BG3 in a nutshell: 95% chance to hit... MISS. MISS. CRITICAL MISS. MISS. 2 DAMAGE.

[Justin_760]

95% chance to... CRITICAL MISS.
85% chance t... MISS
90% chan... MISSSS

(skill checks)

You have +5 proficiency in persuasion and maxed out Charisma.
Roll a 5 or above.
Rolling...
YOU ROLL A 4.

Intimidation check, you have +3 proficiency.
Roll a 7 or abov...
YOU ROLL A 6
Reroll?
YOU ROLL A 1.


I actually decided to test a skill check by reloading saves. Roll a 14 or above to convince Khaga with persuasion. So 14-20 is 7 possible numbers for success on a 20-sided die. Thats roughly a 33% chance.

It took 14 reloads to finally roll above 14. That is about 7% success chance, which is significantly lower than 33%. I'm not saying it is broken RNG, I don't know enough about the games system to say for sure. But it is definitely not behaving as you would expect.

Inspecting the little Tadpole container in the very beginning of the game requires investigation - I had to roll a 5 or above to pass the skill check. That is about 80% success chance, but again, it took 7 reloads to finally pass the skill check. That is a 14% success rate, which is far lower than 80%.

Perhaps reloading saves (scumming) alters the RNG system somehow, which could account for these discrepancies. I'm not sure, however. Something about the RNG skill checks really doesn't add up.

[Ashe]

This thread reminds me of that guy who worked on Civ talking about how gamers view chance. As though if it's above 50% it's a solid choice.

You got the 5%. If the dice don't love you that's just how it be. I play a Warlock who's constantly failing CHA checks. It be like that sometimes.

edit : found it

[Sly_Boots]

Regarding the point about resurrecting dead characters, there's a character you can unlock fairly on who appears in your camp and offers to res people for 200g.

Not wanting to give away spoilers, but it's worth thoroughly exploring the temple you find Shadowheart trying to get into. :)

[djinnxy]

I don't have these problems. Yes RNGesus loves me. The stat window tells me so.

[Space Dog]

The d20 system has pretty big variance built in it.

A completely untrained character with 8 in the associated attribute (-1) only has 30% less chance of succeeding than a proficient character with 16 in the attribute (+5).

This could mean 50% vs. 80% chance of success (DC10). Let's say it's a Dex check to catch an object thrown at you. The worst clumsy least skilled character might easily succeed while the highly trained agile expert fails.

That difference between minimum skill and maximum skill is too small and the biggest weakness of the d20 system.

I'd like to see the skill ranks have more impact in these checks. Or more Take10 checks that work with tresholds.. you either can or can't do something instead of a highly random roll.

[dulany67]

People who can't handle rng should not buy rng heavy games.

[Nikanuur]

引用自 Slapstick：

Have none of you played XCom?

Did you mean, had he played that game that made the button for quickload worn after one gaming session :D ?

[Tensu]

I do think some % are buggy or not taking everything into account so its good to click the combat log where you can hover over every action to see what was rolled and what modifiers were taken into account.

I noticed my clerics sacred flame no matter who I target or how much dex they have is always a 50% hit chance when in reality it hits maybe 10% of the time.

[Romeo.34]

XCOM 2 flashbacks. 99% chance to hit, MISS, RELOAD, MISS, RELOAD, MISS, RELOAD MISS. ALT+F4

[Justin_760]

引用自 DestroShade：

yall just getting some ♥♥♥♥ rolls sometimes in dnd that happens

You're ignoring the statistical anomalies in skill checks. Sure, in combat there are good and bad rolls, thats how it works.

But a 50% chance to succeed at a skill check should not take 7 attempts. That is beyond an anomaly. Flip a quarter and guess which side before it lands. See how long it takes you to call it wrong 7 times in a row. It is an anomaly. Sure, anomalies happen, but they are rare. In Baulders Gate, they are not the exception, they are the rule. For my playthroughs at least.

