I have my players roll. We use the "Heroic" 2d6+6 method. There have been some very "average" characters but nothing "unplayable".
|
Well 4d6 rerolling 1s, then rerolling a stat is off the hook. That's way better than 25 point buy.I usually let them roll 6 times, 4D6, re-roll '1's.
Then, one re-rolled stat,but you have to take the result, even if it is lower.I offer a 25 Pt buy instead, but nobody takes me up on it.
A few players roll in order for fun, build a character from there. I like this fro time to time, as a player.
-Uriel
It ends up with quite a few 12-15s,but rarely an 18.
Point Buy can get you that as an auto, so it does have it's perks.4D6,reroll 1's is how I have done it since 1st Ed.
I don't allow anyone to ever ditch their set of stats (You know, those cry-babies who complain that their stats 'suck') and roll over.
I started the re-roll a stat after one player didn't get anything over a 13. He ended with a 17, and was pretty happy. We've done it ever since.
I always use Pt Buy for games at Cons,PVP Gladiator-style bashes, etc...
-Uriel
Our DM doesn't like point buy. He says it ends up with cookie cutter stat arrays. As far as rolling, I have the best luck with 4d6 method. Mostly above-average rolls with one or two outliers. I tried 24d6 once---did not work out well.
My original game group and all the progeny groups it spawned use the "Ace" method (so named for the person that suggested it). In 1st and 2nd edition there were requirements to even join a class (if you followed that rule). So we decided that each character could tag two ability scores that would qualify. Now we just roll that way because it's second nature more than any specific adoration or theological significance (I still think Odin smiles on it but whatever)
two stats -> 1d4+14
remainder -> 3d6 assigned in order from Str to Cha
no re-rolls
It can generate a broad range of characters depending on how the 3d6 play out, and some are wildly out of whack with the other characters. But no one is ever crappy at their designated class and you don't get weaselly dump statting.
For those curious regarding the probabilities on the 4d6 drop the lowest *and reroll 1s*:
There are 625 combinations since this is equivalent to rolling 5 sided dice and 5x5x5x5=625.
Of those 625 combinations:
1 will result in a 6
4 will result in a 7
10 will result in a 8
21 will result in a 9
38 will result in a 10
58 will result in a 11
79 will result in a 12
94 will result in a 13
100 will result in a 14
91 will result in a 15
70 will result in a 16
42 will result in a 17
17 will result in a 18
If in both cases (reroll 1s or not) we throw out arrays that generate scores less than 7 (since their costs are not defined), we can calculate the expected value for the point cost for a single ability score:
Without rerolling 1s: 3.45
With rerolling 1s: 4.9887...
The equivalent point buy if we don't reroll 1s is: 6 x 3.45 = 20.7
The equivalent point buy if we do reroll 1s is: 6 x 4.9887 ~= 30
Of course, the one thing I'm still assuming here is that we do not throw out arrays that are below a certain threshold (ie a least one 15+ or total mods +3, etc.) These restrictions only serve to further increase the expected point buy of a random array.
Good work guys. What is the formula for the number of combinations for X dice of Y sides dropping Z lowest?
I believe the binomial distribution can be used if no dice are dropped.
Well, my group recently decided to drop 4e for PF. All but one of us hail from 2e, so there was some debate on this. Some wanted to do 3d6/x6 like we use to in 2e. But, the less adventurous side won out and the debate boiled down to 3d6/x8 or 4d6 DTL, with the later winning out.
Point buy was never even considered, and I think that has more to do with the fact that most of us just like to roll the dice :) The 2d6+6 does sound interesting though, and I think the next time we roll up new characters I will try to encourage them to go in that direction.
Now typically on the rerolling 1s, doesn't it stay a 1 if you that's the result of your reroll. If you just keep on rerolling it, it seems like people should just use 1d10/2 +1 to save a step.
I'm not sure *4d6 drop lowest and reroll 1s* is an official rolling method so you could do it either way. My numbers above assume that it would be rerolled until it was a non-1.
