This is all inspired by the "Challenge the player" thread from last year.

I've fallen into a pretty heavy "planned to showcase ability x" rut in my dungeon/encounter design. So I started thinking about challenges I've run before, which ones had worked in the past, and which ones hadn't. There's a definite pattern in the ones that I like, and I'm trying to figure out how to make more of those fun moments.

I think I can make things clearer by talking about a session of our saturday cartoon superhero game. The heroes had been transported to a magical video-game like realm where they had to do some fairytale like quests. Each of these things was pretty blatantly a challenge. To start, they had to prove worthy by beating the test of agility, the test of strength, and the test of intellegence. Also - out of genre powers were replaced with abilities that fit in more. Mostly movement abilities were removed.

The test of agility had the players facing a series of platforms above a pit. There was a ladder leading out by the beginning, but they'd have to deal with damage each time they fell. The dificulty was pretty high, enough so that the acrobat was the only one with a chance to make it through before falling. But the acrobat wasn't here for this session. How'd they finally make it, especially the low agility power-suit guy? It took a while, but eventually they started using a rope as a safety line. Most of the party braced themselves, and the super strong alien would throw the speedster at the next platform. A whole lot of jumping, throwing, shuffling, crazy laughter and trying again, and the party would be on the next platform, ready to do the whole thing over again.

The test of intellegence was a puzzle. There was a grid of numerical tiles with a pattern to them, the left and bottom rows needed to be filled in. It took about fifteen minutes, but those got filled in. The bottom left square? That took forever. There was some arguing, some trying random tiles. Things are getting tense, half an hour in, and arguements are about to break out. The acrobat's player finally arrived - he took one look at the starting puzzle, and handed me a piece of paper with the answer. Eventualy I gave the rest of the players a color coding clue, they figured the puzzle out, and we left that agony behind us.

The test of power had the heroes facing against a giant sized, hammer wielding, magically animated suit of armor so tough they couldn't hurt it. It also hit like 3 tons of bricks. This one was entertaining to watch. They started by swinging a few times, hoping to get lucky and do some damage. Following that, they disarmed him, tripped him, and held him down. The space alien picked up the armor's giant hammer and combo attacked with the acrobat. The acrobat was literally jumping off of the walls on to the hammer to add more force. The result? One crushed suit of armor.

In my evaluation, the test of intelligence failed. It wasn't much more than a variation on sudoku, but it was both too hard and too easy at the same time. It took up a lot of time, we weren't satisfied while doing it, and it wasn't fun afterwards. It also only engaged about half the party.

The test of agility ranks as some of the most fun we'd had gaming up to that point. It also took about half an hour, but was filled with description, laughter, teamwork and problem solving. They had to figure out how to use the abilities they had to overcome a situation where their chances of crossing the pit by just making the rolls was less than 10%.

The test of strength was a lot like the test of agility. Solid fun, took about twenty minutes total, and came with a lot of colorful description, a ton of movement and a good bit of laughter. They had to find out how they were going to even hurt this thing.

All three tests challenged the players. But the tests I liked challenged the players and their stat blocks. I think there's more to it than that though. Like, the powers they had were important, but so was the creative use of the abilites that they had and the tools at their disposal. Where I'm really stuck is figuring out what made the test of agility so much more fun than the test of power. Does anyone have any insight?

I think it'll be hard to point to anything definite, to explain why agility was more fun than power.  I'm going to take a guess and say that the agility puzzle really had everyone working together and improvising, whereas possibly only 1 or 2 people found the solution with the power test.  You may have had some of the players feeling like they couldn't really contribute, even though the party as a whole performed well.  I don't see any reason this would have to be true, since the suit of armor was its own kind of puzzle.  It might've worked even better than it did, but maybe some were feeling tired or a little discouraged by the intellectual puzzle to put in more effort?  It's strange that in rpg's there are so frequently problems with individualistic behavior, when most people I know would agree their best moments were when the party had to pull together to solve something, just like in your agility test.

