I just rolled myself a 100 points with 18/00 strength yesterday (in BG1EE though it should be the same throughout the games). Almost spit my tea on my keyboard when it happened. *total: 100 "oh wow", checks strength > 18/100 "sprfffffff"*.
Technically the maximum is all stats maxed out but you need to get ridiculously lucky for that to happen. The best I've had myself was 102 or 103, though that Kensai roll just now blows that out of the water by having a 1-in-100 chance for that strength.
Assuming I know how this games rolling works (I'd assume it rolls a d18 and if you roll, say 1, the game raises that to it's minimum). So all 18's is (1/18) to the power of 6. That's 0,000000029.
At least here, Lotto is 7 numbers out of 40. If I understand how to calculate that it's (7/40)*(6/39)*(5/38)... and so on which end up with 0,000000054.
Unless I made a mistake in that then yes, it actually is more likely to win in lotto. Almost twice as likely Wow
EDIT: add exceptional strength of 18/00 and you can add a couple of zeroes to that as it adds another 1/100 roll to the mix.
It's actually lower than that if the game is adhering to the 2e rulebook which stipulates that 3 d6 dice are used to roll stats (so rolling 18 becomes 1/216, not just 1/18).
I don't think either of those are the case though. Rolling a paladin for example, with a minimum 17 charisma, you'll have an equal chance of having 17 or 18 charisma.
If you rolled an 18 sided die, and then moved anything too low up to 17, then you'd be getting almost entirely 17's. Same with 3D6.
It would seem that a 2 sided die is being rolled in this situation.
It does roll 3d6, but it discards any results that don't meet the stat requirements and keeps rolling until it gets an acceptable one.
In the case of the paladin the chance of rolling an 18 with 3d6 is 1/216. The chance of rolling a 17 is 3/216. Therefore you will see 3 times as many 17s as 18s for your paladin, i.e. if you rolled 100 times you would see on average 25 18s and 75 17s (as results are random they will vary, but over 100 rolls you would expect to get something not that far from the average).
Jeez, knowing now that it's rolling 3 die and seeing the results of 10 million rolls documented. Just, wow. I never realized high rolls in BG are so harsh.
If you don't mind me asking though, didn't you roll over 100 even once in all those tries or did you just omit it because it's so rare that it gets cut of due to statistical discrepancy (or whatever the correct term for cutting extreme differences out of statistics is)?
Also holy shiiiiieeeet, I'm so going to save that Kensai I got into a cloud now. That was an honest mouse clicking roll and and it looks like it's close to 1 in 500 million! (Which is probably about as much rolls I've done in my lifetime, I'm kinda shamed to admit).
So, here's a question. How does the game handle the 75 minimum during rolling? Does it just go from top to bottom with stats and decide at some point that it can't reach 75 anymore and then reroll stats until it get's at least there? If it rerolls the whole thing instead of applying some sort of "raise to minimun" rules it seems unlikely that 75 would be so common.
I've just checked in my 2e player's handbook and it has this to say about the reasoning behind rolling 3D6:
"This method gives a range of scores from 3 to 18, with most results in the 9 to 12 range. Only a few have high scores (15 and above), so you should treasure these characters."
Which certainly makes me question my obsession with 18s.
I've just checked in my 2e player's handbook and it has this to say about the reasoning behind rolling 3D6:
"This method gives a range of scores from 3 to 18, with most results in the 9 to 12 range. Only a few have high scores (15 and above), so you should treasure these characters."
Which certainly makes me question my obsession with 18s.
Baldur's Gate does seem to be balanced around a character with a high score in their primary stat though.
You can play a mage with 12 intelligence but you won't have a fun time.
You can play a mage with 12 intelligence but you won't have a fun time.
I don't think it makes that much difference actually. The only impact is on spell learning and you can use potions to ensure 100% success for that anyway.
So, here's a question. How does the game handle the 75 minimum during rolling? Does it just go from top to bottom with stats and decide at some point that it can't reach 75 anymore and then reroll stats until it get's at least there? If it rerolls the whole thing instead of applying some sort of "raise to minimun" rules it seems unlikely that 75 would be so common.
I think it does reroll entirely. Have a look at the profile of @lefreut's actual scores - that's consistent with total rerolling rather than a raise to minimum.
Baldur's Gate does not roll 3d6 for each stat. If you go in the game and just hit the reroll button for a while, you'll see that most stats end up being 9+ most of the time, even for fighters.
You might think to yourself that maybe BG rolls 3d6 per stat and then boosts the overall point total to 75 if you didn't naturally make it there. But no, BG doesn't do that either. 18d6 only gets you 63 points on average so 2/3 of characters would be getting the boost, and showing up as 75 point characters, under this system. Yet @lefreut's table shows that only 25% do.
For sure, the game is dropping any character with fewer than 75 points rather than "boosting" subpar characters to 75. I wonder if it's also "stacking the deck" by rolling "loaded" dice for each stat so that you'll average 12-13 per stat rather than 9.