The (1d10/2)+1 doesn't quite work unless you (contrary to D&D standards) assume you're rounding up when you divide by 2. Naturally, you can use any method you like to generate the same numbers as long as you have 5 equally likely outcomes.
|
Here's an interesting dead-lock stat generation method:
Start every stat at 1 and roll 69d6. For every die that lands on a 1 add 1 point to str, for each 2 add 1 to dex and so on.
If your int ends up less than 3 (extremely unlikely) then up it to 3 and call it a "freebie" (since being that dumb is going to be a massive hindrance).
For the less adventurous start each stat at 5 then roll 45d6. Or 7 and 33d6.
In any case it ends up equivalent to a 15 point linear point buy.
(You may also want to set a certain number above which a stat will not rise, like 18, in which case re-rolling the excess dice would be the solution.)
For a "controlled" deadlock:
Have the player assign a set quantity of the points (on either linear or regular point buy) and roll the rest (on linear for simplicity). The more points they get, the more controlled it is. If they get all of them it's full control, if they get none it's zero control (and pray to the dice gods).
My group has done 15 points (linear) and 5 dice for the last couple of characters. One character made a fighter that was just intelligent enough for combat expertise but had 4 dice land on int and was suddenly the smartest guy in the group.
To encourage using randomness you could give players one of three options. (Assume starting at 1:)No control and 74 points, 36 control and 35 points or 68 control but no random points, for example.
Note: If you give the player any control *do* cap the result to 18-20 or you'll end up with people putting 1s in everything but their one good stat and letting statistics reliably give them another 4-5 points on it for one massively unfair stat. If there's no control the probability of greater than a 20 is negligible.
I have used a variety of systems as both a GM and a player, with the 4d6 method being the one used most often. All the methods presented here actually makes me think about whether GMs who employ these methods use the same methods when creating their NPCs?
For example, if you let your players roll 4d4+2 or 3d6 straight across the board do you use the same method for the NPCs, use a different method, use point buy or just create stats on the fly?
Also, if anyone runs modules (such as the APs) do you ever worry that your dice rolling method for players will make them too strong or not strong enough for the adventure (or the NPCs they face)?
I'm trialling point buy for the first time and find that it at least creates fairly equal character's stat-wise. In some ways it's not as much fun but at least I don't have characters with wildly varied stats (like two people who used the old 6d6 method - one didn't have a stat lower than 15 while another didn't have a stat higher than 13 - the power disparity was almost painfully obvious).
I never use a random generation method to create characters (unless I'm playing HOL) and I hardly know anyone who does. First, the most important aspect of game balance is having he characters balanced with one another, and the easiest way to miss that is to have them starting out with imbalanced attributes. Second, I want the players to create characters they're interested in playing, not start out the creation process with a potential disappointment.
Over time, stat rolling techniques develop more and more ways of minimizing the random element. For example, the original method of rolling D&D stats was to roll each stat, in order, on 3d6. Later, players were allowed to arrange the results in any order, then encouraged to roll 4d6 and drop the lowest, then to do both of those and reroll any set with a low total bonus. In addition to each of these alterations increasing the overall scores, they reduce the width of the curve.
My current favorite method for D&D/Pathfinder attribute assignment is to give everyone 18, 16, 14, 12, 10, 8 and allow them to arrange as desired. Also, if they'd like to take from any high stat to add to a lower one, they are allowed to do so without limit.
We do 4d6, re-roll ones, place as you like.
When I create NPCs I either use the standard NPC arrays or eyeball the stats.
|
|
I started playing with the 3d6 in order with one reroll. But we added Luck as a stat and if you were lucky and had a high score you could reroll another ability if you wanted. This way we ended up with more role playing.
The GM allowed for in game improvements for abilities. If you wanted to improve your strength you suffered through barbarian basic training and your strength went up by one. Wisdom could only go up with experience, but since Intelligence was book learning you could study at the library.
Hit points were rolled fresh at every session. So if you rolled poorly that meant your character had a cold or you were still recovering from the last beating you got from a previous adventure.
