the opportunity to gain that advantage is actually lost. Trying to hide is a good example of a case where the loss of an opportunity to gain an advantage is a meaningful consequence. If you don’t successfully hide right now, you’re not going to have time to set up your ambush before your target arrives (or whatever). An example of a case where the loss of advantage for success would not be a meaningful consequence in my mind would be something along the lines of researching in a library. On a succes, you’ll find relevant information, which certainly constitutes an advantage that failure would deny you. But if you fail, no big deal, keep researching for a few more hours and see if you find the information this time (I’m assuming the information in question is available in the library, otherwise we wouldn’t be rolling for a different reason). So, failure doesn’t actually deny you the advantage, it just makes the advantage take longer to gain, sp unless that time is a finite resource, that’s not a real consequence.

 

[Elfcrusher]

I don't see a problem with "A success moves towards the goal, a failure away." as general guidline however, I don't really feel the need to hold it as a hard rule. Sometimes the success is just not failing and failure is just not success, however in order for a role to have meaning at least one of them must be significant to the progress of a goal. For example, with jump over a deadly trap pit trap, if they succeed the jump and can clear the trap, they just continue on their way having avoided danger. If they fail, then they fall in the trap they die. They don't get set further from danger because that was make them safe. Success here is simply not failure, but the consequence of failure is sufficient enough that the role is meaningful.

To assess "toward" and "away" you have to compare success and failure to each other AND to not attempting the action that requires the roll. Presumably the reason they want to jump over the pit in order to get somewhere. If they successfully jump it then they have moved closer to their goal than if they stood there refusing to jump.

Now, it may also be that if they fall in they are still better off, because taking some damage then climbing out the other side may still be better than not getting over the pit at all. (It depends on why they are trying to get over the pit and how important that is.) In that case even though there's a consequence to failure, it's still a no-brainer to try, which makes it an uninteresting choice. Right? If it's the only exit from the dungeon, then eventually everybody is going to try to make the jump, so they aren't really making a decision of whether to take the risk. The decision was made for them, and their only job is to roll dice.

If it's absolutely necessary to get over the pit, there should be other ways to accomplish it.

Alternatively, a PC be short on time but see a valuable gem stuck in the wall and want to try and take a hammer and knock the gem out, grab it and keep running. Success means they swing knock them gem out and pick it up. Failure means the don't get them gem (ether failing to knock it out, or knocking out but not being able to find it and keep running). It doesn't mean there attempt to gain wealth while running out causes them to mysteriously drop their coin pouch.

Sure, if you set it up that way. But why? It's basically, "Make an ability check. If you succeed you get a gem, if you fail you don't." Why would you NOT make an ability check? I don't think this makes for an interesting challenge or obstacle. (The acid test is to ask what would happen if you asked "Does anybody else want to try?" If everybody says, "Hell yes!" then I would say you haven't set up an interesting challenge.)

 

[Charlaquin]

It's something like progress combined with a setback. The player wants to know fact A. The check fails. The DM gives the player fact B. It's not what the player wanted, but it's TRUE. The character now has the onus to try to make fact B useful, if he or she can.

But, of course, the response you're quoting was not the kind of narrated result that @Ovinomancer was talking about in the first place.

This, I like. Not enough to convince me to adopt your “describe a background element that might have led you to know this information” method of adjudicating lore recollection, but I will probably adopt this technique for actions taken to gain unknown information.

 

[Charlaquin]

I'll say this:

If your Apocalypse World games play like your Dungeons and Dragons game it is likely because you are ignoring a vast swathe of the game. The majority of what makes Apocalypse World Apocalypse World is GM facing. It does not have GM advice. It has GM instructions.

It tells you what your goals are. It tells you what your principles are. It tells you how to live up to those principles and goals.

When it tells you not to preplan a story, it reiterates with language not fit for this board. This is not unlike when other players are instructed to play their characters with integrity. Vincent Baker means it.

Yeah, but have you heard TAZ Amnesty? Griffin does not follow those instructions, but I’ll be damned if it isn’t an incredible game of Monster of the Week.

 

[Elfcrusher]

"You character has heard a lot about creatures like this but the rumors have been mixed, often contradictory. Some have listed it as vulnerable to lightning, others fire and some have listed it as resistant to non-silvered weapons and another said it was difficult to affect with spells - some kind of magic resistance."

Now, which of those is true, which is false? Which resistances or other traits were not mentioned?

You the player and your character have knowledge that you have little confidence in, cited from the in-game sources.

Now which of these progress with setback nuggets of wisdom do you choose to try and risk?

Huh.

Let's explore this.

First, let's assume one of those things is true, and the others false, but not the "opposite of false." I.e., maybe it's not vulnerable to lightning, but it's not resistant/immune, either. In that case even though they don't know for certain what its weaknesses are, the players (being used to your DMing style) at least have 3 possibilities to test. So they are strictly better off than if they hadn't rolled.

