Percentages of dice rolls

[Shinizak]

From a purely percentage perspective, how often should characters be succeeding at their rolls?

60%? 70%? 80% or 90%?

What about new characters vs veterans?

[oxybe]

everyday tasks are a free pass. The fisherman fishes. the bakerman bakes. not every pot is brimming with crab or every loaf a winner, but those truly terrible results are outliers on an average day.

~65-75% on things that pose some difficulty or everyday tasks done under some kind of abnormal stress.

% goes down as difficulty rises.

same percentile regardless of veteran or newbie, just the scope of the tasks they consider difficult changes.

[Kaptin Keen]

If you've built your character to be a good shot, or competent craftsman, chances really should be in his or her favor when doing that - unless it's a particularly difficult task. Shooting in combat should have at the very least a +50% chance. Cobbling together a chair should never fail, but a result might decide how happy our craftsperson is with the result.

I suppose as GM I somewhat rarely ask players for rolls they're all but certain to fail. Instead, I tell them outright: This really isn't something you character feels he can do - but it would make the task much easier if he had X .. X being a workshop, a silvered sword, help from someone who reads chaos hieroglyphics, whatever.

[MoiMagnus]

In a very simplified way, I'd follow this procedure: one start at 0% of success, then:

0) This action is pure luck: ignore everything else and set the success rate at 50%
1) The task is reasonably feasible: +33%
1bis) The task is quite unlikely to succeed: -33%
1ter) The task is hard (not reasonably feasible, but not unlikely to succeed either): +0%
2) The character know what they are doing (trained in that action): +33%
2bis) The opponent, if any, know what they are doing: -33%
3) The character is actually good at it (high stat, or advantageous situation): +33%
3bis) The opponent, if any, is actually good at it: -33%

[Grek]

It depends on the stakes of the roll and how frequently that roll comes up.

Let's say we're talking about a game where there's a basilisk enemy that forces everyone in the party to make a save vs. petrification. Even if every character only fails their save on a natural 1, a five person party has a ~23% chance of losing someone every time they face one of those monsters. Even if you only face three basilisks before getting access to Stone to Flesh, you're more likely than not to lose a party member to petrification, and the player of that character will probably be frustrated at having to abandon the character they were invested in.

But if you change the stakes slightly: the basilisk's gullet contains juices which can reverse petrification, and as long as at least one person survives the entire party can be restored. That changes things. The 23% chance of losing someone every time a basilisk shows up becomes a 99.99996875% chance that at least one person survives and rescues the others. You could face thousands of encounters like that and never run into a total party wipe.

Interestingly though, the maximum acceptable danger goes up as players progress through the campaign. The further along in the campaign you are, the more dangers have already passed and the fewer rolls remain between you and the part of the story where such dangers are no longer potentially character-ending. (Even if there's never a cure for petrification, there's always the end of the campaign. At which point your characters won't be making any more rolls, potentially character-ending or otherwise, by definition.)

Likewise, if the stakes are low, the acceptable odds of failure are high. If the PCs are just showing off their skills at a tournament, it doesn't really matter whether they get the best roll, the worst roll or somewhere in the middle - the story goes on either way, with the players taking pride, eating crow or plotting revenge as appropriate to their characters.

[GrayDeath]

Most importantly of all: it depends on how GOOD your character is at what he is attempting to do.

For all other consideration, see the more elaborate posts above mine. ^^

[Anonymouswizard]

For a starting character in their specialty in a stressful situation without immediately fatal consequences I'd say approximately 60-70% of the time. The worse the consequences and the easier the task the higher those chances should be, but IMO if it's above 90% don't roll.

Although as noted it varies from person to person, a lot of people prefer an 80% or higher success rate in their specialty.

[False God]

For most of my games I tend to set things up so my players are succeeding about 70% of the time in at-level-challenges. Enough chance of failure to feel real. Not pointless as a coin flip. Rarely deadly outside of specific areas or facing challenges that are far above them. I run few, but hard encounters though, and each "25%" of success is to me, a measure of how many characters will be standing at the end of a fight. Maybe not permanently dead, but in any given fight I expect at least one PC to get KOed.