I've had bad luck with intellectual puzzles also.  People just don't like them.  Maybe it's because it pulls them out of character -- they have to rely on player-brain resources instead of character-brain resources.

Possibly the more immediate and strong threat, while the animated armour - well, perhaps they are used to winning combats mostly? I dunno if it applies, but alot of roleplay defaults to 'guys only die in combat if they do something really, really, really stupid'. So the armour just didn't seem threatening?

For each challenge, what were your planned consequences of failure? Was it possible not to engage some challenges at all?

For the puzzle challenge, did you plan the color clue as part of the challenge, or did you modify conditions on the fly?

One of the things I really enjoy is providing physical props for my players.

In one game a friend ran, we had to solve a series of puzzles in order to get from one room to another. They were given a reverse-printed word that needed to be viewed in a mirror (provided), a pencil and a strip of paper that had random letters on it, and some other stuff.

It was a simple cryptography puzzle that had most of us stumped but we figured out the correct way to wrap the paper around the pencil to get the next clue (point the wand at the lion's mouth or something like that).

I'm not sure if this was a print module or a bunch of different things from around the world. But it was able to keep us entertained without involving combat and dice rolling.

I ran a RIFTS game once where players were trying to catch a cyborg squirrel (who was pretty quick) and using a "make more successes on a d20 than the squirrel" mechanic along with narrative approach to how the PCs tripped or slid had everyone laughing. No combat involved there either, just a simple roll fest and mini story.

I've found that if you can engage more of the players at once in semi-intellectual puzzles or risk-type games (such as farkle or yahtzee...push your luck kind of things) that combat isn't a necessity.

You do have to take the real-world mentality of your gaming group into account. Not that they're stupid or not, but the range of material, schooling, and just plain random knowledge.

You now know that the sudoku puzzle was a bit out of range for them so look for something a bit different that's more up their alley. Though, the more you can bring physical objects into the game, I think that's better.

To this day, whenever my old gaming group gets together, we laugh about the squirrel chase adventure...

Did the test of power come after the test of intelligence? If so, I agree with Jeff, they could have just been tired and gearing back up to enjoying themselves. The big warrior might also have been more familiar, or a more familiar type of solution than the platforming one.

Apart from that, I think your analysis of the difference between the intelligence test and the other two is accurate.

Hey guys!
I haven't forgotten this thread, but this week has suddenly turned all crazy on me!
I'll have a couple more examples soon.

Daniel.

And I'm back. Sorry for the delay.

Jeff - You might be on the money with that. In the Test of Power there was a lot of everyone doing their own thing, that happened to work out into a nice sequence of events. In the Test of Agility, the Alien and the Speedster had the lead roles, most everyone followed what they were doing.

Callan - I think you're on something on that. When we're in combat, there is definitely a tendency towards "taking their turn and hitting someone." With this particular group, failure wasn't really an option, only setbacks. Having to start over on the jumps probably had had more threat than failing the combat though.

Filip - I hadn't really thought of consequences for failure. It probably would have ended with running away, resting up and starting again. The players also had to get the item awarded at the end of each test. In fact, I pretty rarely think about the consequences of failure beyond the most obvious.  As far as the color code goes, I added that in on the fly.

Excalibur - Yeah, I definitely botched matching the intellect puzzle to my players. Two went "A puzzle? Nah," and checked out immediately. The two remaining ones began to argue about what the puzzle might mean. The remaining player (who sadly, missed most of the session) looked at it and immediately solved it. Heh. No fun for anyone really.

Josh - My players were probably gearing themselves back up, yes. Before last session I would have said combat was just less interesting (except in some wiggly games like D&D 4) but our last session of Mutants and Masterminds proved that wrong. We had a long running combat encounter that had the same flavor of awesome as the Test of Agility.

