Larian Studios Forums Baldur's Gate III Baldur's Gate III - Technical & Gameplay Problems Possibly bugged attack rolls?

My attack rolls seem way off, statistically speaking. I've gone through the blighted village and swamp today; played a few hours and I didn't record a single attack roll over 10 on my PC.
In total I'd say I rolled around 70-100 times.

My skill checks and saves where not affected this way, only attack rolls. The rest of the party and enemies rolled seemingly normal numbers across the board though.
Ofc it is strictly speaking possible if very unlikely but this feels like a bug. Has anyone else encountered this?

Yes. I wrote about it in another thread. I eventually quit that character because it was close to unbearable.
A prolonged streak of bad luck like this is completely possible of course but seeing that you are not physically rolling dice and just have to "believe" the RNG, I still think there should be some kind of failsafe built in that defends against rolling single digits 50 times in a row.

So I did some more testing and I did 50 attack rolls with a new character and the calculated avg. was 6.04 but this time there were the occasional >10 roll, so I guess there isn't that hard limit I thought there was when I was testing before.


ATTACK ROLLS:
3+7+6+4+3+8+3+8+6+5+3+5+2+9+8+6+8+7+5+1+17+4+8+7+1+2+9+8+1+1+5+6+4+8+15+6+7+5+2+6+17+19+4+10+1+4+1+3+5+9 = 302/50 = 6.04.
To clarify, these are the primary rolls, i.e. not the adjusted values after bonuses and penalties.
This looks like some sort of bug considering I made a new character and everything.
I'll keep testing and see how this develops.

I appreciate the testing.
Doesn't look amazingly bad but it's showing already that single digit rolls seem to be a way more common outcome than double digits, for no discernable reason.
Would appreciate some more data on this.

Yeah, while I haven't been very scientific about it, it certainly seems that there's a very high ratio of low rolls. Almost as if disadvantage is applied all the time.

Yeah I'm no statistician and statistics are famously difficult and counterintuitive but the way these rolls play out makes it seem there is some implicit bias somewhere in the code.
These numbers I get are not produced in a controlled environment, they're a result/observation of gameplay so I can't say anything about how or why I get these particular results.

Haven't time to do anymore testing yet, sry.

did you have disadvantage or advantage on any of these rolls? that would throw these numbers off.

Now that you mention it, I hadn't even thought about it. Considering it was in regular gameplay it is likely that I had both advatage and disadvantage. This is a good point, I'll make sure I record that as well when I get to it.

I just did the same, saved at the start of the first (turotial) fight and took a look. Sample size and all, but this was mine over roughly 50 (all no advantage / disadvantage rolls which force you to either take the higher or lower of two rolls which is then displayed -- e.g. I purely melee attacked the opponents from where they could see me):

11 12 9 17 7 11 20 15 7 6 15 9 7 7 19 12 7 15 10 8 19 19 19 18 18 17 19 6 4 14 4 15 1 20 3 6 19 17 18 12 10 3

Average 12.02 (in the long run this should come closer to the ten, naturally)

There could be bugs, but from all of my playing experience too none that obvious. Even considering sample size, yours suspiciously look as if you may have been put at disadvantage a decent amount of time.

So I've done some more testing without avd/disavd. and then I got an avg of 9.74 on 250 rolls which closer to what I'd expect to see.
Kinda embarrassing to start a thread like this and then have it turn out to just be me being crap at the game haha

This actually looks worse than just attacking with disadvantage, the latter would give you 7.175 on average. However, the sample size might be too small for any reliable assessment, I suspect.

Personally, I also have some questions regarding the game's RNG. For example, a suspiciously high percentage of Crits landed by my party happens with targets having 5 or less HP remaining, which makes them sort of redundant. Scoring a Crit on an unharmed enemy, however, even with advantage? Good luck with that. So far, I've seen only 2 such outcomes, one per playthrough, when party's rogue one-shot a phase spider with sneak attack from shadows while on higher ground.

On normal rolls, the long-term, average should be 10.5.
With advantage, it should be ~13.9.
Without, ~7.1

https://www.r-bloggers.com/2020/06/dungeons-and-dragons-advantage/

One time when rolling 100, I had an average of normal rolls of over 11, however that should even out.
https://forums.larian.com/ubbthreads.php?ubb=showflat&Number=709062#Post709062

If anybody's interested and records stuff, you could also post your findings in that thread (there's now over 2,000 rolls recorded inside).

I also wanted to write about it, the game is a "miss fest". I was not blaming statistics though I think the hit chance dice should be adjusted, I thought it was too high for not being able to hit anything. But it goes for the enemies as well, so fights take ages.