Pondering Perception

[delericho]

So, I've been thinking about Monte's "Legends & Lore" article about perception, and the more I think about it, the more I think he's trying to adjust details of the mechanics (that mostly work fine) in order to fix what is actually an adventure design problem.

Whether it's done using a check against a Search skill, or using Passive Perception, or using his proposed ranks system, or whatever else, the rules for the system will basically work just fine. And, frankly, it won't make a great deal of difference which one you use - they all use more or less the same process to get more or less the same results.

But the problem is that using a mechanical solution to find hidden features in the dungeon is boring. The PCs walk into the room, they search the room (using whatever process), they get the information, they move on.

And it gets worse, because in most cases the only way for the PCs to find this information is by searching, and very often the adventure gives no indication as to where to search. And so, the PCs just have to search everywhere. And so every door becomes a matter of "I check for traps. I remove the traps. Is the door locked? I open the lock..." Every room gets searched according to an ever-increasing set of Standard Operating Procedures, coupled with the appropriate rolls, made without any particular care so that the party can (a) get all the loot (or avoid the traps), and (b) move on as quickly as possible.

Instead, I'm inclined to think that elements in the game like traps and secret doors (but, actually, also including locked doors and even outright puzzles) should be in the game mostly as a challenge to players and not to their characters.

So, I would suggest the following:

Mechanically:

- Searches should be limited to "we search the room". This should take several minutes (they can combine it with a Short Rest), and allow a single roll made by a single character. And no, they can't take 10 or 20 on the roll. (I'm torn on whether to allow Aid Another. Basically, either allow it and raise the DC, or disallow it and just assume that everyone is pitching in. The letter is probably preferable as it's quicker, and makes the probabilities more controlled.)

- The DC for hidden elements should generally allow a 70%-ish chance of success. So, if the party rogue has a Perception mod of +8, the DC on checks should be approx 15. (Obviously, this should really be tailored according to PC level rather than an absolute value - see "page 42" or its equivalent.) This means that a skilled character will find most but not all hidden elements on a simple search... but that's fine.

(The same should almost certainly also be true of attempts to open locks and disable traps, as well.)

- Passive Perception should never find a hidden element. It's still of use, but it represents a defence against others sneaking up on you, and not a Spidey-sense for traps.

(If Passive Perception can find hidden elements, then the DC for searches needs to be much higher, or else having hidden elements at all becomes pointless. Problem is, those hidden elements then go from "always found" to "never found", and you're back to constant search checks anyway!)

But in adventure design:

- Include lots of hidden elements. This allows the trained Rogue his chance to shine (as he finds lots of secret doors and traps), while at the same time still leaving some challenge (because he doesn't find everything). (As a rule of thumb, an adventure the size of "Keep on the Shadowfell" should probably include 10 such hidden elements.)

- For every hidden element, include at least three clues to its existence hidden around the dungeon (and not all in the location of the hidden element). This could be a treasure map with the secret door marked, or a scrawl on the wall left by the hobgoblins' last prisoner, or the skull of the last adventure to die in the trap, or scratches on the floor, or whatever. But be sure to leave the clues in places where the characters can find them.

In general, these clues should be found automatically when the party searches the appropriate room. Nesting hidden elements just makes it that much less likely that they'll be found.

- When the players try the specific thing to trigger the hidden element, they should find it automatically. So, if the party are deep inside Peter the Mad Dentist's dungeon and they find a statue with scratches at its base, they might reasonably conclude "we try the statue's teeth"... whereupon it will slide aside to reveal a hidden compartment.

And returning to the question of locked doors and traps mentioned above, the same should apply - in addition to the Rogue simply applying his skill to these things, there should be alternate ways past. Once the PCs have learned "only the penitent man shall pass", they may be able to bypass the trap without a roll. Likewise, if there's a locked door, there should probably be a key somewhere in the dungeon.

- Don't include more than one hidden element in a single place. Don't bother with traps on locked doors, or having two secret doors in a single room, or... Again, either the group will find both with a single roll (which seems a waste of a hidden element), or they'll find one and assume that there's nothing more to see. Remember, you want them to at least have a decent chance of finding everything, at least given a modicum of good play.

- It should go without saying, but... never build a dungeon that requires the PCs to find a specific hidden area to complete. Otherwise, you're going to get some very frustrated players when they blow all their search checks and can't find (or decipher) the clues.