So, let's hit that example

The Challenge: Blow up the cloning facility

Our players are a set of elemental mages, currently trying to help a ghost haunting our Wind Mage. Functionally the ghost acts as his sidekick. Their investigations and help from an antagonistic NPC has lead them to a secret, underground lab where cloned bodies of the ghost are being grown. Their objective is to blow up the research notes and prototype cloning equipment. The Ice Mage is the player of the Alien Powerhouse from the superhero game.

The party bluffs their way past the guards with a "How dare you question me?" A quick infestation leads our party to the head scientist who happens to be our Ice Mage's grandfather. Some short debate later, and the new objective is to help Grandpa fake his death, then blow the place up. The Wind Mage and ghost find a body for the ghost to inhabit, then they proceed to explore the rest of the lab. They're noticed by the head guard who raises the alarm. Everyone at the table rolls their eyes and makes fun of the Wind Mage's player. Combat starts.

The Wind Mage ends up cornered by the guard who spotted him, and a hostile ghost. The Fire Mage runs to help him out, and the next few rounds are the Wind Mage attempting to defeat the ghost with his sidekick's help while keeping the Fire Mage and sidekick out of the ghost's area attack range.

The Fire Mage is mixing it up with the head guard and trying to keep him from double teaming the Wind Mage. After the Wind Mage beats the ghost, he uses a lot of helping actions on the Fire Mage's behalf till they beat the guard.

The Ice Mage seals the hall where most of the guards would come from, then spends the next few minutes in a game of cat and mouse with a huge brute that can break through any of her ice barriers in a single round. This was especially awesome because we both knew the brute could take her apart, and he proceeded to describe what he was going to do as he gave chase. When he finally managed to avoid her snare attack, there was a breathless moment of anticipation around the table, but I whiffed. 

The Earth Mage grabs the explosives and begins using his powers to punch holes in walls, allowing him to avoid direct contact with the brute, and get around more quickly. He plants all the explosives, though a few failed roles mean that some are going to go off early. There's some good natured arguing going on between him and the Ice Mage's player as they keep crossing paths.

The guards finally manage to chip through the barrier and charge into the hall, only to get taken down by the ice and earth mages. The first bomb goes off, and caves in part of the ceiling. Guards from above the compound come in, but are rebuffed by harsh words from the Ice Mage and head back out without getting the rest of the guards.

At this point, the party attempts to defeat the brute as a group, and weaken him enough that when the rest of the bombs get set off early, the collapsing roof puts a stop to him.

The party extracts themselves from the rubble. After a brief conversation Grandpa sneaks away.

This had some of the same sense of fun that the Test of Agility had. It's also the best "move pieces on a grid" combat I've ever run. We were neck deep in what was going on. We used some basic moves in new, logical, but surprisingly effective ways. I was also really impressed how a trio of bluff checks were able to change the face of the rest of the dungeon. The first got them in successfully, the second got them to the scientist, and the third stopped the last wave of guards. Of course, getting to the scientist without a fight just meant that everyone else was still around when the alarm went off.

I should have one more example up tomorrow.

Last example!
This is from a game of Burning Wheel from a while ago.

The Challenge: Question/Rescue the Mob Boss's little brother

For a variety of personal reasons, each of the players are investigating the death of the local mob boss. The players have fallen into two factions as their events have crossed. 3 of the PCs are at an uneasy truce, including a Student with lots of connections. The fourth PC is a big city Adminstrative Investigator here to check out what's really going on. The trail of clues has led to the Mob Boss's little brother, currently captured and beign interrogated by the head of the town watch. As a favor to the other 2 characters, the Student Who Knows Everyone is going to try and break into the building where the mob boss's little brother is being questioned and get him out.

Meanwhile, our Administrative Investigator has tracked things down to the same point. She needs to question the mob boss's little brother as well, but he's resisted hours of pretty painful investigation so far. She manages to bully her way into the building to see someone else, but has no way to actually get the information she needs out of the brother.

