Larian Studios Forums Baldur's Gate III Baldur's Gate III - Suggestions & Feedback Character creation ability scors

Yes, Rag... there's a fair chance to get above the expected average. A roughly 50% chance, in fact. So, three of your sets were at or below average expectations, and three were above - not seeing nay argument against rolling so far.

Back to the time question then: since you've said about 6 rolls in 30 seconds (given that you need the time to roll, cognitively process what you're seeing, confirm it doesn't have what you want, and then re-roll), let's be generous and make the maths simpler and say 5 rolls in 20 seconds. Thus, 15 per minute. So... to roll the expected average of 10000 rolls you'd need to reasonably expect a roll of three 18s, you'd need to spend a little over 11 hours doing it, with no breaks, no stops, no slowdowns and no rest. If someone wants to do that for a benefit that will be neutralised by eighth level, I can't really see any complaint there. It's silly, and self-destructive, but it's also their call.


Point stands that your claim of making a character with three 20s at level 4 off a three 18 roll set is false. There is no 'easy' way to do that, and I'm not certain there is Any way to do that legally.

Again, I'll ask a second time: Please enlighten me how you're getting that.



No, you gave a disingenuous account of someone with three 20s at level 4. As well as being, to my knowledge, impossible, the three 18 case that it's built off is one I've already discarded as irrelevant to consideration due to its odds being more than one in ten thousand. We're discussing the practical and common reality of a 4d6 drop lowest providing a player with one 18, which will generally happen about once every ten roll sets. In point to that, I explained quite directly the ways in which the supposedly over-powered place of rolling was not so when compared to another character with the same play-values using ability scores reflected by another system.

Bounded accuracy is precisely WHY the difference disappears very quickly and becomes a non-issue.

In other older systems, there was no bounding - your 20 start became mid-twenties, became thirties, became the chase for a 50 stat by the end of a campaign, and it was ridiculous. A head start *Stayed* with you all game in those systems, and you *Always* had and kept that leg up, all the way through.

In 5e, that doesn't happen. It doesn't happen, because, as stated, multiple times now... even if you start with an 18 roll that you make into a 20 at level 1, that's it: you're at cap... and in just a very small handful of quick levels time, your colleague who didn't roll so well, or who used a fixed system, will *Also* be at cap for their important stat, and the edge is neutralised. It doesn't go any higher, and your initial leg-up edge is gone.


No, I've done no such thing. My contention is as it has been from the outset here: The claim that rolling for stats stands to make unbalanced and unfair characters compared to players that didn't roll is utterly ridiculous, especially so since the edge you might gain from a good roll set disappears by fourth level.

We're talking about roll sets, and a decent portion of the discussion has revolved around comparisons to other players using different methods, because most folks here seem to agree that quibbling over it in terms of single player is pretty pointless. I used the word 'table' by habit, and for that I'll apologise, but it does not change any element of what I said; it's a group of friends playing a game together.

Edit:

Thanks to ArvGuy for producing that roll sheet - I actually went looking for a generator I could use online to quickly generate a 10k set of 4d6 drop lowest rolls and couldn't fine one... I'm not as versed in whipping such things up in excel as I probably should be. I appreciate you taking the time to do it! That 100 sum is a pretty impressive roll. Mostly 17s, so that's begging for a standard human race choice ^.^

I feel like the numbers you present support both sides to the debate in different ways; we can see from the generation that the likelihood of getting an extremely high stat line across the board is not good; it's not a likely occurrence at all, although it's slightly more likely than getting a full set of terrible rolls. The idea of it coughing up multiple 18s and creating characters capping multiple scores from the outset or very early on is also a pretty unlikely, nigh on foolish worry. At the same time, it does also demonstrate the greater swing potential of rolling, compared to the more rigid systems - and that swing is a consideration in games (if it wasn't, this debate wouldn't happen). Thanks for putting it together. I wouldn't mind seeing the full set, if you've got it to hand (and the excel algebra you used to generate it...)

They're pretty good rolls, yes. But none of them are OP. Taking a +2/+1 stat increase, only two of them allow two 18s. Only four of them even allow a single 18, and one of the rolls doesn't even allow two 16s.

Above average statistics are good. Most people don't want to play underpowered characters, and higher stats allows for more freedom to choose cool feats! Especially in BG3, where the max level will be ~12 meaning most classes only get 3 ASIs, having a single stat start out at 18 instead of 16 will make players feel so much better about sacrificing an ASI to take a feat.

If people want to spend 10+ minutes repeatedly rolling for amazing stats, let em. It'd probably be more time-efficient to install a mod that allows you to choose stats at that point though...

One player running at super high stats has no impact on the person that prefers to roll lower stats. In multiplayer it might have some impact but everybody involved will have the same option.

I know my GMs on the tabletop prefer people to have high stats because they prefer to run for stronger characters and also like to encourage taking feats because they consider them more full of personality than raw stats.

Well, i would say "up to" ... since it really can be any of those 10.000 rolls, right?
But that is my point, there will be such people and Larian must expect them too.

Once again ... exactly my point.