As for the issue of the "Standard Operating Procedures", these are a playstyle issue, much like the "15 minute Adventuring Day". In fact, they're probably a response to DMs who have hidden elements in their game, require the players to figure out exactly where they are, and who have provided insufficient clues to enable them to do that. Therefore, the group is reduced to exhaustively going through the dungeon interacting with every single element in every possible way to find the one hidden item in the dungeon. This is combated by making sure there are means of finding the hidden elements, by not simply springing death traps on them without any warning, and then by encouraging them to move on reasonably briskly.

(It is probably worth noting that almost all of this applies to all sorts of hidden elements in the game. The application to secret doors and traps is obvious, but it can also be applied to other things. Perhaps the bandit leader's lieutenant is willing to betray his boss if his crimes are forgiven - the PCs can then negotiate him with Diplomacy (as per the 70% rule-of-thumb above), but alternately if they've put together the clues they can go directly to the solution for an auto-success.)

 

[Aberzanzorax]

I like what you're saying here.


I've always been bugged by traps. The whole "search roll, disarm roll, done" paradigm is boring. Heck, sometimes my players WANT the trap to go off so they can see what it does, and have a bit of excitement.

I'd much rather something more dynamic, even something everyone in the party can participate in, but that the thief/rogue would be naturally suited for.


For example: crushing celing trap.

The celing is going to come down, either quickly or slowly. How well the rogue rolls on disarm will modify the speed. Then all players move through the area, perhaps fighting skeletons along the way, and roll acrobatics each round to avoid the spikes protruding from the ceiling. This is also modified by the rogue's disarm roll (e.g. a good disarm means a slow trap, so the pcs get as much as +5 to acrobatics rolls).


I think, if a trap is "disarmed" that means the players still need to contend with it. It just makes it easier. Then the trap is a true hazard, and not a binary on/off switch of boring.



Same for perception....there needs to be degrees of perception. (E.G. on a poor roll in the woods - you hear a branch cracking. On a very good roll, you actually see the bandits.)


EDIT: Forgot to address your "auto win".

I love that idea. Sometimes it's a puzzle (the dentist teeth), sometimes it's a hazard (in my example) and sometimes it's both. (Like if they didn't solve the teeth statue, the crushing room would exist...and if they did find it, the crushing trap never goes off.)


Adventures as written currently enourage exactly what you say (roll dice at every door) and discourage active thinking (because you STILL need to roll at the statue). That needs to change.

 

[(Psi)SeveredHead]

I tried this setup last weekend.

The PCs were given the task of destroying an inter-dimensional portal that was allowing demons into Athas (City by the Silt Sea adventure, but much reduced to their level).

I made the DCs to notice the turret traps and panels moderate (which meant they were auto-spotted by passive Perception) because I wanted them to see these things. Heck, the panel was practically being worked on when the PCs walked into the chamber. What they didn't notice were the trigger plates (hard DC, impossible with their passive Perceptions). Since finding the panel was easy, one PC spent three rounds (and six actions) actually taking control of the turrets to use against the enemies.

There was an even nastier trap, the portal itself, which kept electrocuting PCs and throwing them down the stairs (it was a warder trap, based on the Blood Tree in the Dark Sun Creature Catalog, if anyone wants to know more detail).

The PCs got scared because they noticed only one opponent, and quickly discovered he wasn't an elite. Also, turret traps are nasty because one trap gives two turrets, and it's not elite.

Unlike previous editions, 4e specifically gave traps roles. "Lurker" traps have been used poorly in the past, especially a pit trap in the middle of nowhere (well, a place where the PCs would go). A trap should be part of an encounter, where everyone can do something. Even a pit trap has a purpose now, you just put it in the part of the room where the PCs are likely to get good cover from ranged (put the mouse trap where the mice will go) and then watch as the ranger, wizard or what you falls down in the midst of raging combat. All of a sudden, the "cost" of the trap isn't the loss of a little time, the loss of a few hit points, the loss of a few spells, or the use of a few healing surges. You have a PC stuck down there, not dishing out damage or effects, maybe they don't have the Athletics skill and need rescuing, maybe it's the fighter who fell down there and so there's no one anchoring the front line... it's even worse if the kobolds rush forward with lamp oil and torches and throw them down the pit, or there's a swarm of scorpions down there or something.

I usually give non-lurker traps an easy to find panel (moderate DC), often easy to find traps (moderate DC, it's moot, the trap springs when you enter the kill zone anyway, and you can't search for traps in the midst of combat) and hard to spot trigger squares. WotC, even now, makes the control panel hard to find and trigger squares not-so-hard, which is missing the point. If you're reading an adventure with wonky DCs, switch 'em up DMs!