A little later the Student has managed to circles and bluff his way in. Trying to walk around and check out the situation, he and sees the Investgator sipping tea and waiting for an opportunity. They're both quite familiar with the other by now. There's a hesitant moment where each wonder if the other will blow their cover. When that doesn't happen, they briefly discuss what they're trying to do, and make an arrangement. If the Investigator helps the Student get the brother out, The Student will help the Investigator get the answers she needs. The investigator intimidates her way into the room, and they manage to get some info from the little brother, but he's pretty badly beaten up and needs medical attention.

The two make a plan. The Student will raise a distraction out front, and the Investigator can sneak little brother out the back. The student manages to re-form an angry mob from earlier, and direct them at the house. The Investigator drags little brother out of the house in the confusion.

So, things that have been tickling in the back of my brain while writing that last post. I totally have expanded what I mean in challenges. Sorry about that. Those were just the situations I've come across that felt similar to the Test of Agility. There's that same flavor in all of the encounters. Let's look at our ingredients so far. Based off of people's comments I have:

1) Require some thinking
2) Need to use the character tools/resources available
3) Description
4) Consiquence
5) Unfamiliarity/Unpredictability

So, let's zoom on two exchanges on the test of agility.

The speedster is on the third platform with one end of the rope. The powerhouse is holding the rope on the second platform. The wizard and powersuit are also on the second platform. The powerhouse jumps and just barely makes it across to the third platform, but drops her end of the rope. The powerhouse takes the speedster's end of the rope, reels it in, then tosses the other end back to the second platform.

The powersuit ties it around his waist as a precaution before jumping. The powersuit jumps, and fails pretty bad, slamming into the pole holding up the platform. He fails his toughness save, ringing his helmet aginst the pole like a bell. He goes limp, and due to a failed roll on the powerhouse's part, jerks the rope out of her grip. The speedster dives for the rope, making his attack roll, but barely failing his strength check. So now he's got the rope, but is falling over the edge too. The powerhouse makes a grab for the speedster and makes it, then braces herself and pulls him back up. The two slowly reel up the powersuit. The powerhouse ties the rope around her waist so she can't lose it. Then they throw the rope to the wizard.

Looking at that, I'd add a 6th ingredient
6) Plausability.

There was no rule that "a rope will give you an XYZ check if you tie it around your waist and fail making a jump". It made sense that they could climb it, and there were rules for that. But every thing they did was perfectly plausable. Tying the rope around their waists? Perfectly plausable. Later portions, where the wizard tries rapelling up the pole? Perfectly plausable, but not explicitly options.

I think this has a pretty good sense of balance to Unpredictibility. Like "I should be able to gain an advantage by doing this reasonable thing with these items at hand. However, something unexpected, but within reasonable limits could still happen."

What's plausable, of course, varies from group to group.

Daniel,

Filip - I hadn't really thought of consequences for failure. It probably would have ended with running away, resting up and starting again. The players also had to get the item awarded at the end of each test. In fact, I pretty rarely think about the consequences of failure beyond the most obvious.  As far as the color code goes, I added that in on the fly.

Filip,

Your insight highlights everything that I'm trying to improve in my play. Going for the sense of challenge instead of an actual challenge sounds like how I settled on running games about 4 years ago. And boy, I can go on tangents about the what and whys and hows. Particularly, my recent 4e game ended over "challenge" vs "illusion of challenge" issues. Two players are not in my current game due to being diehard on believing that the illusion of challenge is the way to go.   

I'm happy to say that I am improving, and am committed to improving further. I should have included a time-line in my notes to make this more evident. The tests of Intelligence, Agility and Power are from 4 years ago, and I've been moving slowly towards real challenges since.

