Quote from: jhkimSeriously, could those in favor of repetitive rolls give some more concrete examples of when they think it is significantly better than a single roll?

For me it's all about the math and the discrete decision/evaluation points.

I remember a discussion I had a long time ago about interpreting "pick locks"--in general, not specifically D&D, probably working from RQ as a baseline. Could you try to pick a lock multiple times? One answer was basically the "Let it Ride" rule: only once your skill goes up. Another was, each roll takes a discrete amount of time. Another was, each subsequent roll takes an exponentially longer amount of time. Another was, try as many times as you like, but if you crit fail, you've jammed the lock. Any of these can be simulated with a single probability function, so in theory you could do it with one roll, it's just a matter of calculating what to roll. E.g., you could look at someone's skill, ask them how long at most they're willing to keep trying, convert that into a table, and roll percentile dice to see if they open or jam the lock (and how long it took), or if they give up after the time allotted.

But for any of these latter approaches to work (except maybe "jamming" on a crit fail), time has to actually matter in the game. It's implicit in the mechanic itself; if a GM doesn't enforce passage of time, then the rule doesn't work. If they do, then iterative dicerolls are an effective way of representing just that fact: there are some tasks that you never know how long they'll take, or even if you're just about to succeed or will never succeed. Like trying to catch a fish using a baited hook, for example. A single roll with "let it ride" introduces an artificiality to the game, where people give up after a single brief try. I know that BW has rules for working slowly & carefully; I forget if you can try that after an initial failure, but at some point you're back to iterative rolls.

The other thing, decision/evaluation points, is what I pointed to when I was writing about sneaking into a fortress. It looks like Tim's method is to use iterative rolls for each guard, which I'm not sure is actually in the book, but it's a good solution to answering the question of when & where someone finally notices you. It's not available though if you're testing your climbing ability to scale a 20' wall or a 100' cliff.

Elliot hits the nail on the head, as usual.

One of the big problems with RPG best-practices is that the baseline from which the game is being discussed is very hard to nail down. So I'm going to give an example of where I think multiple rolls are a good idea.

In the JAGS Revised drama roll system a situation may call for a number of rolls (usually three) to accomplish a task instead of one. Drama-rolls are used when there is a game focus on some non-combat skill (such as drivers in a racing game or the act of picking a lock in a tense thieving adventure).

There are two key elements of the Drama Roll:
1. Each roll may cost something like time or money (i.e. the lock has a difficulty of +20, meaning you have to roll until your aggregate total of successes is 20 or more and each roll takes 30 seconds and there's a guard coming).  In this case, the decision to keep going or bail is important since the player knows how close they are to success and gambles his ability against the risk of being discovered.

2. The character can take actions between rolls to change the outcome. An example is in a driving race where the driver who makes their three rolls by the most wins the race--a driver can choose riskier maneuvers (possibly resulting in a crash) to gain more successes. In short, it takes a single abstract roll (driving) and turns it into a mini-game where the success is drawn out and the player gets to make decisions "during" the resolution.

Now, all this is cheating in response to the question since the many-rolls-thing is probably meant to be "a bunch of rolls for 'no reason.'"

The idea of building tension by having multiple rolls for success is not alien to me--but it's not my cup of tea either (and if we're going that way, the GM better do a really good job of building tension--not just having me roll ... and roll ... and roll ...).

Secondly, representing "reality" or some rule-guideline is okay (saying that you must make a climbing roll ever X-feet makes it riskier to climb a high cliff than a low wall). But there are reasons for rule-zero type things to cover situations where the players don't want to do this--and this is one of them. The GM can decide to just have one roll where that element of verisimilitude isn't wanted.

A final note: it's very possible for GM's to "fail their GMing roll"--in fact, it happens, IME, especially to me. After a game I can think of things I shoulda done better but, being on the spot (GMing is a performance art) or just not at the top-of-my-game, I didn't.

I think a lot of what gets ascribed to bad / malicious GMing is really just a GM casting about for something that'll work. Some people do this better than others--but I think an all-too-often combination of generic poor communication and lack of stated goals leads to situations where the GM may feel that having a bunch of rolls could help and the players are dug in against it.

