Character Generation - Stat Draft Method

[[OMENRPG]Ben]

You could do the heroic method if you want your players a little on the strong side. All stats start at 6, roll 2d6, re roll 1's. That way, every attribute is at least ten, with most trending around 15/16. If this is too strong, you can just roll and have no rerolls. I've always given my players the choice between point buy and the classic 3d6 seven times, drop the lowest.

 

[Olaf the Stout]

I got rid of ability scores in favor of ability bonuses several campaigns ago, and haven't regretted it at all. The deciding factor was a group with, like yours, a wide disparity in ability scores (the high character had a total of +11 in ability score bonuses; the low character had +4). I settled on +6 as a good starting point.

Characters start with +0 (ie 10) in every stat), and add bonuses as they see fit. They can gain up to 2 more points by taking an equivalent penalty. I usually restrict starting characters to a max bonus of +3 in any one score. (You could increase this to +4 for demihumans in their bonus stat).

It's a perfectly clear, simple system that I've been 100% happy with. You can't min/max it by tweaking your odd/even stats. Everyone has the same bonus total.

If you really want 3-18 scores, do as above, and then roll 1d6 for each stat. Odd means the stat is odd, even means the stat is even. There's almost no mechanical meaning to 3-18 scores in 3e/3.5e, though (weight allowance is the exception, I think, and that's easily calculated.)

Under your system, what is the downside to taking an odd number in every stat?

Odds and evens don't matter for the purposes of bonuses (10 and 11 are the same for that purpose). However, an odd number in a stat means that at level 4 you can put the +1 in it and increase the bonus.

The actual stat number also makes a difference when Wish spells increase the inherent bonus to a stat, or when a PC takes ability score damage.

Olaf the Stout

Edit: It is also still a point buy system, which means you will still end up with boring min-maxed stats. Fighters dumping Cha, Int or Wis, Wizards dumping Cha or Str, etc.

 

[Olaf the Stout]

I was going to talk about modelled (or constrained) stats , where you use point buy to simulate what you roll, but when I asked Invisible Castle to give me a test set to work with, the first thing it gave me was 12 8 10 15 13 14. Very funny, Invisible Castle. Very funny.

The second go came up 15 13 13 9 14 14 . Perhaps not the greatest set for my purposes, but this needs to be able to work for anything. I prefer Pathfinder, so I will use its 15 pt scheme for my example. That array works out to 22 pts. You could turn that into 14 13 13 13 12 9, which has a similar spread, or 15 13 12 12 12 9 if you think at least a 15 in one stat should be retained.

Another is 13 12 14 17 15 16, a whopping 40 pts, which could probably be reduced to 15 13 12 12 11 10 or 14 14 12 12 11 10. Or even 17 12 11 10 10 9, if again you want them to preserve their high stat.

Finally I got a low score, 11 8 13 13 13 9. I think I would turn that into 14 14 14 11 10 9.

Hmmm... I suppose this sort of system requires a bit of finesse. There is no set way to deterime how stats get changed, though the general idea is to preserve the "spread" in a reasonable way. I am pretty sure that I, as a player, would be happy to play in a game where stats were developed this way though.

Edit: I missed that you were using 28 pt buy, which rules out a Pathfinder conversion, .That would the conversions something like:

2) 15 14 14 13 13 9 = 15 14 13 12 12 9
3) 17 16 15 14 13 12 = 17 13 12 11 10 9
4) 13 13 13 11 9 8 = 15 14 14 12 10 10

Can you explain a little more what you are doing Kaodi? I have no idea how you are changing the stats around or increasing/decreasing them.

Olaf the Stout

 

[Olaf the Stout]

You could do the heroic method if you want your players a little on the strong side. All stats start at 6, roll 2d6, re roll 1's. That way, every attribute is at least ten, with most trending around 15/16. If this is too strong, you can just roll and have no rerolls. I've always given my players the choice between point buy and the classic 3d6 seven times, drop the lowest.

That method doesn't really interest me to be honest. I like that PC's can end up with a stat below 8. Weaknesses make for fun situations. Last campaign (which used point buy) had a Warblade with Cha 6.

Instead of just keeping his mouth shut in all roleplaying situations, he instead decided to play that Cha 6. He would often make comments that the PC's would regret, whilst making for a lot of fun for me and the players.

By the end of the campaign he had actually spent quite a few points in diplomacy to give his character at least a slight clue of what not to say and do. It was great to see that growth as the campaign progressed.

Olaf the Stout

 

[Olaf the Stout]

Hmmm, another option could be to use the 3d6 method, number of rolls = number of players, but then I add in a 16, 17 or 18 for each stat on top of that (perhaps rolled randomly d6 roll, 1-2=16 3-4=17, 5-6=18).

I'm just trying to find some way to end up with all PC's at a similar point buy level, while at the same time having some high and some low stats.

Olaf the Stout

 

[Kaodi]

Can you explain a little more what you are doing Kaodi? I have no idea how you are changing the stats around or increasing/decreasing them.

Olaf the Stout

I will do my best to explain.

All of the initial stats were rolled with 4d6 x 6. The general idea is to replicate with 28 point buy something similar to what you rolled.