Time-line wise, the challenges go in this order:
1) Test of Speed/Agility/Intelligence (3-4 years ago)
2) Rescue the Mob Boss's little brother(2 years ago)
3) Blow up the Cloning Facility (last month)

And let me tell you, the players' plans were all sorts of derailed if they messed up on the cloning facility. At best, the villains would have gotten to continue refining their clones of one of the most powerful wizards alive. The players may have been captured. The worst possibility would have been getting run out after being identified. Grandpa and the Ice Mage's family would have been under a lot of scrutiny, and possibly had attempts made on their lives. How this ended was going to have the most long term impact of anything in the game, and everyone knew it.

So, I know that I can step up to that level of challenge, but I need to learn how to make it consistent.

As to why the test of intelligence was less fun, I can walk through that list. Have I mentioned that I love this list?
1) Require some thinking. It very much required some thinking.
2) Use the tools/resources available to the characters. High int? Nope, doesn't help. This really kicked us out of the game's world. Had we thought about it, the speedster could have just tried all 1000 possibilities for the last tile.
3) Add description as a matter of play. Nope, there wasn't any description or addition to the game's world by working on it. We had the starting diagram on the battlemat, and the players were working on a piece of paper. There was no movement or change involved in putting a number down. The challenge was pretty much a block until completed.
4) Have a game-world consequence for any action. There wasn't much consequence. Aside from the time invested in trying an answer, nothing was lost in a bad guess. Even having it whittle away some health periodically would have been an improvement.
5) Have outcomes be unfamiliar or unpredictable. Nope, it was only unpredictable because that the players didn't know what was in my head. Putting a tile down was entirely predictable.
6) Game world plausible actions as a viable strategy. Plausibility wasn't part of the puzzle. There was no way to 'trick' the puzzle. The tiles couldn't be stacked, or illusioned or punched or shot. The puzzle generally didn't follow any of the rules of the setting except "it's magic and it knows when you're done". I had my answer and they were just going to have to give it to me.

So, of the list of 6 things, I failed on 5. Yikes!

Just a quick question on #6, which may be off base - why does it have to be plausible?

I'll grant I have an inclination towards it, but I'd describe why I indulge that inclination rather than just pursue it. Basically like lion cubs wrestle each other as practice for real prey latter on, I think there's an instinct in people to 'wrestle' with an imagined situation to better prep for real life. The desire for plausibility is to make it somewhat (it's never fully) applicable to real life.

However, I think in pursuing plausibility - well, I think where simulation sometimes comes from is an obsessive pursuit of plausibility, not as a means to something else, but obsessive pursuit of plausibility for it's own sake and the original purpose of the instinct subverted entirely. It's something to watch out for.

Callan,

Why Plausibility? This is kind of a tough one to answer.

First, I think I need to clarify what I mean and what I don't mean. First, I don't mean plausible as in "like the real world". Nor do I mean plausible as in "Fully specified."

What I'm looking at is some sense of having enough of an idea about how the game world works to make an informed judgment and create a strategy with a reasonable expectation of success on its own merits. Let me break that sentence down into parts and see if I can give you a better handle on the idea I'm trying to express.

First is having an idea about how the game world works. I think that's fairly straightforward, SIS type stuff.

The players are able to make an informed judgment and create a strategy. The players at the table should be able to come up with a plan that is strongly based on the mental picture of how the game world works. An example of this might be getting around a trap by marking the cobblestones that set it off and avoiding it. Using the rope in the test of agility is another example.

The strategy's success should be based on its merits. Basically, the odds of this plan working should be based on knowledge of how the game world works. The rope idea didn't work because it was cool, it worked because that's about how I'd expect a rope to work.

We could probably merge Plausibility Description and Consequence into something like "an active, changing game world", but for me the separation is very helpful.

Now, to actually get around to answering your question. It needs to be plausible in order to have enough internal consistency for actions that aren't explicitly enumerated to apply. But how about this statement "The world needs to make enough sense so players can come up with novel ideas."

