How do you feel about fudging die rolls?

[phaedrusxy]

I complain about this often enough, but I think part of the problem is the d20 - a flat distribution of results means you're as likely to fuck up catastrophically as you are to succeed beyond your wildest dreams. A better designed system might have results like the kind I'm describing be so improbable as to be not worth discussing.

You do know about this variant, right?

[Bauglir]

I do, and I also know it'll never become the D&D standard. This is a system so heavily dependent on the d20 that they named all the other games that use a similar design after the die. It's also not a perfect solution; you're still as likely to fail catastrophically as critically succeed, but you're much likelier to see average results. I would actually expect the bell curve to become skewed as you gain levels, rather than just get shifted.

@Unbeliever

Sure, that's why I said "well prepared" and emphasized that. Having rerolls available is an important part of preparation if you're in a game with players at that op-level, but you could also expect somebody in a less well-optimized game to rely solely on modifiers. My claim is that there are situations where the dice can negate player agency by feeding them a result that's improbably bad (or good, in theory, but it'd require players to be actively trying to be worse). What qualifies as "improbably bad" will vary from game to game, but at some point in D&D, you're almost certain to have a situation where the dice are telling you, "Nothing you did mattered, you fail."

[Unbeliever]

That's interesting.  I will have to give it some more thought.  I was mostly considering the agency concerned with (i) character creation, and (ii) expenditure of resources during play.  I don't know if I have given adequate consideration of the possibility of a die roll being the functional equivalent of "rocks fall, everybody dies." 

If you've got time, can you tell me how you think those thoughts play out in a game like M&M?  That's a game that has an explicit reroll mechanic, hero points, which are essentially action points.  It's also a game where the differences in your saving throw mechanics are a key way of differentiating characters -- everyone is supposed to have the same totals, but the Flash is a lot better on the Reflex while the Hulk is a lot better on the Fortitude.  It is still d20, so swingy, but I think we appreciate the issues there. 

Finally, is leaving that sort of thing in the DM's sense of fairness/judgment the most elegant solution?  I can see the argument for that -- if you can't trust your DM to be reasonably fair when something crazy happens, then maybe you shouldn't be playing with him or her.  Is it, in a game like M&M, largely taken care of with an action/hero point mechanic, so either DM fudging or magic items in baseline D&D a kludgy papering over of a flaw? 

[Bauglir]

Those are some good questions. I'd have to look at M&M, which I really should anyway since I hear almost nothing but good things about it. It sounds to me like those are good things for shaping the way the dice tell you what happens. Selective rerolls (action point types of things that let you reroll when you feel you need it) are a fantastic way of accomplishing the bell-curve skewing that I think a system designed to construct a narrative needs, especially if you're attached to the d20 or a similarly unskewable RNG. The different defense types are, equally, a core aspect to building a good game of that kind - you don't want characters to succeed at everything they try no matter what it is. Rather, I think it's good to make characters excel at their own specialties.

I know it is, by definition, impossible to get rid of freak accidents when you're using dice, and they lend a certain spice to the game that's part of the point of using them instead of flat modifiers. I don't think designers should try and fight their medium to that extent, but I do think that you need to give characters a feel that their success rate isn't just higher, it's more reliable, for the things that define them.

One of the things that has bugged me for a while, and this started occurring to me during my group's foray into Unknown Armies, was that they used a flat percentile system for everything. Even if you were a badass professional, at the top of your field, you had a 70% chance of success (as a normal human, anyway) at something if you were under stress. The whole system is an exaggerated version of the d20 with no modifiers, and they threw in a few mechanics that "felt right" to try and even that out for your special skills, which of course meant that if you wanted to feel useful you had to figure out which were likely to come up in the campaign or else compete with the other players and the DM to ensure that the story proceeded on your terms. Leaving aside their bizarre polarity decisions with the RNG, which I've ranted about at length elsewhere, you're locked into a paradigm with no room for diminishing returns or reliability, so your dedicated astronomer has the same chance of forgetting the Earth orbits the Sun as Stinky Joe the Hobo, because you share the same chance of rolling a 100.

Fighters in D&D have a somewhat stealthier version of the same problem. As is well known, a high level Fighter will roll natural 1s more often than a low-level Fighter, by virtue of iterative attacks. Basically, it seems like, your skill level shouldn't just change your odds of success, because that feels hollow and incomplete, particularly when you adopt rules denoting certain rolls as "special" (which I'm not inherently opposed to, since they add a vague sense of drama to each roll). In a well-designed system, your skill level needs to alter what failure even means, and it needs to alter the reliability of certain kinds of success. A level 20 Fighter should never drop his sword, no matter how poorly he rolls, without outside influence. It breaks immersion in the story, even if it's a tiny and inconsequential part of it. This is something we've all talked about before, it's just part of the overall thought process I'm trying to describe.