Its only matter of amount of frustration you involve in your system ...
I dont have data, Larian do ... but since there are quite hard double-check (or even triple-check) rolls in game ... as it was in case of book Necromancy of They ... Larian should see how often people reloaded just to get past all check ... and THAT are people who will spend half hour, hour, maybe even longer re-rolling their stats to get "as optimal result as possible".
So the only question in my opinion here is: Do Larian aim to please hardcore players, who accepts any outcome even if that was not exactly what they expected ... or will they also take under concideration people who play a little dirty, and allow us infinite stats-buy? :P

//Edit:
I have found another dice roller, to speed up whole proces:
https://rgbstudios.org/dnd-dice/char?r=

I just tryed to "roll" 50 characters ... it took me around 5 minutes.
And maybe im just child of luck ... but i had there maybe two, or three bad rolls ... and around 12 really great (and yes, there was even one that had two 18 and no score was under 12 )

@ Niara

Yeah, that 100 roll is crazy. It's the kind of ridiculous statline that I'm sure a DM would play around with and gently rib the player for having, spending the odd moment to describe how monsters and pawns alike are "in awe" or maybe even slightly infuriated of how chiseled and perfect the person is looking, whenever it can be fitted in. Maybe even making it a minor theme at some point that Mr Perfect is having a bad hair day and a runny nose.

I actually went back and peaked over the numbers a bit more and there are zero cases of three 18's. 935 rolls had a single 18, 29 had two. Chances of getting a character with actually three 18's and having those be in the right abilities and not ending up with ridiculous strenght values is horrible.

That being said, the chance to get a truly catastrophic roll is somewhat lower since just one roll of 5+ will essentially guarantee somewhat playable stats of at least 7 or more. Not getting a single 5+ with four dice is freaky. But of course it can happen and with strict rules you would be out of luck. You can have your 18 charisma warlock with 16 int and 17 con, but your dex is 3 and your strength is 4. That might be great fun at a table, it could lead to personal quests for a girdle of not being a weakling and maybe some elven gloves of not falling on your backside all the time. But would really anyone play that character in a video game?

However, if we consider the relaxation that Bioware decided on back in the day then it becomes very clear why it was so relatively easy to roll stat monsters in the Baldur's Gate games. If we can trade points between abilities after the roll then all that matters is the sum of dice. 297 out of 10000 rolls had a sum of 87 or more. That's roughly 3%, meaning that at 15 per minute, you'd almost certainly get a very good roll within 5-10 minutes.

Lastly, regarding cooking it up in Excel, it was indeed a bit finicky. I started out with a formula based approach, just put a randbetween(1, 6) in 24 columns, then using sum() of column 1 to 4 minus min() of column 1 to 4. And then 5-8, 9-12, 13-16, and so on. It worked but it was slow. Too many formulas needing calculation and something about formulas just rub me the wrong way. So I gave my lazy bone a swift kicking and wrote a macro instead. It isn't super optimized, but it gets the job done.

Well, assume you roll 3 18s. Now take Half-elf. But +2 in CHA (+18 = 20 on level 1) put you free ASIs in CON and DEX or STR (2 19s at level 1). With you ASI at level 4 put 1 point in the other 19 stats. There you have a very very powerful level 4 Paladin/Bard/Sorcerer/Warlock.

You see it is not hard, and doable with everything we have in the game right now. If they implement Tashas rules you can make the +2 ASI also whatever you want and can have any 3 abilities at 20 at level 4.

I'd like them to at least experiment with handing out more feat picks for this reason. 5e is way too stingy with fun level up choices.

EA already has a lot of level ups where you can't choose anything but just click "ok". Meh. Having to choose between +2 to your main attribute and a feat is not a fun choice, and a pretty obvious one at that. This is a big downgrade from 3e especially in a video game where character building is more of a focus than in tabletop. Single player without social interaction and unlimited creativity and all that..I would replay BG3 for character builds instead but the building is very barebones.

If you can roll three 18's, mind you. I couldn't do that in 10,000 rolls.

That was not the premise tho. So the point still stands. Also: a single 18 is pretty damn great, because with the standard-array/point-buy the best value you can assign is a 15. Add to that a +2 from race and you have nice fat 20 on level 1. That is you mainstat being done. No more opportunity costs from that. If you have a SAD character like a Rogue, you have a pretty huge edge.

Rolling dice is better than standard array in around 56% of the cases. So you don't need to roll too much just to get an advantage. There is a reason why things like the AL don't allow for stats rolling.

LoL I am guilty of rolling to get a certain stat set. I have a few character archetypes (involved multi classing, feat combos etc)I have played for almost 30 years (yes late 40’s) in crpgs. There just wasn’t a “class” type or kit that fit as set up in a crpg. In table top, the DM rules, but did let us home brew as long as it was balanced. I have other “types” I play that don’t need this.

With that being said - I prefer the way the OG BG series did it. It IS easier in 5e to get closer to what I want because it doesn’t have all the ridiculous stat and alignment requirements the older editions had. The closest to the archetypes that I like are:

Paladin oath of the ancients (Fighter/Druid archetype)

Paladin oath breaker (Anti-Paladin God Hater archetype)

Shadow Druid atm doesn’t even have a kit/sub class type that I am aware of. BUT if I can actually be recruited to their side AFTER, agreeing to help lock down the grove that would be cool. The game mechanic that can open this up could be to know Khaga’s plans and affiliation, WITHOUT exposing the Shadow Druids… Then as a Druid, you can get an official invite or be put on a hit list if you decline.

These aren’t even guaranteed to even be in the game… So, for me it wasn’t about Min/Maxing but getting a certain “feel” in the games.

Spore Druid is probably closest to the the idea of a Shadow Druid to be honest. But really, most subclasses are alignment independent. Vengeance, for example, can cover most of the alignment grid.

The spore Druids abilities to some extent, but not the philosophy… “The land has been spoiled by civilization. A reset is needed, by force if necessary”.

Being a Paladin of Myrkul with the Oath of Vengeance could come close to what views my Anti-Paladin believed, “The only god I respect is the god of death. Once I am done with the others, he will be next”. Odd, sounds like the PS game God of War, honest though, I had this guy before Kratos lol.

Not all my characters are this dark, it’s just that these are the hardest to make in a CRPG.