I want to PCs to interact with the traps, even if that just means smashing the turrets or mechanisms - fighters need to have a role in trap-breaking too. PCs might walk into a room with obvious scratches caused by a clockwork slicing-from-the-ceiling trap, with the panel clearly visible on the other side... too bad there's kobolds armed with crossbows there, and of course none of them will enter the trapped area.

 

[delericho]

I tried this setup last weekend.

Cool example!

Unlike previous editions, 4e specifically gave traps roles.

That's a very good idea. Ideally, there should be a "Book of Traps" giving lots of different traps with different roles and different levels, plus a toolkit for quickly and easily building custom traps, and lots of guidance in how to use those traps.

Whether such a 'book' should be stand-alone, part of the DMG, or part of the MM is an open question.

Also, a "Trap Builder" for the DDI would be nice to have.

"Lurker" traps have been used poorly in the past, especially a pit trap in the middle of nowhere (well, a place where the PCs would go).

Agreed.

A trap should be part of an encounter, where everyone can do something.

Mostly agreed. I think there is a place for the occasional 'gotcha!' trap. Also, there's a place for what 3e called "Encounter Traps" - a trap that isn't part of an encounter but rather one that is itself an encounter.

Even a pit trap has a purpose now, you just put it in the part of the room where the PCs are likely to get good cover from ranged (put the mouse trap where the mice will go) and then watch as the ranger, wizard or what you falls down in the midst of raging combat. All of a sudden, the "cost" of the trap isn't the loss of a little time, the loss of a few hit points, the loss of a few spells, or the use of a few healing surges. You have a PC stuck down there, not dishing out damage or effects, maybe they don't have the Athletics skill and need rescuing, maybe it's the fighter who fell down there and so there's no one anchoring the front line... it's even worse if the kobolds rush forward with lamp oil and torches and throw them down the pit, or there's a swarm of scorpions down there or something.

Yep. Good example of use of a classic trap that has traditionally been misused.

I want to PCs to interact with the traps, even if that just means smashing the turrets or mechanisms - fighters need to have a role in trap-breaking too. PCs might walk into a room with obvious scratches caused by a clockwork slicing-from-the-ceiling trap, with the panel clearly visible on the other side... too bad there's kobolds armed with crossbows there, and of course none of them will enter the trapped area.

Agreed on all counts.

 

[I'm A Banana]

But the problem is that using a mechanical solution to find hidden features in the dungeon is boring. The PCs walk into the room, they search the room (using whatever process), they get the information, they move on.

I think boring is kind of the point. Auto-success on skill checks is never going to be exciting, because there's no real tension. It showcases character skill.

I like the distinction between "defensive passive" and "aggressive active" on display here. Stumble into a trap, Passive Perception might help you avoid it, want to find that magic sword, a roll is the only way you'll do it.

I'm also a fan of the "specific solution trumps a roll" thing, but I'm cautious about it -- if the idiot barbarian solves the riddle that the genius wizard could not just because their players are clever in the first case and burnt out from a long day at work int he second case, that's not exactly satisfying gameplay for me. There's gotta be a mechanic that I can use to say, "Okay, how does your character do it?", to protect a character's role in the adventure. I maybe making this problem bigger than it might otherwise be, but it's a concern that any system like that needs to at least be able to address, that distinction between "I know the solution!" and "My character knows the solution!"

I think you're right in that Peception isn't a skill problem, it's a design problem, but there's gotta be a more satisfying way than yelling out guesses and making rolls...hmm...

 

[I'm A Banana]

Doubleposthaste

 

[delericho]

I think boring is kind of the point. Auto-success on skill checks is never going to be exciting, because there's no real tension. It showcases character skill.

Indeed. I have no problem with showcasing character skill. However, I think to a large extent these hidden elements in the adventure should be about player skill, and that boiling these down to nothing more than a dice roll is a problem - and that seems to be a mistake that post-3e adventures (at least) have fallen into.

I'm also a fan of the "specific solution trumps a roll" thing, but I'm cautious about it -- if the idiot barbarian solves the riddle that the genius wizard could not just because... There's gotta be a mechanic that I can use to say, "Okay, how does your character do it?", to protect a character's role in the adventure.

Honestly, I think the answer to this particular one is "my character doesn't solve the riddle; the wizard does - I'm just giving Bob here the answer." Or whatever other handwave is required to make it work.

(Of course, it's also perhaps worth noting that the dice sometimes give odd results, too - sometimes the weakling wizard manages to lift something the buff barbarian can't just because of how the dice fall. So some turnaround sounds like it might be fair play anyway!)

Ultimately, D&D is a team game, and if the goal here is "challenge the players", I'm not sure it matters which player comes up with the answer - the important thing is that the team gets the satisfaction of beating the challenge.