Some of the resulting characters were rather interesting. Rob named his character "smith" because he rolled all 10s and 11s. He figured that Smith would died quickly and he could roll up something more interesting later. But Smith went on living and became Rob's favorite character even with the poor stats just because he had had so many interesting adventures.
But now that we play so many Living campaigns the point buy gets used quite often. Dave was surprised when we started a new game because we were using point buy and he was still rolling. I think as we start King Maker I will suggest we try rolling again just for the fun of it.
|
|
@Phil. L : I always use the same stat generation method for the "heroic" baddies as I do for the PCs, but I don't for generic enemies or regular monsters. I also try to keep the stats no more than a few points outside the normal range, though (25 point buy equivalent is my upper limit, above that and it's one-shot-only range).
@Blueluck : Look at the random stat generation method I proposed two posts above yours. It's deadlock, it forces you to work with what you get, but also ensure that every player has the same total ability score. Random != inherently unbalancing, you just have to use the random element at the right spot (assigning location of stats instead of quantity).
Good work guys. What is the formula for the number of combinations for X dice of Y sides dropping Z lowest?
I believe the binomial distribution can be used if no dice are dropped.
The total number of combinations is what it is. I didn't use a formula to "drop the lowest" it's just that your only using three dice out of four to calculate your score. However, since all four dice figure in your result, the total number of combinations is simply 6 to the 4th.
The result is a discrete probability distribution, so you'd have to use the correct formula for that. I already did so and, as reported in a prior post, the mean was 12, and variance was 8.04 and standard deviation was 2.84.
Erm, my bad, I mistyped what I was trying to ask about. Yes, the total number of combinations is Y^X, where Y is the number of sides and X is the number of dice. What I meant to ask about (instead of what I can see I asked about now that I reread my post) is the formula for calculating the number of combinations to that would yield a specific number (again for X dice of Y sides dropping Z lowest). So for example, you post that the number of combinations to get a 15 when rolling 4d6 drop the lowest is 131, so my question is what formula could be used to calculate that (instead of just using a brute force approach, although that would also work).
@Phil. L : I always use the same stat generation method for the "heroic" baddies as I do for the PCs, but I don't for generic enemies or regular monsters. I also try to keep the stats no more than a few points outside the normal range, though (25 point buy equivalent is my upper limit, above that and it's one-shot-only range).
Yep, I do the same thing! The major opponents (or allies) get to use the same ability score generation as the PCs and the minor ones don't.
I don't however stick to a 25 point buy equivalent. I like the characters to be overpowered, so I don't have to hold back. :)
@Blueluck : Look at the random stat generation method I proposed two posts above yours. It's deadlock, it forces you to work with what you get, but also ensure that every player has the same total ability score. Random != inherently unbalancing, you just have to use the random element at the right spot (assigning location of stats instead of quantity).
Yes, the method you posted is pretty interesting. It is not at all unbalancing, but on the other hand, it can be more restrictive than rolling with more conventional methods - i.e. it is pretty likely that players will not get the play the character they want.
Personally, I am not that concerned with the intra-party balance and thus don't care much if the PCs have disparate scores, so long as all the PCs are significantly above the world average overall.
Still rolling here also.
We did 3D6 3 times keep highest per stat in 1st Ed.
We Did 4D6 drop lowest in 2E - 3.5E
Now in PF when a character buys the farm it's 4D6 drop lowest and reroll 1s.
Always put the stats in the order you like.
Tried Point buy once and the players kinda just were unhappy with it.
Have Fun out there!!
~ W ~
Just my two cents.
I just ran a bunch of new players through D0. There were four of them, and I used 2d6+6 and had them roll hit points too, though I let them re-roll ones.. Highest stat rolled (just one) was a 17, and there were several 8s. The druid had 2hp.
Gotta say, it seems like a lot of the material out there is written assuming a highly optimized party. Including D0.
I really really had to be nice to them not to wipe the party with the tatzwyrm and the witch's cauldron. They had fun, but it was a tough adventure.