This isn't so much a many-rolls thing or a let-it-ride thing. It's more about the GM maybe realizing that time-pressure in the game will lead to a kind of tension that just asking for a bunch of rolls isn't doing ... and the players maybe realizing that there's a lack of energy at the table and rather than resenting the GM, just taking a step back and talking about it might help (and if none of that does help it's the people--not the game).

-Marco

Quote from: jhkimSeriously, could those in favor of repetitive rolls give some more concrete examples of when they think it is significantly better than a single roll?

Take a look at this map from Jeff's blog.  It's from Palace of the Vampire Queen, the first adventure module ever published.



If a character wants to try using stealth or find traps on that long central hallway from the entrance to room 51, I think it's better to have them make a check each turn than any kind of "let it ride" rule.

Rather than getting 1 success -- and letting walk right down the length of the hall, they need to consider whether to keep risking failure, or if they should stop pressing their luck and duck into one of the side doors or hallways.

Let it Ride is great for a storytelling game.  Not always so great for a traditional RPG where the focus is on exploration and adventure.

Quote from: StuartIf a character wants to try using stealth or find traps on that long central hallway from the entrance to room 51, I think it's better to have them make a check each turn than any kind of "let it ride" rule.

Rather than getting 1 success -- and letting walk right down the length of the hall, they need to consider whether to keep risking failure, or if they should stop pressing their luck and duck into one of the side doors or hallways.

Let it Ride is great for a storytelling game.  Not always so great for a traditional RPG where the focus is on exploration and adventure.

OK, but...

As a player in dungeon crawls, I hate rolling for find traps every turn.  To me, it would just drag things out pointlessly.  This isn't out of a desire to make it into a storytelling game -- this is just that I'd prefer to move on to the monster-fighting and other more interesting action.  For example, here's an LJ post, Avoiding Trivial Stakes where I talk about being frustrated with having to make a bunch of rolls dealing with a set of pits in dungeon in a HarnMaster campaign.  This isn't finding traps, but it is a similar static challenge where we had to make repeated rolls.  

As a DM, I might roll in advance for the PCs for when they find a trap.  Alternatively, though, I might just have them either automatically notice traps or never notice traps, because neither randomizing it nor the time management of doling out rolls are exciting to me.  

Now, to Marco:  Yes, if you're choosing a notable maneuver with each roll, that's different.  For example, James Bond 007 has an interesting chase system with maneuvers rolled each round.  I haven't played JAGS, but the maneuver system in the Drama Roll system sounds potentially interesting.

Similarly, I don't consider combat to generally be repetitive rolls because there are a variety of options each round.    Though if combat reduces down to only one reasonable choice rolled over and over, I feel the same way.  For example, I recall a one-on-one duel between two high-level cavaliers in an old AD&D game.  It became a real snooze since they only had around a 20% hit chance, and they needed a huge number of hits to take the other out.

QuoteOk, but...

Address the Stealth check.

Quote from: StuartAddress the Stealth check.

Sure.  I don't like rolling Stealth every round either, for similar reasons to not liking to roll for finding traps.  As for how I'd handle it...

In D&D, I'd either do a secret set of rolls for the party again, not make it an issue, or just make something up.  

In Burning Wheel, the party would roll for Stealth once -- getting a given number of successes (or failure).  That would then be the base target to beat for a faction in the dungeon to notice them.  If there are multiple creatures in a faction (i.e. a group of 7 orcs), then I'd roll for the lead orc with other orcs adding helping dice to represent the benefit of numbers.  That could be modified by particular circumstances for a given creature.  

No, this isn't statistically the same as rolling each round, but it works fine for me.  I'm not convinced of the realism of any particular probability curve anyway, especially not for an activity like sneaking.

Quote from: TimMaybe it seems like a dumb rule to good kind-hearted wise GMs like yourself, but I fail to see what harm is done by having good GMing techniques codified in the rules.

Things change when they go from being optional to being mandatory.  Maybe you think desserts are a good way to end a meal, but they can be a problem when you are forced to eat one whether you want it or not.

Quote from: jhkimAs a player in dungeon crawls, I hate rolling for find traps every turn.  To me, it would just drag things out pointlessly.  This isn't out of a desire to make it into a storytelling game -- this is just that I'd prefer to move on to the monster-fighting and other more interesting action.