So instead let's say that one of those things is the "opposite of true", for example immune to lightning. Again, the players are used to your DMing style so they know one of these things is true, and one or more of the others is the opposite of true. It might tweak how they go about experimenting ("Don't use a high level slot on lightning...try a cantrip or level 1 spell first.") but they have clues to explore. Aren't they still better off than if they hadn't rolled?

In both cases, when the player proposes the action and informed there's a roll, there's zero reason to not roll. They're not taking a chance.

 

[ClaytonCross]

I get the impression this question is directed at Elfcrusher, but my two cents for anyone who might be interested is, yes, but only if the opportunity to gain that advantage is actually lost. Trying to hide is a good example of a case where the loss of an opportunity to gain an advantage is a meaningful consequence. If you don’t successfully hide right now, you’re not going to have time to set up your ambush before your target arrives (or whatever). An example of a case where the loss of advantage for success would not be a meaningful consequence in my mind would be something along the lines of researching in a library. On a succes, you’ll find relevant information, which certainly constitutes an advantage that failure would deny you. But if you fail, no big deal, keep researching for a few more hours and see if you find the information this time (I’m assuming the information in question is available in the library, otherwise we wouldn’t be rolling for a different reason). So, failure doesn’t actually deny you the advantage, it just makes the advantage take longer to gain, sp unless that time is a finite resource, that’s not a real consequence.

That's very good point. So for Stealth I also do count the lose of advantage, but its very true for any repeatable task the lose of the advantage of success is not sufficient a lose to be a "meaningful consequence of failure". That kind of goes back to "any many cases it is a judgment call by the GM, but a roll is generally called for when their consequence of failure and the result is a success moving towards the goal or a failure moving away from the goal." is a pretty good guild, but not perfect. … Failing stealth check has a clear consequence of being known. Even if you hide in battle, they search for you because they might not know exactly where you are, but they do know your around and they are not going to continue on their way keeping you from a goal of surprising them or holding you up and preventing you from just moving on. While success allows you allow the enemy to pass or the party to attack them hidden in a surprise attack.

 

[ClaytonCross]

In both cases, when the player proposes the action and informed there's a roll, there's zero reason to not roll. They're not taking a chance.

Why is it bad about players wanting to roll without having a reason not to want to roll? I mean a melee attack is kind of weapon skill test, but players are generally going to make those attacks without any thought of lose for attacking once the fight starts. What is gained by making players question as many rolls as possible? The idea listed above about having false information and sorting it out... seems like fun roleplaying to me. I feel like I am missing something here as to whey this type of "failing forward" would be bad. Can you explain, your reasoning here? I am really curious what your view of the benefit is.

 

[5ekyu]

Huh.

Let's explore this.

First, let's assume one of those things is true, and the others false, but not the "opposite of false." I.e., maybe it's not vulnerable to lightning, but it's not resistant/immune, either. In that case even though they don't know for certain what its weaknesses are, the players (being used to your DMing style) at least have 3 possibilities to test. So they are strictly better off than if they hadn't rolled.

So instead let's say that one of those things is the "opposite of true", for example immune to lightning. Again, the players are used to your DMing style so they know one of these things is true, and one or more of the others is the opposite of true. It might tweak how they go about experimenting ("Don't use a high level slot on lightning...try a cantrip or level 1 spell first.") but they have clues to explore. Aren't they still better off than if they hadn't rolled?

In both cases, when the player proposes the action and informed there's a roll, there's zero reason to not roll. They're not taking a chance.

In my experience in actual play the choice to experiment through say four options trying safer bets first is often a bit less effective than just relying on your strengths.

Leading with a cantrip is a good eample where if it wasnt opposite then going cantrip first slows your progress towards victory.

But, whether or not this rises to your own personal value of significance of consequence is to me meaningless.

The position i was addressing was the claim about the player knowing the roll as bad on a knowledge check created this conflict where the player knew the info was bad by the roll but needed to pretend it was not so.

Your own view over whether this was enough of a consequence to encourage the player not to roll is another subject and not one i tried to reply to in that post.

 

[Charlaquin]

Why is it bad about players wanting to roll without having a reason not to want to roll? I mean a melee attack is kind of weapon skill test, but players are generally going to make those attacks without any thought of lose for attacking once the fight starts. What is gained by making players question as many rolls as possible? The idea listed above about having false information and sorting it out... seems like fun roleplaying to me. I feel like I am missing something here as to whey this type of "failing forward" would be bad. Can you explain, your reasoning here? I am really curious what your view of the benefit is.

I believe that this is a guiding principle of Elfcrusher’s, in the same way that “never tell the players what their characters do” is one of mine. There’s nothing necessarily wrong with everyone in the group wanting a try at a check, but it’s something Elfcrusher finds undesirable at their table, and so they (he?) tailor the way they adjudicate actions to avoid it happening.