No real difference between new players/low-level-characters and experienced players/high-level characters. If the players start wanting more or less challenge, I'll adjust things.

I do however run my gothpunk horror campaign at a <40% success rate, but many of the challenges are not death. So it's survivable by any given character, but they might not recognize themselves by the end. It's a specialty campaign for a number of reasons and most people aren't up to the demands of the game.

[Pauly]

From a purely percentage perspective, how often should characters be succeeding at their rolls?

60%? 70%? 80% or 90%?

What about new characters vs veterans?

Generally speaking

A task it is possible to achieve but they are physically (or mentally) unsuited for and unskilled at ~ 10%
A task that they are physically able to do for but unskilled at ~ 25%
A task they are skilled at but is up to 25% beyond their normal physical abilities ~ 50%
A task that is “character appropriate” ~ 75%
A task that is either “easy for the character” or “character appropriate, but without time pressure” ~ 90%
A trifling task or a task they can take unlimited time to do - 100%

I always scale it to the character’s level and skill.

[Jay R]

Consider shooting an arrow at a single target. If it's close enough, there should be a 100% chance of hitting it.

As the target gets further away, the probability goes down, until there is a distance at which the PC has a 0% chance to hit it.

The DM's job is to set up the situation. The player's job is to try to arrange things to increase the probability of success (increase her PC's skill, get better weapons, sneak up closer, etc.).

So there is no base expected probability. If the PC aims at a target a mile away, the probability should be zero. If she gets right up to it, it should be 100%.

[Telok]

Paranoia: The Three Stooges doing a Soviet Russia skit written by a 4chan taught neural network learning program.

Success rates from 25% to 75% with degrees of hilarity & awesome replacing degrees of failure & success.

Champions: Point buy supers game.

Success rates starting from 50%ish (11 or less on 3d6) for Every-man and starting value skills. Rolled only in challenging or opposed circumstances like driving too fast on a road called 'Dead Mans Curve' in a rainstorm or when being shot at. Want to be a world class surgeon? Buy the skill up to 18 or less on 3d6 and you can only significant penalties can put you at risk of failure.

D&D: Depends on edition. In theory a zero-to-hero system. Has versions that go from 'nope to auto-succeed' and versions that go from a 15% success rate for untrained and handicapped 1st levels to a 15% failure rate for 20th level super-human experts.

What sort of success rates would I like to see in games? Untrained neophytes vary from 'nope' to needing luck depending on the task, high level experts auto-succeed at the stuff beginners have trouble with.

Except for Paranoia. It does the slap-stick comedy quite well with the 25%-75% spread for everything from nuclear engineering to boot licking.

[Kelb_Panthera]

That's a question with a few variables to the answer.

Taste is an obvious factor. Generally, you'd expect anywhere from 50% up unless the GM is just a power-tripping D-bag or one of the other factors is at play.

Genre is also a factor. Comedy and horror both expect failure a bit more often than heroic adventure with the consequences of such failure being obviuosly more severe in the latter.

Level also might play a role; starting nearer the 50% area at low level and moving ever upward toward nearer 100% with certain tasks by whatever your system considers high level.

Finally, and obviously, I presume that the question is about things the character spec's for; the things the character is supposed to be good at. Naturally, both level and taste might just as easily play a similar role in adjusting the odds for things the character is supposed to be mediocre or bad at from as much as 75% as often as with tasks the character is good at to as little as little as virtually no chance at all. Genre tends to be pretty even about both the things a character is good at and the ones he's not.

[JoeJ]

I think it depends critically on how often players are making die rolls. If most things are done without a roll, then the success rate when they do roll can be lower without the players feeling like their characters suck at everything. Also, the success rate can be lower if there is a mechanism like hero points that gives the players some control over which rolls are more likely to be the successful ones.