Naturally, video games emulate probability using computerized RNG systems that are (often) crude representations of real-life laws of probability.

[Justin_760]

引用自 Arcanum Elite：

I have failed litterally every skill check ive had when i played.... except one with the dying mindflayer... which just ended up getting me killed as i failed the second wisdom check to break free of its control...

And this is with alot of it having bonuses to the skill checks... I once needed to roll a 3... and i rolled a 1.... so I must have the absolute worst luck in the world.


speaking of which....


is there anything i can do if I have dead characters i cant revive in my party... i cant even rest as long as they are dead and with no way to revive them that means i can never replenish my spells... ever... so can anything be done in this situation or am i just screwed?

The zombie dude you find during the "Explore the ruins" quest will appear in your camp and offer to resurrect party members for 200 gold each. He is at the end of the quest in a sarcophagus.

Other than scrolls, Gale has a special dialogue that triggers when he dies with some comically complex method to revive him without a scroll. As far as I know, it can only be used once.

[Lady An'wen]

pretty much XCOM level of misses are what I keep seeing.

[NamThunder]

Calm the heck down. But still, that's RNG and dice rolls for you.

But considered that there have yet to be the option to choose and change game difficulty just yet, I would just be patient and see. When they introduced difficulty levels, we might just see the modifiers they might put into this randomness for each level.

[Snobby Hobo]

引用自 pandariuskairos：

引用自 Justin_760：

You're ignoring the statistical anomalies in skill checks. Sure, in combat there are good and bad rolls, thats how it works.

But a 50% chance to succeed at a skill check should not take 7 attempts. That is beyond an anomaly. Flip a quarter and guess which side before it lands. See how long it takes you to call it wrong 7 times in a row. It is an anomaly. Sure, anomalies happen, but they are rare. In Baulders Gate, they are not the exception, they are the rule. For my playthroughs at least.

Naturally, video games emulate probability using computerized RNG systems that are (often) crude representations of real-life laws of probability.

So what happens is that using an RNG can actually cause it to become 'weighted' over time, so that it successively reproduces the same numbers continually. This may be (I don't actually know what Larian is doing, but something is strange about their RNGesus system) what is occurring. Digital RNG isn't truly random, and particularly not after the first roll.

One way around this is to generate from scratch a new random number generator ever time. In fact, they could have the game call up a random number generator from the internet, or on their servers, that essentially creates a "fresh copy" of the generator each time a die roll is called for. This would be the safest way to ensure that dice rolls aren't, over time, becoming increasingly biased. I wonder if Larian isn't doing something like this and thus some people's rolls just get worse (or don't change) over time. Like, you might be setting the RNG essentially at the beginning of a game with a couple of bad rolls and then it just keeps rolling bad for that entire playthrough.

A seed is nothing but an offset, if that is what you're referring to.

I think you have the right idea but you worded it poorly. If you put RNG in every micro decision the combined outcome will always gravitate towards the average value you'd expect from a fair die. When throwing a fair die with 6 sides 3 times the average may seem arbitrary, but when averaging 1000 throws it'll always be roughly the same. It's a solid point to make and a very logical one too.

Sadly there is already strong evidence of humans not being rational in the slightest. People are much more inclined to remember when they were done wrong than good. As such people will remember "that one time I missed three arrows in a row!" - but are unlikely to recall the 3 crits.

[Justin_760]

引用自 Zacharia：

引用自 ⚜：

all you guys complaining about rng need to play games like battle brothers and darkest dungeon.

Battle Brother- brother here. I confirm the above.

At least with Battle Brothers the %'s are cut and dry. Most bro's have poor accuracy, but that is reflected in the % chance to hit.

BG3 will say 80%+ but it will really be ~50%. Battle Brothers will just straight up tell you that your bro has a 50% chance to hit BEFORE you attack.

[Zouls]

1) people are bad at math, will see 60% chance and think "this will hit a bit over every second time" which isnt guaranteed

2) this is compunded by the fact that most games cheats in your favour, so an 85% is closer to 95% in most games