On a 1d100? On 10d10? Or as a total for your stats on character creation?
If you mean the last one, then it depends on the class/race combination you are going with. Classes and races influence minimum rolls, thus making it easier or harder to roll a given number, depending.
Total stats for character creation. Just started BGII today and re-rolled an Elvish Ranger on account of not knowing how to transfer my BG chars and stats and the fabled 100 roll appeared. A screenshot is in the thumbnail above. May have to double click to see.
Someone calculated this not long back, you might want to try a search. If I remember right, in an hour of constant rolling, you have a probability of hitting it once.
It's fairly rare, so congratulations! Rangers along with Paladins are certainly among the classes most likely to see it, as they have high minimum rolls (being an elf helps, too).
Appreciate it everyone! On another note, does anyone know how to effectively tranfers BG character files, potraits, sounds, and game files to BGII? Tried to drag and drop the files into BGII's folders but the game kept crashing. Literally had to de-install and re-install for it to function properly again.
I had a program set to roll stats for an elven ranger x times and report the highest result. I can't remember what I had to set x to before it started giving results above 100, but it was... a lot. Millions at least. So congratulations, you just won the lottery of BG stat rolls.
@Musigny it probably doesn't account for the minimum roll system in BG
All I see is that sums over 75 (min BG value) are common. I have no clue about how they do it. For instance how many times 3d6 are rolled for each stat ? When do they add offsets and so on ? This being said, for very high values a simple model may give a hint about the order of magnitude. When rolling 18d6 (once) and according to this site, odds to roll a score >= 100 are around 1/650000 1/65000000. In BG odds are higher. How much ? I don't know.
You'd better play the lottery and be a second/third rank winner.
Edit: reading mistake. Those odds are far from what the nice guys report on this forum. See comments hereafter.
Appreciate it everyone! On another note, does anyone know how to effectively tranfers BG character files, potraits, sounds, and game files to BGII? Tried to drag and drop the files into BGII's folders but the game kept crashing. Literally had to de-install and re-install for it to function properly again.
Portraits are just pics, dumped in the appropriate portrait folder. It shouldn't cause any crashes so a reinstall was probably a good idea there as likely there was another cause. Char files are designed to be able to be imported in BG2, so also again, simply putting them in the folder should have worked.
Appreciate it everyone! On another note, does anyone know how to effectively tranfers BG character files, potraits, sounds, and game files to BGII? Tried to drag and drop the files into BGII's folders but the game kept crashing. Literally had to de-install and re-install for it to function properly again.
You can/should also be able to start a game by importing your last save game from BG1. That will transfer the character for you.
@Musigny it probably doesn't account for the minimum roll system in BG
All I see is that sums over 75 (min BG value) are common. I have no clue about how they do it. For instance how many times 3d6 are rolled for each stat ?
Do we even know for sure that the game rolls 3d6? I know that's how it works in D&D, but based on my experience it seems like it could very easily be just generating a random number between 3 and 18 with a uniform distribution.
@Musigny it probably doesn't account for the minimum roll system in BG
All I see is that sums over 75 (min BG value) are common. I have no clue about how they do it. For instance how many times 3d6 are rolled for each stat ?
Do we even know for sure that the game rolls 3d6? I know that's how it works in D&D, but based on my experience it seems like it could very easily be just generating a random number between 3 and 18 with a uniform distribution.
As I said, I have no idea about their implementation. This could be uniform as you suggest. I can believe it. However if the distribution is uniform then this means that a roll equal or greater than 100 is much more frequent -> around 2/10000. Please forgive me but I am a bit cautious as regards the various claims.
And btw I was wrong with the anydice.com results as they provide % not 0 to1 probability. I will edit my previous message to fix it. Two orders of magnitude that's a lot. This strengthens your own assumption.
Do we even know for sure that the game rolls 3d6? I know that's how it works in D&D, but based on my experience it seems like it could very easily be just generating a random number between 3 and 18 with a uniform distribution.
A while back I did a bunch of experiments on highly controlled sections of the distribution (paladin Charisma scores were highly useful in this regard, since they only have two possible values). First, the system does appear to roll 3d6, or at least uses something closer to that than to a uniform draw from 3-18. We can tell this because the game rolls three times as many 17s for paladin Charisma as 18s, which is what we would expect from 3d6. Second, we can tell from the same results that the game appears to reroll any individual stat that is below the minimum for the race/class in question. Or perhaps it merely approximates a reroll. We know it doesn't just roll 1d2+16 or something.
We can also reasonably assume that no stat total below 75 is ever produced, given that none of us (to my knowledge) have ever seen one lower than that. It's harder to test exactly how this is accomplished, but the distribution suggests that the game simply rerolls any results that total below 75. The distribution above 75 appears to be approximately what one would expect from such a system, although individual stat minimums make this difficult to verify in practice. In addition, the fact that the reroll system is used for individual stats suggests that it was a method the developers approved of, which, combined with the other evidence, makes it the most likely candidate for the implementation of the 75 minimum.
The application of both these minimums in combination makes exact probabilities very difficult to calculate. I've never seen someone incorporate them successfully, and I know for certain it's beyond my skill. Hence the brute force approach I mentioned earlier. If someone thinks they know how to compute the probabilities given the constraints described, I'd love to see the math. This problem has been frustrating me off and on for some time now, and I'd love to see it concretely solved.
Noloir
Member, Mobile Tester Posts: 380
Comments
If you mean the last one, then it depends on the class/race combination you are going with. Classes and races influence minimum rolls, thus making it easier or harder to roll a given number, depending.
If you don't know the model used by BG then it will not help you much.
This being said, for very high values a simple model may give a hint about the order of magnitude.
When rolling 18d6 (once) and according to this site, odds to roll a score >= 100 are around 1/650000 1/65000000. In BG odds are higher. How much ? I don't know.
You'd better play the lottery and be a second/third rank winner.
Edit: reading mistake. Those odds are far from what the nice guys report on this forum.
See comments hereafter.
http://i.imgur.com/UfmQbeU.jpg
And btw I was wrong with the anydice.com results as they provide % not 0 to1 probability. I will edit my previous message to fix it. Two orders of magnitude that's a lot. This strengthens your own assumption.
We can also reasonably assume that no stat total below 75 is ever produced, given that none of us (to my knowledge) have ever seen one lower than that. It's harder to test exactly how this is accomplished, but the distribution suggests that the game simply rerolls any results that total below 75. The distribution above 75 appears to be approximately what one would expect from such a system, although individual stat minimums make this difficult to verify in practice. In addition, the fact that the reroll system is used for individual stats suggests that it was a method the developers approved of, which, combined with the other evidence, makes it the most likely candidate for the implementation of the 75 minimum.
The application of both these minimums in combination makes exact probabilities very difficult to calculate. I've never seen someone incorporate them successfully, and I know for certain it's beyond my skill. Hence the brute force approach I mentioned earlier. If someone thinks they know how to compute the probabilities given the constraints described, I'd love to see the math. This problem has been frustrating me off and on for some time now, and I'd love to see it concretely solved.