Larian Studios Forums Baldur's Gate III Baldur's Gate III - Technical & Gameplay Problems What a lying, lying game!

Windmill fight, tired of making them run away.

Lae'zel has 3 attacks against her
19
19
18

Other 3 are around the corner, hiding in smoke. 18 arrow on tav

Back to Lae'zel, another 19 on her


HAS ANYBODY LOGGED ANY OF THIS STUFF TO PROVE OUT A HIDDEN ANTI-KARMIC, FRONT-LOADED DICING. THIS IS NOT A D20 GAME.

IT IS NOT A BONA FIDE DICE GAME IF THE ENGINE CHEATS AND FRONT LOADS HIGH ROLLS FOR THE GOBLINS AND LOW ROLLS FOR YOU, AND THEN QUICKLY FIXES UP THE STATS TOWARDS THE END OF THE FIGHT BY GIVING YOU HIGH ROLLS. THIS IS NOT A BONA FIDE DICE GAME.

So pissed, went to murder the lovers.

Interrupt their fun
Bugbear Crit on TAV, downed
Bugbear then runs next to Gale, then runs away, Gale gets a successful opportunity attack
Gale uses off hand attack (he's holding two torches), critical fail
Gale runs away
Bugbear rolls 5 = 11 = 11, enough to hit Gale, how convenient barely enough. Gale does not go down, barely, Gale throws pot on TAV
Shadowheart void bulbs the two back into the barn
Lae'zel booms twice (action surge) and pommel strike, 11, 9, 13, enough to hit Ogre 3x
Bugbear rolls 18, hits Lae'zel
Shadowheart's spirit trident rolls 10, enough to hit
Gale uses grease, bugbear goes down
Shadowheart rolls 16 on fire bolt on ogre, hit, grease goes up
Lae'zel booming blades the barn door for some reason
Ogre hits spiritual weapon, still alive
Bugbear rolls a 3, misses
Tav MMs ogre for second time
Gale misses ogre with fire bolt, 6
Shadowheart rolls 19 on Ogre with firebolt, rolls a 1 for damage.
Lae'zel rolls a 2 for booming blade on Ogre
Ogre rolls 2, misses, tenacity gets Lae'zel for 5 anyway
Bugbear throws something, rolls 3
Gale fire bolt rolls 3
Shadowheart FBs 11, enough to hit ogre, it goes down
Lae'zel's booming blade rolls a 2
Tav skips this round because idiot thumbs spacebar by mistake
Spiritual weapon rolls 2 on bugbear
Bugbear rolls 18 on Lae'zel, but she shields
Tav MMs bugbear
Gale FBs for 6, a miss
Shadowheart 16 FB, a hit
Lae'zel booming blade 7, a miss
Spiritual weapon 3s on bugbear
Bugbear 3s on Lae'zel but 2s from tenacity anyway
Tav uses last MM and wins the day.

I got TPK'd in the Kargha combat. A combat I never lost in the past. Tav (bard), Gale, Astarion and Lae'zel, all at lvl 4.

Simply missed everything. And so did Silver and Rath. Silver was also killed, only "good" character still standing was Rath, when my last partymember fell.

Oh yes and Astarion and Tav have "favourable beginnings" tadpole power and Astarion was "happy" (+1 to attack rolls.)

Obviously these things are technically possible, and with millions of playthroughs across what, hundreds of thousands of copies, there will be some people who get insanely bad sequences.

What I am saying is these are way too frequent, especially early in combat, where the game may front load good rolls to them, bad rolls to you, then fix up the overall statistics later in the combat, boosting you when it little matters anymore, or worse, in later rolls after moving on.

I'd like to see a combat logger, but it should record all rolls, and persist across save scumming restores, along with character, enemy or ally, including the attack bonuses and penalties breakdown, as if hovering over a "hit" or "miss" or damage roll line in the combat log.

I can even envision a little live window, maybe one of those election night live dials, showing statistical probabilities of how well each of your characters and the opponents are doing, individually and as teams.

It would take lots of rolls to separate this out from any individual battle where things could go wrong.

Or maybe this has already been done. Not as a mod I can find, but I did read somewhere someone logged about a thousand rolls during fights and found enemies averaging 15 and party only 10. This would be pre-mod raw rolls. My bleats above are this, and do not count tactician bonuses and whatnot.

There are enough gods against us already. One more RNG god literally isn't fair!

Isn't there a country song, "Your Cheatin' Dice"?


