What classes give the highest overall stat rolls
JDow
Member Posts: 71
Hi guys,
Thought I'd share this here.
https://youtu.be/DYA55ILfPFc
It's a 5 minute video on what classes give the highest stat rolls.
Paladins sure, but the take away is that specialist mages (not conjurers sorry) roll high, and rangers roll low!
And yes when it says shadow masters it means shadows dancers
.
Do let me know what you think, or if you disagree.
********** Edit: Due to the comments I've realised my results are wrong. I worked on the basis that rolls are made, if they're below the minimum values for the race/class they're raised to those minimum values, and then the whole result is rejected if it's below 75. However, that's not the case because if you keep clicking re-roll for a paladin and watch the charisma score, rather than the rolls being 15:1 in favour of a 17 (as opposed to 18) there's way too many 18s come up. So, it's looking like the game doesn't raise the roll to the guaranteed minimums but rather keeps rolling until the roll itself is over the minimums. This drastically changes things. So I'll go back to the drawing board!!! *********
Thought I'd share this here.
https://youtu.be/DYA55ILfPFcIt's a 5 minute video on what classes give the highest stat rolls.
Paladins sure, but the take away is that specialist mages (not conjurers sorry) roll high, and rangers roll low!
And yes when it says shadow masters it means shadows dancers
.Do let me know what you think, or if you disagree.
********** Edit: Due to the comments I've realised my results are wrong. I worked on the basis that rolls are made, if they're below the minimum values for the race/class they're raised to those minimum values, and then the whole result is rejected if it's below 75. However, that's not the case because if you keep clicking re-roll for a paladin and watch the charisma score, rather than the rolls being 15:1 in favour of a 17 (as opposed to 18) there's way too many 18s come up. So, it's looking like the game doesn't raise the roll to the guaranteed minimums but rather keeps rolling until the roll itself is over the minimums. This drastically changes things. So I'll go back to the drawing board!!!
*********Post edited by JDow on
2
Comments
- why don't all classes have an equal chance of getting a 108 total (I'm assuming that you haven't included racial adjustments in your figures)?
- how the basic fighter (with the lowest minimum requirements) can be better at rolling higher value total scores than other classes with higher minimum requirements.
As I understand it the game just rolls as many times as required until it gets a roll which meets the minimum requirements for the character - and that roll is then displayed. There are thus no adjustments made to the roll, it's just that you see a lower proportion of rolls for characters with higher minimum stats and those rolls are skewed to the higher end of the total distribution. Do you think there is something else going on that affects rolls?
There is a bias towards certain classes having higher rolls due to how stat minimums are handled - the roll is made and then afterwards if the stat in question is "too low to be legal" it gets bumped up to where it should be, which can yield as many as 14 "extra" points (specifically, a paladin who rolls a 3 in Charisma gets 14 "extra" points to bring them to their required minimum of 17)
There are a few caveats to this.
1) This modification is otherwise completely invisible - if you have a paladin with 17 charisma, you have no way of knowing if you rolled a 17 or a 16 or a 3. All you know is that you ended up with a 17
2) The modification more heavily impacts poor rolls rather than good ones. Our example paladin has a forced 17 in Charisma, so they then have 58 spread out across 5 stats (for a spread of 11/11/11/11/11/17 with 3 points left over)
3) Racial adjustments are applied after the rolls are made, not before - a halfling paladin would have 10/12/11/11/10/17 with 4 points left over (halflings have -1 to STR and WIS but only +1 to DEX, with a theoretical maximum roll of 107 as a result and no chance for percentile strength at chargen). Note that it is only as of the Enchanced Editions that you can legally have a stat less than 3 due to racial modifiers - in the original releases, racial modifiers were ignored if it made the stat less than 3. This affects Dwarves (DEX), Halflings, Gnomes (both WIS), and Half-Orcs (INT)
4) This is still an array of 6x(4d6 drop least). You have the same chances of rolling three sixes six times over regardless of if you abuse the Forced Minimums or not. You will eventually roll the perfect 108(00) fighter... though if you're like me, you'll fly right past it out of inattentiveness (my highest rolled character is a 93, my highest seen roll is a 97)
5) If you want a better idea of how each race/class combination sits on the rolls, you need to have a higher statistical base. 1000 rolls for each combination would be a good starting point. (And yes, at least one other player has tried to create an appropriate statistical base)
(yes I know that Paladins have other relevant forced minimums than just 17 CHA but the full truth clouds the principle I'm trying to explain - our example paladin would legally have a 75 with 12/11/11/11/13/17 with no points left over).
Some numbers (see also full table here:
Number of expected visible rolls for a 95+ score, rounded to three significant digits:
Dwarf fighter: 75500
Human fighter: 27500
Human conjurer: 9470
Human bard: 3200
Human paladin: 1270
Human ranger: 996
Elf ranger: 722
You seem to think that rangers don't get great scores, and that there's very little difference between human and elf for them. In fact, rangers have the absolute best chances of 95+ scores; paladins only catch up to them in the 100+ range. Also, the difference between human and elf rangers is more about the elf's -1 Con penalty than the relatively low Int and Cha requirements.
Incidentally, the chances of a perfect roll are entirely determined by how many rolls get discarded. A paladin that only keeps one in a thousand rolls is a lot more likely to see that 108 than a fighter that keeps one in nineteen.
@Pokota you may be referring to one of the many possible ways to generate characters in P&P, but this is definitely not how Baldur's Gate does the stat rolls. People have tested using autorollers to generate hundreds of thousands of sets of stats and the distribution of results is consistent with rolls of 3d6 - but not with 4d6 dropping one.
You also don't need to do large samples to confirm that the way minimum stats are generated is not the way you describe. Taking the paladin as an example, if all charisma scores less than 17 were bumped up to 17 you would get far more 17s than 18s. In fact you will quickly see if you do some rolls that you get about 25% of 18s (as there are 3 more ways to roll 17 with 3d6 than to roll 18).
I couldn't figure out how to tell Excel to use the right numbers for the horizontal axis, so I put that as a disclaimer in the axis label.
Also, I broke out my calculations into their own spreadsheet. Attached here, so other people can play with it and see how the math works. The chart is in that sheet, too.
I was working on this assumption:
(1) All six attributes are rolled (3d6 each)
(2) If a particular attribute is lower than the guaranteed minimum for race/class, it's raised to that minimum
(3) If the total score, after this, is less than 75 it's all discarded.
However, as @gondo pointed out - on that basis - a paladin's charisma should have many more occurrences of 17 coming out than 18s. But, I've just tried it and way too many 18s come up. So, my methodology must be wrong and therefore my video and it's conclusions must be wrong. It's looking like the computer keeps re-rolling until it get's a value over the race/class minimums - rather than doing any adjustments.
So, I need to go back to the drawing board!