So, plausibility isn't a goal, it's a step to get "fun" challenges (for this specific value of fun). So we only need as much plausibility as the situation requires. This pretty much neatly sidesteps your issue. I don't care how plausible it is, or what the motivation for it is, as long as there's enough internal consistency to make a reasonable guess as to how a plan may turn out. Otherwise it's salting to taste.

Also, if you have a better idea of a word I should use, please let me know.

Mmmm, once again I'm bewildered by gamers talking about vocally described fiction as if it exists, and not talking in terms of what actually exists at the gaming table. Marshall did this to me the previous time. Because this 'game world' doesn't exist to be plausible. It's like talking about having a giant invisible bunny walking around with you, but instead of talking about how he's an invention of your or someone elses mind and how each element you might describe of him is someones invention, you just talk about what qualities the bunny 'has', quite detached from any human source.

The 'game world' you refer to doesn't exist to 'have' any quality of plausibility. Someone could invent something and describe it and their invention might sound plausible, but even then that's just how it sounds as well as it being something they made up rather than something that exists.

I dunno, what your refering to just bewilders me. So that's where I'm left. Just noting it. Not saying it as if it requires attention or a responce.

Of course, no gamew world really exists.  But you can treat it as if it exists.  In addition, anything that has been established to exist can be defended by appeal to the social contract.  If a thing has been established in the SIS< and that is known by the group, then contradiction of that thing can be prevented; you will, in short, get called out for doing so.  This can be extended further to things that are only probabilities or potentials on the grounds of "reasonable expectation" or extrapolation from our own, objectively existing world.  The IS exists in the same way that rules do - by our awareness and consent.

Hi Gareth,

looking at this with some scientific rigour, beyond just strong assertion from any party, can you provide evidence for your assertion that you are aware? I'm not looking to open that up as a subject in this thread - I am saying it because I am not acknowledging your point at all at this moment, but neither am I just ignoring it because of my own strong assertion. I really am willing to look into some evidence. But otherwise no, I think your unaware.

Hey Callan,

Are you doing this to try and get a reaction out of me? Are you trying to understand what I'm saying?

I ask, because it seems disingenuous to say that I'm not talking about what goes on at the gaming table. Take a look at my prior examples, like the Test of Agility and the Mob Boss's little brother. If that' not what you're looking for, then I'm afraid I'm going to have to ask you to put up an AP example that does illustrate what you're talking about.

If the existence of the SIS  this is something you'd like to talk about, I'd be glad to discuss it with you. But I need you to do something first. I need you to acknowledge a few points.

a) A SIS can exist
b) People who use and acknowledge a SIS are not crazy, stupid or delusional
c) My viewpoint on this is potentially helpful to you.

If not, there's really no point to trying to discuss SIS with you, no matter what angle I come at it at. 

But, In terms of a brief reply, a thing doesn't have to be real for us to build a model of how it works. To use your example, If I talk to my little sister about how my giant imaginary bunny is blue, and then later about how it's red, she'll call me on it and say that it was blue because that's what her brain has built into it's predictive model.

Everyone else, hold of on bugging Callan for a bit?

Daniel,

At the forge we aren't like alot of other forums - on other forums if someone says something to the contrary, you totally have to engage it or lose some sort of intellectual turf. You don't have to here. My post was supposed to be a wrap up post, kind of like a post it note tacked on, which I honestly estimate most people will ignore anyway. But also at the forge it's not considered rude to politely disagree.

And everything else I'm biting my lip on. Maybe someone else would know how to respond in a genuine engagement. I don't. Maybe I'm wrong but it seems like it's going sour to me. Pretty much everyone else will take it, to use your own words, that the SIS exists. So don't worry about one voice disagrees but who's said he's not going to post further anyway.

Hey Callan,

Thanks for the feedback!
I'm honestly trying to engage an interesting question you brought up. "Why do people believe in and use a SIS?" Because I think engaging you in the topic would be an interesting direction to take this thread, particularly since the SIS is important enough to take up 3,4,6 and a bit of 1 and 5. It's central to what I'm trying to accomplish.

