D&D 5E Using 3d6 for skill checks

[DEFCON 1]

The big thing for me is that due to CoS being a lowish-level adventure... the numeric swing between untrained and bad-abilitied vs trained and strong-abilitied when it comes to a particular type of check is not that great. A 1st level PC with a WIS of 8 and not proficient in Survival gives them a -1 on Survival checks. A 1st level PC with a WIS of 16 and proficient gives them a +5 for Survival checks. So that's only a swing of 6 points between absolute best and absolute worst. And if you throw in the Rogue and Expertise, the swing might be 8 points between absolute best and absolute worst.

For some people, an expert having a 30% better chance (under the probably-faulty mathmatical assumption of each point of a d20 being 5% better) to do things that a fully unqualified person could do might be fine. Especially if you add in concepts like "Only the trained can roll" or any other qualifiers like that which various players use to adjust the ability check paradigm for their table. But for other folks... maybe those 6 points aren't enough of a divider between expert and completely unqualified? If that's the case... then changing the die roll such that those 6 additive points have a greater impact on the odds of success is a way to go.

Maybe not the best way... maybe not even better than the normal way. But until the individual DM tries it, they'll never know if it works for them.

 

[I'm A Banana]

Well we certainly need to keep in mind that we have different ways of running the game so that the context into which the rules of choice operate can vary a lot.

For example, one element of my own context is to never place a skill check (or ST, etc.) at a critical point of the adventure such that failure in the check means failure in the adventure. So whether you don't have a Rogue, or your Rogue fails at lockpicking, the game won't be screwed up. If you really need to get past that door, there are always many alternative, and if all you can think during our game, is a thieves' tools check, then you're not thinking enough

I also don't encourage people to design the party cooperatively at all costs. I rather encourage cooperation during the game, but not beforehand, because I don't like to force latecomers feel they have to comply with leftover roles.

If a GM told me to create whatever character I wanted because part of the social agreement here is that only the skills I actually took would be relevant, I'd personally feel like the decision of what skills to take would be pretty meaningless.

"What skills do I have? Does it matter? If I take thieves' tools, we'll encounter locks. If I don't, we won't. If I take Persuasion, we'll need to do persuading. If I don't, we won't. Lets just say I am Skilled +x, at Whatever The Plot Demands, and lets get to the decisions that matter."

Also, after character creation, a system where someone who is untrained has essentially no chance encourages a party split. "Oh, this is a scene where Persuasion needs to happen, everyone who dumped Cha, just sit and watch." / "Oh, here's some locks we need to pick, guess Ed's gotta do it, I'm gonna go get a soda." / etc.

It'd be a bit like in combat if you had a Fighter Who Could Fight Humanoids, a Dragon-Slaying Dwarf, and a Barbarian What Kills Wolves. If this fight has humanoids, the Fighter gets to participate, but the rest sit out. If this fight has Dragons, the dwarf can participate, but the rest sit out. If this fight has wolves, the Barbarian can participate, but the rest sit out.

The more swingy d20 helps everyone participate - a -1 to a roll means you can get a high roll and still contribute a success.

...which isn't to say the other way is bad, just that there's some tradeoffs for that more strict division, and whether or not those trade-offs are worth it is going to vary from campaign to campaign.

 

[Springheel]

"What skills do I have? Does it matter? If I take thieves' tools, we'll encounter locks. If I don't, we won't.

You could just as easily take the same approach from the other side....

"What skills do I have? Does it matter? If I don't take thieves' tools, I'll have nearly as good a chance of success as if I do."

I think the entire point of having races and classes is because we want characters that are good at specific things, which should translate into them not being so good at other things. You play a fighter because you want to be good at fighting; you play a rogue because you want to be good at stealth, etc. Obviously we want some specialization in the game or there would be no need for different classes at all.

One possible tweak to this that occurred to me while reading is to use 2d10 when the characters are in relatively calm conditions (casually playing an instrument, shooting for target practice, trying to remember a history fact during a council meeting, picking a lock in an empty room) and use a 1d20 during highly stressful situations (combat, picking a lock before the guards arrive, playing a performance in front of the notoriously unpredictable king). That way you'll still have the chances for wild successes or failures when it has the most story impact, where under normal conditions more skilled characters will outperform their untrained peers more regularly.

 

[Li Shenron]

If a GM told me to create whatever character I wanted because part of the social agreement here is that only the skills I actually took would be relevant, I'd personally feel like the decision of what skills to take would be pretty meaningless.

"What skills do I have? Does it matter? If I take thieves' tools, we'll encounter locks. If I don't, we won't. If I take Persuasion, we'll need to do persuading. If I don't, we won't.

No no, it's not like that...

You will probably encounter locks anyway, as long as those locks make some sense in the adventures. If you have a Rogue (or lockpicker) you'll pick the locks, and if you don't have it, your party will find other ways to deal with locks.

Same thing if the players build a party low on combat capabilities. It just means their story will be more about avoiding combat rather than throwing yourself into battle automatically. But they can do it by themselves, it's not me who always needs to force the story so there is no combat, I just need to make sure I don't force the opposite either (i.e. that I don't make something unavoidable if they can't handle it). *

I don't "retrofit" the challenges, because my assumption is that if everybody wants to play fighters then they probably want to fight, while if everybody wants to play a bard then they probably want to deal with quests differently. It's their choice, by I don't necessarily need to adapt everything around the party, I just need to make sure that they can find a way to progress in the story whatever they capabilities.

That said, making adaptations isn't wrong either, but I tend to do it more about the theme or the general type of adventure: for example if a player designs an undead or dragon hunter PC, then I do feel responsible for featuring undead/dragons in the setting (theme), and if they all want to play explorer-types I have to make sure the exploration pillar is represented plentifully. But there's no need to remove all locks from the world just because there is nobody in the party who can pick them using the standard skill check.

* I suppose this is more understandable, if you know that as a DM I try to be more like a referee and less like a movie director, and I favor a "sandbox" approach to campaign design

 

[Bruce Vistani]

I don't see skill checks as binary succeed/fail rolls. I see the modified roll as being along a continuum from spectacular failure to remarkable success and success or failure is a bit grey around the DC. Sometimes I don't even have a DC in mind when my players roll. I let the result define the narrative.

You don't need a bell curve roll if the results are already curved.

 

[Mercule]

So, if that's your definition of swingy, that 3d6 can bin you into auto-success/auto-fail faster than d20 can, then, yep, no arguments. Odd definition of swingy, though.

Agreed. My second post was a half-hearted admission of the poor word choice. Not sure what the right word is, though. Binary?

 

[Professor Periapt]

This mechanic for skill checks based on character attributes appears in Dragon #1. I haven’t tried it out.

Roll d00 and add character attribute to the result.

Compare it to a percentile table divided into five equal ranges representing five different dice from fewest sides (01-20 = d4) to most sides (81-00 = d12).

Roll the die indicated and multiply it by the attribute in question. The result is the percent chance of success.

Attempting things properly in the domain of a different class (thievery, using magic items) does not get a multiplier. The chance of success is the character's raw ability score. DM's discretion to halve or quarter the chance of success and introducing chance of catastrophic, backfirish failure.