@Elfcrusher: If I’ve correctly identified your reason for not wanting to call for checks that don’t have outcomes worse than not attempting, have you considered adjusting the way you handle teamwork? Personally, I think the Group Check rules are kind of rubbish. What I do is, when there’s an action that the party is working on together as a group, I determine if the action would succeed if any individual succeeds (for example, when sweeping an area for a hidden enemy) or if it would fail if any individual fails (for example, when trying to travel stealthily as a group). In the former case, I ask that the character with the highest modifier make the check, and grant them advantage if anyone in the group would have advantage. In the latter case, I ask that the character with the lowest modifier make the check, and impose disadvantage if anyone in the group would have disadvantage. I find this covers most situations where everyone would want a go at something.

 

[Lanefan]

And this is also a common solution to an issue that the method you prefer. That, no metagaming, secret rolls, no retries even if the situation may suggest otherwise - all fixes for a problem created in the adjudication process. Why do any of them if you can just make an adjustment to how you adjudicate?

Because the 'fix' only serves to create bigger problems: loss of mystery from the player side, player knowledge put at variance with character knowledge (leading to either players having to self-police or to metagaming, neither of which is desireable), and loss of realism are but a few.

 

[Lanefan]

And, many find this outcome very problem causing. For one, the player likely knows they rolled poorly, so they know the information is bad but have to act otherwise. This puts a strain on authentic portrayal of the character. In effect, you've now tasked the player to play in a way that's best for the story rather than be the strongest advocate for their character possible.

Hear, hear!

Also, this method involve you, the DM, providing false information to the players. This can (and usually does) erode player trust in the DM.

It shouldn't, assuming the players are aware that not everything is going to be handed to them and sometimes that which is, is wrong.

Otherwise, it'd be mighty hard to play a spy-based game, or to insert a double-agent or doppelganger into the party with any hope of success.

Yes, you can do it, but the system is bad at it and you will get poor results. If I, as a player, can assert fiction in play by leveraging my best scores and playing to gain advantage, then I'll start directing play in ways the GM has little control over. Since I'm then creating my own problems and then their solutions, we're now in a degenerate situation for game play. This isn't good.

Again, absolute agreement from here.

However, I agree that GMs will often incorporate new fiction based on player action declarations (or out-loud thinking) because that sounds fun. The point I was making is that this kind of thing is based on the GM's approval, not any mechanical functions in 5e. The GM decides is the only means of new fiction, and the system is built to enable and work with this. The resolution tools in 5e are, after all, only engaged after the GM considers the situation and the action and determines there's uncertainty and a consequence for failure. Note that this only happens if the GM decides.

Even if you go with players asking for rolls, it's still the GM deciding what happens for any outcome, not the player. Again, GM decides in the controlling mechanical structure.

What I'm discussing here is player initiated fiction introduction in a direct manner. 5e is not built to support this.

I'm not sure any edition of D&D is built for this, at least by intent. Which kind of means, given that this thread is in the D&D-specific forum, that discussion of direct player initiated fiction isn't likely to get all thatfar.

 

[Lanefan]

I don't think false information works well outside of the context of secret rolls. Secret rolls and false information also require a significant amount of buy in that I do not think are a given for most Fifth Edition players. You absolutely have to do the work to show that you have no agenda for how things turn out before you can really utilize asymmetric information.

Hmmm...this makes it sound far more difficult/challenging than I've ever found it to be.

Varying mileage, I suppose...

 

[Lanefan]

To assess "toward" and "away" you have to compare success and failure to each other AND to not attempting the action that requires the roll. Presumably the reason they want to jump over the pit in order to get somewhere. If they successfully jump it then they have moved closer to their goal than if they stood there refusing to jump.

Now, it may also be that if they fall in they are still better off, because taking some damage then climbing out the other side may still be better than not getting over the pit at all. (It depends on why they are trying to get over the pit and how important that is.) In that case even though there's a consequence to failure, it's still a no-brainer to try, which makes it an uninteresting choice. Right? If it's the only exit from the dungeon, then eventually everybody is going to try to make the jump, so they aren't really making a decision of whether to take the risk. The decision was made for them, and their only job is to roll dice.

And no matter how they get across it, they've now put themselves in a position where they've cut off their own escape route...

If it's absolutely necessary to get over the pit, there should be other ways to accomplish it.

Why? If the goal of the pit is nothing more than some potential resource attrition (where resources can be anything from time to hit points to stealth) then making them work to get across it is the whole point.

And, as I just said above, the presence of the pit behind the PCs once they do get across it makes escape a no-longer-straightforward prospect.

Sure, if you set it up that way. But why? It's basically, "Make an ability check. If you succeed you get a gem, if you fail you don't." Why would you NOT make an ability check? I don't think this makes for an interesting challenge or obstacle. (The acid test is to ask what would happen if you asked "Does anybody else want to try?" If everybody says, "Hell yes!" then I would say you haven't set up an interesting challenge.)

I'd much rather they all say "Hell, yes!" than "Nah, that's boring."