I think you're right in that Peception isn't a skill problem, it's a design problem, but there's gotta be a more satisfying way than yelling out guesses and making rolls...hmm...

Indeed. That's why I'm trying to put forward a compromise solution:

- The mechanics allow the party to do a general search, which will often (but not always) get results. But where the dice don't go their way, that's it - the mechanics won't allow them retries, or more detailed searches, or anything else. After that one roll, it's down to player skill.

- At the same time, if the players have been canny, they have (or will) pick up on the clues that allow them to hit on the specific solution to the problem, in which case they are able to solve the problem directly (via application of player skill). But in this instance, they need the specific solution, and not just "I check the statue. I check the table. I check..."

The problem of players who manifestly don't have a clue but are just calling out suggestions to fish for a solution is something that the DM really needs to solve, because the mechanics can't. Either the mechanics will let them keep rolling until they find the thing (in which case you might as well just give it to them), or they'll hit on the exact solution and the DM will have to say "no" because they've blown their roll. Neither of these is going to be a satisfactory solution. Basically, the DM has to just have the mandate to say, "moving on..."

 

[delericho]

One other thought springs to mind.

There's gotta be a mechanic that I can use to say, "Okay, how does your character do it?", to protect a character's role in the adventure.

Adventure Backgrounds tend not to be terribly well used, and could well serve to fix this problem.

Most adventure modules come with a fairly extensive write-up of the adventure background... much of which the players never get to learn, and most DMs never use. It occurs to me that this could be handled better.

Basically, instead of giving the DM a wall of text, the adventure should provide a set of "what you know" lists for various skills. Players should then get the handouts for their character's Trained skills. (There might be handouts for some or all of Arcana, Dungeoneering, History, Nature, Religion, Streetwise, Bardic Knowledge, Artificer Knowledge...)

These would then serve as a major source of the clues for various hidden things in the adventure. And, since the genius Wizard will generally be trained in a lot more of these mental skills than the idiot Barbarian, that player will generally receive a lot more of this information, playing up his character's role as a font of useful knowledge.

(Arguably, this is better done with a roll than with a binary Trained/Untrained check. However, since it's easier to provide fewer handouts...)

 

[KidSnide]

I think the OP is onto an excellent point. If 5e is to really improve the quality of the non-combat experience, the designers will need to focus on creating well design adventures and better guidelines for writing the same. It's not just a mechanical question. It's about helping GMs write more balanced, more interesting and more immersive adventures.

This post is about exploration. How do you write good exploration gameplay? You need to create interesting things to find, and you need to provide useful clues to help the PCs find those things. Looking everywhere needs to be a helpful strategy, both to find clues and to have a shot at just finding the interesting elements on their own. And, if finding things is to be a challenge, "success" needs to be possible even if you don't find them all.

That isn't to say that mechanics aren't important. They are. Mechanics need to support good adventure design in the sense that they need to interact with the good design in a way that is fun and allows the PCs to interact with the good design in ways that (1) highlight the relevant proficiencies of their characters and (2) do so without bypassing the fun part of the good design.

But you can't build these mechanics in a vacuum. You need to understand (and promote!) good design before the mechanics are actually helpful.

-KS

 

[mmadsen]

But the problem is that using a mechanical solution to find hidden features in the dungeon is boring.

Absolutely. If the players don't get to make interesting decisions, then the process isn't interesting. The choices then are (1) streamline the process, or (2) make the process revolve around interesting choices.

I suppose "oh, the elf spots a secret door!" streamlines the process enough that the game can move on to the good part, but I think most of us would prefer making the search interesting.

Every room gets searched according to an ever-increasing set of Standard Operating Procedures, coupled with the appropriate rolls, made without any particular care so that the party can (a) get all the loot (or avoid the traps), and (b) move on as quickly as possible.

A dose of reality may help here. Real live humans can't check everything. They can go through the motions, but if they aren't rewarded rather regularly by finding something, they lose interest and stop spotting whatever they were looking for, even when it does show up.

Cumulative -1 penalties for each search until you find something could model that.

For every hidden element, include at least three clues to its existence hidden around the dungeon (and not all in the location of the hidden element). This could be a treasure map with the secret door marked, or a scrawl on the wall left by the hobgoblins' last prisoner, or the skull of the last adventure to die in the trap, or scratches on the floor, or whatever. But be sure to leave the clues in places where the characters can find them.

Having explicit clues is important. So is having implicit clues -- if the "dungeon" makes sense, then the players can think through how things should work, where valuables might be hidden, and what kind of traps might be involved.