The conclusion I've come to in all this is that your RNG needs to scale with level. Kind of like nWoD, only less dumb. I don't think leaving things in the DM's judgment is, in itself, the most elegant solution. DMs exist partially to deal with exactly this kind of thing, but I think the system needs to include guidelines in its DM advice (which every system seems to have). Every game needs to be suited to its genre, but I think it's important to note that every DM needs to make an effort to suit themselves to the game, as well. Understand that, in D&D, for instance, you're dealing with Big Damn Heroes if you're past level 5, and you should rule their successes and failures accordingly. If you think it would've been a good thing to include in any of Luke's duels with Darth Vader, then it's probably okay - just understand that for adventurers, basically every day is like that, unless you're adopting a very different paradigm of play than the narrative construction one that I kind of assume for this sort of discussion. For instance, if you're playing a kick-in-the-door dungeon adventure with random encounters and loot, it's probably fine to let crit fails happen and fuck over a player, who can just roll up a new sheet. It can be pretty funny, and you usually don't have a mental expectation of your character's skill or the campaign's verisimilitude.

If you've got your system even vaguely well-calibrated, then the sort of situation I'm describing should be incredibly rare - maybe once every couple of years, if you're gaming regularly, will you run into this kind of situation in a party full of decent optimizers (utilizing rerolls and other ways of thwarting the RNG once you've got your modifier jacked up high enough). You need a lot of shit to go so badly that it's obviously the "wrong result" solely because of what the dice said. So, I guess in retrospect, I might be giving this sort of thing a little too much weight in my estimations. Still, it seems to be worth considering, for me.

[Bloody Initiate]

Generally I see this on a spectrum like:

Less GM Preparation<-------------------------------->More GM Preparation
Lots of fudging and improvising<-------------------->Very little fudging and situational improv

The better you prepare the less you'll have to work and improvise come game night. However no one is perfect, or even if they are their schedule won't be, so most people can't prepare games as meticulously as they might like.

Perhaps most importantly, you shouldn't expect your GM to have their shit together when they just found out they were GMing 2 minutes before the game started. However as a GM you should prepare as much as possible if you know far enough ahead of time. Basically players and GMs both have a job to do to make the game work, and neither should have unrealistic expectations of the other.

I personally like to let things run their course as much as possible, but have enjoyed the mercy of the GM too often to issue a blanket denial of mercy when I'm GMing. I PREFER to have things arranged and plotted so well that encounters will be engaging without being super deadly to players. However I am never that prepared because I'm lazy and don't GM that often.

That being said I voted for the "GM can fudge as much as they want" option in the poll because it doesn't actually contradict the other options. The GM CAN fudge as much as they want, they're the GM after all, but each will have personal preferences for how much they do it and the different situations will call for differing levels of GM interference. My personal preference is much stricter, but if I see fit to fudge a roll I don't want any fuss about it. I CAN fudge as much as I like, I WILL fudge as little as possible, aiming for "never," making my way there one game at a time.

[Unbeliever]

Hmmm, there's a lot there.  I have only 2 things that occur to me.  Oh, btw, M&M has an SRD:  http://www.d20herosrd.com/.  And, I don't want to be misunderstood, that system has its problems, but there are some things it does quite nicely.

I think I had similar complaints with the Warhammer 40k games from Fantasy Flight as Bauglir did with Unknown Armies.  Even the crack shots of the Emperor's Chosen have like a 70% chance to hit someone in an open field. 

Thought #1:  in games like D&D, which pull apart the attack and the damage/effect roll, there might be some difference between what a high-level character can do on a "hit" or a low level one does.  Fighter 5 does some damage, Fighter 15 does an immense level of damage and trips/stuns/disarms/cuisinarts at the same time.  At least, that sounds like an attainable ideal. 

There's probably something similar going on with spells, though it's a bit less obvious.  Higher level spells and maneuvers tend to do more awesome things. 

Thought #2:  barring wonky critical fumble house rules, the high-level Fighter miss option is pretty benign.  It's a cost for simplicity of the system and for making dice rolls always matter -- i.e., there's always a chance of success or failure.  It's true the classic high-level Fighter-type is bound to roll more 1s, they also matter a lot less to him.  Missing with 25% of your attacks is less of a big deal than missing with 100% of your attacks (i.e., your whole round).

But, it's a total kludge -- no doubt.  I probably tolerate it just b/c I'm used to it.  And, b/c the alternatives don't thrill me.