Baldurs Gate 3?

For sure. It’s just seems odd that instead of the game’s tables themself be adjusted to make things more forgiving you just make a table then fudge the rolls. And this is coming from someone who has never expected someone to play with a bunch of shit rolls during character creation. Anything less than a 10 is outright ignored. But I guess if fudged rolls can be built into the design, why not.

The system is based around a 5% chance of critical success or failure in a lot of ways though.
Nat 1 and Nat 20 matter in 5E quite a bit, so making that 5% event a 1% event is a pretty big shift.

It’s a problem I actually have with D20 based systems (if everything in life had a 5% chance of failure, no one could do anything and everyone would be dead since driving to the store would kill hundreds of thousands every day and we’d all fail at our jobs to the point that we’d starve in the cold anyway), but it feels weird to like… change random to fix a system that is founded on that level of random. Especially with 5E where the scaling is generally not kind for some things. Like in a bell curve, you’re basically never making any save you aren’t proficient in past level 5 or so. Ever. But you’ll basically always succeed if you are proficient.

I don’t mind a little randomness in a game to prevent things from being too deterministic, but when the player can just immediately reload whenever they fail a skill check, it is a bit pointless.

Seems weird to me. Dealing with bad luck is part of the game.

Hi,
Have you ever seen people…gamers(!) discuss missing an 85 percent shot in Xcom? Its…eductional!

While I’m old enough to have loved and played many dice games, including D&D but many others, I also think it’s actually kind of a bad game mechanic by modern standards and doesn’t simulate chance in a particularly good way. An ancient lich who has ascended to demi-god status has a 5% chance for his mind control to fail on a peasant. Try writing a book with a 5% fumble chance and it will be a comedy, not a drama.

I’m not married to the d20 system is all I’m saying I guess. There are other ways to explore probability.

Oh I want randomness, but in other ways. Combat rolls. Damage. Encounters. Loot. Maybe even a dungeon.

For CRPGs, skills checks have long since been solved with thresholds instead of random die rolling. With a D20 the failure chance is stupid high even if you focus all your points into the skill being checked. It’s totally obnoxious and feels like a regression.

Some of the best and funniest session stories from CRPGs come from those fumbles and aberrant rolls. As a Shadow monk, adept at hiding in the shadows and moving undetected, I once fumbled twice in a row, not spotting guards and making a ruckus. The fun the party had undoing that mess made for an awesome end of a session. Dice might not be an awesome or balanced game mechanic. But as a storytelling and mishap generation device, they are awesome, IMO.

I can see how they are less interesting in a computer game though (but the game is implementing D&D after all).

Disco Elysium solved the skill check thing pretty well - all you have to do is let the player fail forward in funny or interesting ways.

Obsidian’s JE Sawyer weights in on this issue. I agree with him. And CRPGs are fundamentally different from PnP.

I’ve always been a fan of the d100 and combat mechanics of PoE 1 & 2.

I feel like these objections could be ironed out if failures were not catastrophic but rather were interesting to deal with. I’m pretty sure that was suggested above and I agree with that philosophy. Of course, that’s probably 100 times more difficult to implement then just having a failure mean you’re spotted or you fumble or whatever.

I believe Larian is doing that to some degree, but just to dialog and other choices. The concerns, as far as I can tell, tend to be from not being able to get into a chest or a locked door, or when you fail an arcane or history check and miss out on interesting information or details. The reality is stuff in locked chests or behind locked doors in games like this don’t tend to really be a big deal - just a little more loot to sell later, really. It’s just that it’s a “feel bad” moment.

Yup. Other PnP RPGs have built this sort of storytelling feature into the system, like Fate that mandates “complications” or “consequences” when a skill check fails. You don’t just get a negative result. You get a result that allows the story to continue, just not in the way the player intended. A locked door is just a dead end, but a door that opens but alerts all the guards is a story opportunity.

This is stuff an experienced DnD GM will do anyway, but in a CRPG it’s much harder to implement that kind of story-conducive failure.