To me, it's not pointless if the game is played by experiencing the moment instead of getting to the "interesting action".  If I want to travel to California and getting there is my goal, it makes the most sense to fly because I'll get there quickly but I won't get to see any of the scenery along the way, except from tens of thousands of feet in the air.  If I want to see the scenery along the way, it makes sense to drive.  It will take me a lot longer to get there, but I'll see a lot more along the way.  And which one of those options is right for a person will depend on what they want out of the trip.

In many ways, I think impatience is one of the biggest features of many Indie games.  It looks an awful lot like a Reader's Digest or even Cliff's Notes version of role-playing to me.  (ADDED:  And if all you've ever read is Dickens, I can see where a Cliff's Notes approach to role-playing could seem mighty appealing.)

Quote from: John MorrowTo me, it's not pointless if the game is played by experiencing the moment instead of getting to the "interesting action".  If I want to travel to California and getting there is my goal, it makes the most sense to fly because I'll get there quickly but I won't get to see any of the scenery along the way, except from tens of thousands of feet in the air.  If I want to see the scenery along the way, it makes sense to drive.  It will take me a lot longer to get there, but I'll see a lot more along the way.  And which one of those options is right for a person will depend on what they want out of the trip.

But...  repetitive rolling isn't interesting scenery.  I mean, I'm a Harn player -- one of the most detailed fantasy worlds there is.  For our game, we made out a detailed map of Bowdyn's shop, discussing about how the heating would work.  We have sermons written out and so forth.  

But rolling every round doesn't add anything interesting for me to the dungeon.  It's not like there is some fascinating details that are revealed if only you'll sit around and roll on Find Traps a whole lot, is there?

Quote from: jhkimSeriously, could those in favor of repetitive rolls give some more concrete examples of when they think it is significantly better than a single roll?  

I can think of many times when repetitive rolls wouldn't be so bad, but I can't think of any cases where the rule actually causes any problems for me.  (Though this might be a difference in taste compared to others.)

Here's where I think multiple rolls are a good idea... but before I start, nothing I'm going to say defends a D&D-style rolling for traps each turn thing.

In general, I see asking for multiple rolls as being advantageous in a few circumstances:

1) Represent a high degree of success over time (e.g. combat, hacking rules, etc.) -- for rules where success is additive and failure isn't absolute asking for multiple rolls ensures that average results are more likely than really unusual ones. This makes characters perform more reliably and gives high-skill characters their due

2) Represent situations where failure is -- ultimately -- likely (e.g. sneaking all the way through enemy lines), except for extremely high-skill characters. This would be the case when *any* failure represents failure of the whole exercise. This is important because most systems don't give good alternatives to modify success chances in this way.

In the sneaking-through-enemy-lines situation, I'd rather give modest skill-level characters multiple rolls (with a very high total-chance of failure) than a single highly-modified roll.

I'd want to work out how many rolls with the players ahead of time (e.g. "To sneak through the encamped army to the General's pavilion, you'll need to make 12 Sneak rolls... the last 2 will be through his Elite Guard and will be at DC 15).

One added advantage to this is the opportunity to modify approach and plans during the attempt (so the PC's might seize an opportunity to back out if they realized they're not likely to make it), rather than simple and immediate binary success or failure.

3) Represent passage of time. Maybe fixing the spaceship requires 3 successful engineering rolls and each one takes 1 week -- so it's a minimum of 3 weeks, but potentially a much longer period. I'd only do this if time mattered, of course. And it would be more meaningful if the PC's could make decisions in between rolls that make a difference (e.g. "We fire our crappy engineer and hire the babe he was sleeping with...")

There are single-roll methods to do all these things, of course, but I find multiple rolls more intuitive and a better way (in these instances) to represent  these kinds of in-game situations.

Now, thinking about the "roll for traps" which is kind of the poster-child for Don't Do This -- if I were running the PC's traversing a mine field and time was *any* kind of factor I'd probably ask for multiple rolls: I think it would add tension to the scenario, with the chance of disaster hanging over every roll.