@jsaving there's no need to assume any minimum stat rolls beyond those required by character class / race. As you say the overall minimum of 75 is already well above the average for 18d6 rolls, so it's inevitable that the bulk of stats will be above average. You can check that conclusion just by looking at the stats posted by @lefreut - if you really had minimum scores of 70+ already you wouldn't observe such significant reductions in the probability of scores as they go above 75.
This behavior has also previously been extensively tested empirically on the old Bioware Social Forums to confirm there is no need to assume anything more than the 75 minimum to result in the actual stat scores observed.
Seems reasonable. A quick way to test this might be to track a paladin's charisma, which has to be either 17 or 18, over many rerolls. 17s should show up much more often if fair dice are being rolled.
(Edit: just rolled 100 paladins and got 17 charisma for 70 of them. That's way too high for "loaded" dice but in line with what you'd expect from simple 3d6 die-rolling.)
You can play a mage with 12 intelligence but you won't have a fun time.
I don't think it makes that much difference actually. The only impact is on spell learning and you can use potions to ensure 100% success for that anyway.
Oh? How many such potions exactly do you have as a level 1 mage fresh out of candlekeep?
Sure, by the end of Throne of Bhaal you'll be up to your armpits in potions, but there's a whole lot of game leading up to that.
@Chronicler well there are quite a few on sale at the FAI, which is not exactly a long way from Candlekeep. I accept of course you will need to save scrolls up and only learn them at a few points in the game. Having played like that many times it doesn't make much difference to the way I play the game, but you may feel differently of course .
@Chronicler well there are quite a few on sale at the FAI, which is not exactly a long way from Candlekeep. I accept of course you will need to save scrolls up and only learn them at a few points in the game. Having played like that many times it doesn't make much difference to the way I play the game, but you may feel differently of course .
Yeah, in my books having only a few opportunities to learn spells in a 20+ hour game would be a pretty big impact on the fun.
Comments
Apologies, I must have expressed myself badly. As far as I am aware the maximum possible roll is 18 in everything.
Having said that I have no idea how the roller has been programmed and whether it is truly random.
Perhaps rather my bad (non-native speaker's) understanding of the term "unless" – apologies by me.
Technically the maximum is all stats maxed out but you need to get ridiculously lucky for that to happen. The best I've had myself was 102 or 103, though that Kensai roll just now blows that out of the water by having a 1-in-100 chance for that strength.
At least here, Lotto is 7 numbers out of 40. If I understand how to calculate that it's (7/40)*(6/39)*(5/38)... and so on which end up with 0,000000054.
Unless I made a mistake in that then yes, it actually is more likely to win in lotto. Almost twice as likely
EDIT: add exceptional strength of 18/00 and you can add a couple of zeroes to that as it adds another 1/100 roll to the mix.
If you rolled an 18 sided die, and then moved anything too low up to 17, then you'd be getting almost entirely 17's. Same with 3D6.
It would seem that a 2 sided die is being rolled in this situation.
In the case of the paladin the chance of rolling an 18 with 3d6 is 1/216. The chance of rolling a 17 is 3/216. Therefore you will see 3 times as many 17s as 18s for your paladin, i.e. if you rolled 100 times you would see on average 25 18s and 75 17s (as results are random they will vary, but over 100 rolls you would expect to get something not that far from the average).
If you don't mind me asking though, didn't you roll over 100 even once in all those tries or did you just omit it because it's so rare that it gets cut of due to statistical discrepancy (or whatever the correct term for cutting extreme differences out of statistics is)?
Also holy shiiiiieeeet, I'm so going to save that Kensai I got into a cloud now. That was an honest mouse clicking roll and and it looks like it's close to 1 in 500 million! (Which is probably about as much rolls I've done in my lifetime, I'm kinda shamed to admit).
"This method gives a range of scores from 3 to 18, with most results in the 9 to 12 range. Only a few have high scores (15 and above), so you should treasure these characters."
Which certainly makes me question my obsession with 18s.
You can play a mage with 12 intelligence but you won't have a fun time.
You might think to yourself that maybe BG rolls 3d6 per stat and then boosts the overall point total to 75 if you didn't naturally make it there. But no, BG doesn't do that either. 18d6 only gets you 63 points on average so 2/3 of characters would be getting the boost, and showing up as 75 point characters, under this system. Yet @lefreut's table shows that only 25% do.
For sure, the game is dropping any character with fewer than 75 points rather than "boosting" subpar characters to 75. I wonder if it's also "stacking the deck" by rolling "loaded" dice for each stat so that you'll average 12-13 per stat rather than 9.
This behavior has also previously been extensively tested empirically on the old Bioware Social Forums to confirm there is no need to assume anything more than the 75 minimum to result in the actual stat scores observed.
(Edit: just rolled 100 paladins and got 17 charisma for 70 of them. That's way too high for "loaded" dice but in line with what you'd expect from simple 3d6 die-rolling.)
Sure, by the end of Throne of Bhaal you'll be up to your armpits in potions, but there's a whole lot of game leading up to that.