Yeah, Fail Forward design from stuff like Fate and PbtA is the optimal solution to “failing rolls sucks,” but requires a high degree of responsiveness – especially if you want to go with a particularly striated roll-result possibility chart (Fate accounts for Failure, Tie, Success, and Success with Style–akin to a crit). It can be tough as a GM, and takes some real training to remember not to force yourself into that position unnecessarily by calling for skill checks at times when there’s not really any compelling tension to be resolved. For instance, a competent thief should be able to just pick a standard lock if given sufficient time and privacy. If either of those factors aren’t guaranteed, the downsides/negative outcomes of poor rolls are way easier to generate on-the-spot, even if they still get some or all of the loot.

As an aside, both of those systems also use bell-curve distribution core dice mechanics. Fate has the 4 Fate dice (d6s with two sides each for -1, +1, and 0) that cluster around a flat result (so the character’s Skill rating – generally +0 to +4-- is extremely important) with very narrow tail possibilities of +4 or -4, and a typical baseline difficulty of about 2 for “not guaranteed but not crazy challenging” style tasks. PbtA just uses 2d6 with a more regular curve peaking at a 6 on the dice and usually relatively modest +0 to +3 bonuses from abilities or situational perks; the result bands are 0-6 fail, 7-9 mixed success, 10+ full success (and some advanced abilities unlock a crit-style perfect succcess on 12+). In either case, you leave off the crazy-town swinginess of d20s where, hey, your 10-year veteran soldier has an equal likelihood of beheading a red dragon with a single swing of his vorpal sword as he does of cutting off his own foot every time he makes an attack. Ugh, d20.

Something like the Crisis system from Torment, where major challenges were carefully hand-designed with lots of interactivity and cascading result-possibilities, are basically the best way to handle this in a CRPG where the “GM” can’t just pull an interesting, story-advancing failure-result out of their ass, since said GM is an unintelligent machine. But they also require a ton more work than just “chest need check DC 20 hurrdurr,” and hey, game development’s expensive and time-consuming.

On the other hand, you can’t save and reload your game on table top.

Also rolling a 1 on an attack or save does not mean you cut off your own foot or shoot yourself in the face. It just means you missed. It doesn’t automatically invoke the spirit of Wily E Coyote .

No no I’ve been assured by nearly every D&D game I’ve ever suffered through that nat 1s are prime opportunity for “hilarious” GM fuckery!

Edit: okay that’s spectacularly ungenerous to my regular d20 GMs; I swiftly remove the gamers who play that way from my social circle

Indeed. There is not really such a thing as a critical failure in 5E. A 1 on an attack means you will miss and is bad on a death save (2 failures). But you aren’t cutting off your own foot.

As for ability checks and saving throws, they do not necessarily fail on a nat 1 (though you likely face an uphill battle to meet the check requirement).

It can be a lot fun to describe in more detail how your critical hit or your missed attack panned out from a roleplaying point of view. But they aren’t really the place for GM fuckery, and I wouldn’t play with a GM who thinks so.

It might sound strange, but story has been one area where the data has been a boon. At one point during our talk, Vincke pulls up an image to show me what’s essentially a graph of the paths players have taken through a specific story dialogue, and what outcomes they received. Vincke estimated it covered about 30,000 dialogues’ worth of data, with some results being clear majority and others, ones that might require use of certain mechanics or knowledge, having fewer players reach that conclusion.
Vincke describes these analytics tools, which the studio built up after the success of Divinity: Original Sin 2 , as a “gold mine.” It helps the studio identify what players are doing and make inferences, and potentially adapt if they’re seeing a choice being taken too often or not often enough; or even, if the choice is one they want to feel special and earned, keep it that way.

“I call it a very modern way of doing game development, because you finally have the data processing to be able to do these kinds of things,” Vincke says. “And it’s, it improves it, literally, I mean, there’s nothing to be done about it. It makes the games better. They just need time, they need to cook.”

He advised to “try and keep some surprises for yourself,” because things are still changing. Baldur’s Gate 3 is still changing, and there are aspects of this game that might be different in some way I could overlook, thinking I had seen them already in the Early Access. The worst that could happen is that when Baldur’s Gate 3 finally launches, I’ll skip through a scene or not try out a choice because I think I’ve already seen and know it. That’s not what Larian is planning on.

“You won’t know it,” Vincke says. “It will be different, and so that’s the promise.”

Is there any indication when this game might be ready for release?
Already feels like a bit of a has-been blip in the rear view mirror…