But I wouldn't be much interested in running or playing in a game where *every* move was played out like that.

Tension, over-done, becomes frustration and boredom.

Cheers,
-E.

Quote from: jhkimBut rolling every round doesn't add anything interesting for me to the dungeon.  It's not like there is some fascinating details that are revealed if only you'll sit around and roll on Find Traps a whole lot, is there?

I don't think the characters need to roll for every square.  But I wouldn't necessarily ask for one roll to get to orc killing, either.  If I had a 100 foot corridor in D&D with 2 traps in it, I'd ask the players for a roll.  If they make the roll, I'd have them find the first trap.  If they fail the roll, I'd have them miss the trap and potentially trigger it.  Then do the same for the next trap.  And also do the same if they say they are looking for traps, even if there aren't any.  Similarly, stealth could be handled on a per-observer or per-observation point basis.  Secret doors are a bit different, because failing to find them can mean walking past them.  I'd probably ask for some non-zero number of rolls in that case, regardless of whether there are any or not.

Perhaps the real problem here is the idea of asking for skill rolls even when they aren't in response to anything.  In other words, the character rolls stealth when they start sneaking, not when they are observed. They roll their trap-finding skill when they search a square, not to see if they find a trap.  I think the latter can often work better, though some obfuscation is needed when asking for a roll can give the players information (e.g., if I only ask for a find secret doors roll if there is one there, and I ask for 2 roll sand the players fail the, they'll know that they missed 2 secret doors).

I think thats right John Marrow.

In my game there would be multiple rolls. But only when there is a chance of failure meaning anything. Only when the orcs have an opportunity to hear the sneak. That wouldn't be every round or every five feet.

I have something else to add, sorry if it isn't part of the conversation, but I as DM would make the stealth rolls. The player would let me know they are trying, they wouldn't know if they made it or not nor even how bad or good they really did (OK I might give them some idea).

Why doesn't doing that make sense?

I had to READ "let it ride" again out of BW.  It reads to me as a rule for minimalizing rolls and making the roll matter in regards to the players intent.  

Climbing checks are the first thing that came to mind.  Let's say that your intent as a player is to scale a cliff.  In D&D, pulls out book, you move at 1/4 your speed up the cliff.  Each move action that includes climbing requires a seperate climb check against the set obstacle.  3.5 uses taking 10 or 20 to eliminate the chance of eventual failure.  Looks at those rules, I'll be damned climbing is the example given.  A player cannot take 20 because there is a penalty associated with failure (falling).  A player cannot take 10 when they are being threatened or distracted.  So I could take 10 as long as nothing is happening to scale the wall without multiple rolls.  

The "let it ride rule" allows for one roll against the obstacle to determine the success or failure of scaling the cliff.  If there is something happening, that wouldn't allow me to take a 10 in D&D, the obstacle is increased in BW but still would only warrant one roll unless the distraction happened during the climb to change conditions legitimately and drastically.  

It seems to me a preference of how much entertainment one gets of fighting their way up the cliff.  I think the reason taking 10 or 20 were implemented into D&D is that a fair percentage of people don't.  Let it ride seems like a natural extrapolation of that.

Quote from: John MorrowThings change when they go from being optional to being mandatory.  Maybe you think deserts are a good way to end a meal, but they can be a problem when you are forced to eat one whether you want it or not.

Yes, I always try to take my meals in the Gobi or Sahara. :raise:

Anyway, you're the one eating the highly decorated three layer cake that's ALL the same flavor and took four hours of prep to make, not me.

Quote from: Elliot WilenIt looks like Tim's method is to use iterative rolls for each guard, which I'm not sure is actually in the book, but it's a good solution to answering the question of when & where someone finally notices you. It's not available though if you're testing your climbing ability to scale a 20' wall or a 100' cliff.

No, I would make one roll for each unit (for the lack of a better word). So, you might make it past the mook guards, no problem, but run into difficulties with the necromancer's elite undead sentinels.

If we were working at a very detailed level, individual guards might get some situational modifiers (light levels, high terrain, that sort of thing) that could give them a boost or a penalty.

Rolling for each individual guard would totally subvert the purpose of Let it Ride and be a pain in the ass, in my opinion.