Larian Studios Forums Baldur's Gate III Baldur's Gate III - Suggestions & Feedback Natural 1 on skill checks

There is no such thing as critical success or failure on skill checks since 3rd edition of D&D. And as of 5th edition critical successes and failures doesn't apply even to saving throws. Those applied only to attack rolls.

Yupp. And despite what Larian my see in a lot of memes around DnD critical fumbles and fails on those checks are simply one thing: terrible. I will never understand people who enjoy them, especially critical fumbles on attack rolls. Those people are making their game effectively worse.

Many people have repeatedly informed Larian of this. The best we can do is to keep reporting it as a bug every time it comes up, and hope they listen.

You can find their official bug reporting form here: https://larian.com/support/baldur-s-gate-3?ver=4.1.1.1311526#modal

I keep linking it, because it's actually pretty well buried, compared to their general feedback form... but it's the best way to make a report that guaranteed gets read.

Critical failure and success can be done properly for skill checks, but it's much-much harder than just having nat 1 and 20 rolls. Second edition Pathfinder returned critical rolls on skill checks and saves since aging AD&D times. But it also have system of "locked" checks behind proficiency tier.

As example: wizard asked GM for an arcana check. GM approved, wizard rolled and failed. GM replied that wizard could not recollect any knowledge. Naturally wizard was at very least proficient in arcana.
Then party barbarian jumped in and asked if he can do arcana check as well. Obviously barbarian not proficient in arcana at all, and not even sports good intelligence score for that. GM nevertheless allowed barbarian to roll for check. If barbarian simply roll below required score - GM will reply the same, that he couldn't recollect any knowledge. However if he would roll natural 1 or below 10 of requested DC - GM instead will proved him with entirely false knowledge. And finally, even if barbarian would roll natural 20 - he would not succeed on check, since it would be locked at having at very least been proficient in arcana (with tier in PF2e progressed as: non-proficient, proficient, expert, master, legendary).
There could be skill checks that would not be locked at all, for example trying to recollect fairly common knowledge. But naturally lockpicking would always at very least would require you to be be proficient. You can't pick lock just with sheer luck, if you don't actually know how to do it.

Nevertheless, that's about 5th edition D&D, not 2nd edition Pathfinder. And in D&D 5e critical rolls applied only to attack rolls.

The concept of only being allowed to roll in things that you are proficient with is something that 5e does, too. Or at least most DMs and modules have some of those. Arcana, like you said, is the prime example. Its not something everyone can do. But other things could work as well. This gives skill proficiency more meaning instead of just rolling a little better.

I also think BG3 already does this at least with Arcana at a few points in the game.

(Edit: Started writing back to Redwyrm, just took too long ^.^)

That said, you're describing a simple (and slightly cruel) version of the way appropriate skills checks are designed to work in 5e anyway: checks are individual, DCs are individual, and the results are also individual.

In your situation, at a 5e table with a good Dm, it might look like something this:

The DM might internally set different DCs for different characters, as well as setting different sets of information. This is normal, attentive DMing that helps everyone feel like they're contributing, somewhat prevents the extremely unsatisfying ridiculous check moments, and helps give the people who should have a better shot at doing something a legitimately better chance of contributing meaningfully.

Here, an adventuring party are making Arcana checks to see what they know about a sigil pattern. The DM looks at the group, and, in their mind they might say to themselves:

The Wizard: Has a good background knowledge regardless - they are going to get information pieces "A" and "B" regardless, as long as they make a check... Their DC will be about... 8 to get information piece "C", 12 to get information piece "D" and 15 to get information piece "E"... and I'll give them some juicy lore-info if they crack 25, because I know they could potentially get that high.

The Bard: Has some background knowledge, but no detailed training... DC 5 to get "A" and "B", 10 to get "C", 16 to get "D" and 21 for "E", and their information will be largely the same as the Wizard's, but with an allegorical twist.

The Sorcerer: I know they're trained in Arcana, but they operate more in feeling and instinct for the magic, and don't have any academia behind them... They'll get Information piece "D" automatically if they check. They can't really get information "A", not really, but they'll get a variation on "B" and "C" if they hit DC 10, and a variation on "E" if they get DC 17... and if they break 25, they'll get something cool based on the sense and feeling of what's here in front of them right now.

The Barbarian: No training, but they can check if they want. They can't really get information piece "A", but if they hit DC 15 they'll have heard of rough approximation of information "B", and I'll give them colloquial accounts related to "C" and "D" if they get 19 (which is their nat 20).

I haven't played with critical fumbles in ages, but it really depends on what system for it you are using (automatically stabbing yourself in the foot is always going to be the least fun way to do it) and what kind of game you are running. I'd be a lot less interested in having them in a long, more serious campaign.

But there's also a big difference between critical fails and critical fumbles. Critical fail is just 1 = failure regardless of modifiers and I think there is some legitimacy to having them for certain kind of skill checks where failure is always an option.

But that aside, I'd rather BG3 didn't do them at all.

3.5 and PF1e usually proposed fumble like: weapon slipped out of your hand and dropped on the ground. You axe not just missed target, but strike tree behind it and stuck in it.
In all cases usually means you will have to spend next action to "recover" from critical miss.

I always just felt as though "you failed your action, and thus wasted your turn achieving nothing" was punishment enough for the failure, personally.

Are you calling my life a punishment niara

Yeah I know the difference, thats what I meant by "especially critical fumbles on attack rolls". Those are just an insane tool, if you really hate your fighter player.

Well, that's why fighters specifically have Great Weapon Fighting.

This is either a joke I dont understand or maybe you did not get what I was saying.

Is there something off the screenshot that shows critical failure, instead of just a failure? This image shows a DC of 10, with a Roll of 1. Since no bonuses to the skill check are shown in this screenshot, a roll of 1 to 9 would be a fail.

It never even applies modifiers when rolls natural 1. Also no, for that roll in particular the very minimum Astarion could roll is 10. Specifically 1 + 5(dexterity) + 3(proficiency) +1(guidance).

Aww, come on now, Dex, I'm sure you've achieved plenty in your life so far ^.^

=

Just to confirm the topic - it is as Redwyrm says. If you roll a 1 on an ability check like this, it just jump cuts directly to the failure animation, without ever adding the modifiers, even if those modifiers should mean that you cannot fail the check.

Also, Halfling luck has not been implemented at all and doesn't happen in game, despite it being listed on your character sheet.

So you're objecting to the possibility of failing a skill check? If that's the case, why even have them? There should be a chance that you can fail something.

For that purpose DC exist and varies. Astarion might automatically succeed on opening trivial DC 10 lock. But still will struggle with hard DC 20 lock.

Plus. It neither me nor Larian who made D&D rules...

Indeed, and reading through the thread, there's an optional rule to apply a fail state on a roll of 1. It looks to me like they're using that, but as I'm not a developer working on the game, I can't say. Anecdotal evidence, such as your screenshot, would suggest that they are.

it's just a regression, it worked fine on the previous patch

yes. and so does WotC :