However, since doing this would require the amount of effort that goes into a (very small) research paper, I wanted to make sure that it was going to work out. Hence my rather draconian list of demands. Since you've made it clear you weren't attempting to engage me in conversation about the SIS, I'll let it drop. Thanks for the participation in the thread, I found it pretty helpful.

Hiya,

Through no one's intention, this thread got effectively hijacked. Let's dial it back to the middle of the first page, not in the sense of ignoring what's been said since, but in the sense of what the thread's directly addressing. Daniel, if you want, feel free to write a one-sentence mission statement for it, or if you think you've done so already, quote yourself here.

Best, Ron

Ron, I guess my goals for the thread have drifted a little as well.
Mission Statement: Help me understand encounters that get levels of engagement like The Challenge of Agility.
I feel like I'm still missing a couple of social factors to the whole thing.

I guess I have two sub goals as well. I want to:

• Understand what I need to do to craft 'fun' encounters
• Understand what/how game systems can make this easier or harder

Each of those might have to be its own thing though.

Something occurs to me about patience vs "challenge"; there are many real life challenges where you can try to get something one way and fail, say a ball on a roof. In fact I remember seeing a situation with two kids trying various things to get it off. They kept failing at different methods, but I could see that they would very likely eventually get the ball, even if it meant having to wait and get their parents to phone the park-keeper.

In this situation what made it fun or not fun was the fact that failure at a specific task wasn't the end; they were getting a bit despondent, until me and some of the other people walking through started providing them with multiple alternatives that they could try out. As the idea of victory (as opposed to dead stop) began to become a conceivable possibility, they started cheering up, and trying bigger sticks, looking for ones with built in hooks etc.

So inevitable victory and total failure can be fine, so long as they can keep trying alternatives until their patience runs out.

An important part of what made this real life situation have that property is that they could keep expanding the scope, bringing in other elements and working out how to deal with the complications that might introduce (angry dad?). Also the more the scope expanded, the more likely complications were to be long lasting (although not always).

Hang on, I thought I'd posted a response to your other examples before, must've lost it.

So digging that up from memory:

I notice your definition of an enjoyable challenge is quite self-less: You enjoy your players enjoying it, which is nice. It also suggests to me that to make an all categories good challenge for you will require a toolbox long term, as you come across players with different styles. At the same time, I imagine that there is a portion of the task that can never be simply reduced to "are the players having a good time?", instead there will be things about it that will be uniquely interesting to observe or do when setting up the challenge, and if we can find that, awesome, because you'll be able to put that in the intro to games you run as a sort of introduction/specification.

To get there we will need more examples of challenges that were crap, unfortunately, although I'd much rather talk about good ones. Have you had some challenges that you didn't like but your players did?

On the successful challenges:

You cover a lot of what I notice in your current specification, but I also notice that one feature all the good things share is overlayed objectives/purposes, criss-crossing over each other in the same space. Sometimes these are sub-strategies with players in large scale agreement, sometimes these are but an important thing is that the teamwork isn't too rigid; there's no real command hierarchy, and there is also no conflict between players, instead you have overlapping and partially assisting attempts by the various players.

One thing that makes this happen is that mucking about is resolutely allowed; there's no time limits requiring people to be businesslike, and there aren't any scarce resources they must judiciously guard (or are there? I get the impression you were mostly using "at-will" type stuff).

Secondly, and related to the first, organised cooperation is not required for survival, only success. That stops people reacting with fear when other players muck about and the overlapping stuff becomes fun instead. That's not to say there isn't danger, but people can focus on mitigating danger and do so, even if it makes them a bit useless for a bit.

Then finally when players do cooperate success is not a matter of "assist rolls" but the sort of "reasonable effect" power use you've referred to before.