In Balthazar's underground base, trying to unlock the 30 difficulty door.

These are the raw rolls, with advantage!

5
11
13
11
11
8
8
14
15
15
1
18
9
17
12
15
18
10
Crit
My bonuses were +9 to +13

I literally have 30 lockpicks in my group, and 50 more back at the base. That's not the point. Come on, man.

That's 18 failures of advantage, the equivalent of 36 rolls. .95^^36 = 15%, not not impossible, but unlikely. Keep in mind rolls of 17 or above would have to fail the 1d4 guidance on top of that. There were three such rolls.

As a side note, there's a meme on the Internet that the game will yank thieves' tools from the chest so you don't have to carry them around. Nope.

Do you have Karmic dice turned ON or OFF?

This mostly just sounds like normal RNG honestly. Humans are really bad at intuitively judging randomness, especially when we remember the bad streaks much more clearly than the average ones.

In long sessions the rolls usually even out pretty close to expected probabilities.

Statistical spikes are always memorable. On tabletop many players have their lucky dice others aren't allowed to touch, almost superstitiously using some dice for attacks, others for saves and yet others for skills. One Star Wars D6-session we played in the mid-2000s was especially memorable. As the GM, I was rolling for both the gunners of a pirate transport with the player characters aboard, one also manning a turbolaser and their Imperial adversaries. They were intercepted and chased by TIE Bombers. When we rolled combat, all the pirate NPCs and player characters rolled normally, but when it was my turn to roll for the Imperial pilots, the Wild Dice rolled straights of 1s, read: One critical failure after the other. This pattern was stable not for one round, but for three consecutive combat rounds even when I switched the sequence. This resulted in the TIE Bombers annihilating each other or themselves. The in-game explanation was that there must have been sabotage by Rebels, but at the table we were like, what the Force was that? In BG3, I remember both some memorable 20s at the Waning Moon accompanied by weird crashes I couldn't reproduce later and Astarion having really bad luck with that mined intersection in Act 1 near the Goblin Camp. "Hey, should we go back to camp and ask Volo?" Other than that, I think the dice behaved pretty normally if you kept Karmic Dice unchecked.

Go to fight the mephits south of Teahouse.

I have 19 AC

Them:
Whomp
Whomp
others still on the way
Me: Whiff
Them:
Crit
Crit

Fight Kagha, me:
WIff
Lae "wiffy" zel
Wiff

After I finally approach the end fight, should I use feinting attack to gain advantage? Hypothis: No, because the cheating dice engine is now attempting to repair the statistical BS frontloading by adding extra to the rolls towards the end.

Normal roll: SMASH
Normal roll: SMASH
Fight over. Hypothesis made, theory demonstrated.



Go into Zhentarim hideout
4. [Persuasion] I just want to trade, use guidance
1 reload
2 reload
1 reload
6 reload
9 reload (10 is needed)
7 reload
20 critical
This is a joke. A monster sequence of bad rolls, followed by an attempt to fix up the overall statistics with a crit.


Is there no mod to record rolls so we can feed it into a statistical analysis to detect frontloaded BS rolls that are later "fixed up" towards the end of a fight?


These are not one offs! This is not cherry picking bad roll sequences! This is regular operation! <------------- Is anyone paying attention?

Technically, this isn't what is happening, as reloading effectively removes the roll from the game's memory.

Thus, every save scummed roll is being done under the same logic.

Which can lead to horrible results when factoring in buffered statistic logic - I.e. If your roll happens to be one with a large penalty to it due to a string of good rolls earlier, then all save scummed rolls will have this large penalty as the logic won't update.

As such, if the case is that buffered statistic logic is being used, then that crit was just a lucky crit despite a large penalty on the actual roll. Not a "You rolled a bunch of bad rolls, here's a pity crit" scenario.

Which is where the benefit of Inspirations comes in. As they don't remove the bad roll like save scumming, but do another one instead, it allows the logic to update with the fact you got a bad roll and then improve your chances.

As such, you might be interested in the No Inspiration Cost mod to utilize in lieu of save scumming, letting you spam inspiration rerolls for free instead of reloading a save to do repeated rolls (Which would provide better statistics on overall roll logic than save scumming results would)


There's the Records mod that will allow you to track rolls, though I'm not sure if it will allow you to see a dynamic array of stats (I.e. The proposed scenario of bad rolls being buffered later on for a statitical average). But will let you track overall hits/misses on your dice, both inside and outside of combat.