So on roll #2, I noticed that all of the stats were clumped together in the mid-high range, except for that one lone 9. When I changed the stats, I thought that perhaps it was best when using this method to maintain the high score in most circumstances, so I kept the 15. Then with the others I tried to preserve than "clump" of abilities that were 12+. 15 14 14 13 13 9 is equivalent to 29 pt buy... hmmm... Ack... it seems I made some mathematical errors (hence the confusion, I bet). So, really to make it 28 pt buy, the thing to do is to make either one of the 14s a 13, or one of the 13s a 12. Reducing one of the 13s to 12 to get 15 14 14 13 12 9 might be preferable, since that has a higher total bonus.

In #3, the key thing is that the abilities start at 17 and then descend in order. So I thought that the thing to do in this case was to keep the 17 and make sure the other abilities still descend in order, because if you recuded the 17 as well you would pretty much just get the standard array, and that would be kind of boring (though you could do that I suppose, and get 15 14 13 12 11 10). 17 costs 13 points, and then the 13 12 11 10 9 costs another 15, to make 28.

In #4, none of the abilities were high, but the top 3 were identical, and the bottom 3 in order. If you bumped every stat up by 1 you get 14 14 14 12 11 10, but that is only 25 pts. My intuition then was to bump up a 14 and then 11 to get the extra 3. Alternatively, I suppose you bumped the higher abilities in 13 13 13 11 10 9 to make 15 15 14 11 10 9, which is also 28 pts.

I said this was a finesse system, hehehe...

 

[Nellisir]

Under your system, what is the downside to taking an odd number in every stat?Odds and evens don't matter for the purposes of bonuses (10 and 11 are the same for that purpose). However, an odd number in a stat means that at level 4 you can put the +1 in it and increase the bonus.

+1 bonus to an ability score (bonus) at level 4, 12, and 20. Or 8 and 16, if you want to make them wait. Or determine odd/even randomly, as I've already spelled out.

The actual stat number also makes a difference when Wish spells increase the inherent bonus to a stat, or when a PC takes ability score damage.

Wish spells refer to inherent bonuses. I can't find any reference to the "actual" stat number, though I'm not sure what system you are playing. I believe pre-3e D&D systems had "maximum" stats, but that was replaced in 3e with a maximum +5 inherent bonus. I don't see any reason my system wouldn't work with wish spells.

Halve ability score damage. This also makes ability damage work the same for every character and every stat, rather than the odd (and admittedly rare & outlying) circumstance where a character takes 1 point of ability damage but suffers no meaningful consequence.

Edit: It is also still a point buy system, which means you will still end up with boring min-maxed stats. Fighters dumping Cha, Int or Wis, Wizards dumping Cha or Str, etc.

OK, I skimmed the thread and didn't realize point-buy was out altogether. Mea culpa. Roll 1d6 to pick a stat. Roll 1d4-1 to determine bonus (-1 total = +0 ability bonus). Variant A) Repeat until bonuses total +6 (or +5, or +7...you're the DM. Make the call.)
Variant B) Repeat, rolling 1d4-2 to determine 2nd bonus (-2 total = -1 stat bonus, and adds +1 to the bonus pool). Repeat, rolling 1d4-3 to determine 3rd bonus. If necessary, repeat, rolling 1d4-1 to determine 4th bonus, and 1d4-2 to determine 5th bonus. 6th bonus is whatever number of bonus points remains.

I'm sure you can think of variants that would suit you.
Cheers.

 

[delericho]

I've used a lot of stat generation methods over the years, and found every single one of them somewhat dissatisfying for some reason or other. Random rolls do seem to generate power disparities, point buy does seem to generate "cookie cutter" characters (and is more complex).

The organic method described up-thread is quite fun, but still generates power disparities, and runs the risk of creating a character that just sucks.

These days, my inclination is simply to offer three methods side-by-side, and let each player make a choice:

1) Fixed array: 16/15/13/12/10/8, arrange to taste.

2) Random roll: 4d6 drop lowest, arrange to taste. Reroll if the highest stat is 13 or lower, or if the total modifier is 0 or less. (This method can only be used if rolling in front of the group.)

3) 28-point buy, using the standard costs from the 3.5e DMG.

On average, #2 is roughly equivalent to a 30.5-point buy, but doesn't allow for the fine-tuning of point buy. #1 is a 29-point array, but again isn't fully optimised. #3, which does allow for full optimisation, then, is very slightly weaker in terms of raw power.

This should avoid 'cookie cutter' characters, as players each choose different methods. Unfortunately, including #2 does lead to the possibility of power disparities in the group. Of course, you could always drom #2, and only allow the fixed array or the "weaker" point buy?

 

[Olaf the Stout]

I think I am going to go with Option 5 - Roll 4d6, drop the lowest for stats. Number of rolls = number of players (so with 5 players there will be 5 Strength stats to choose from). There will also be a set of Wild stats. This Wild stat can be picked as any stat.

Players will roll to pick first in round 1. The order will then go as follows:

Rd 1: 1, 2, 3, 4, 5
Rd 2: 2, 3, 4, 5, 1
Rd 3: 3, 4, 5, 1, 2
Rd 4: 4, 5, 1, 2, 3
Rd 5: 5, 1, 2, 3, 4
Rd 6: 1, 2, 3, 4, 5

Once they have picked all their stats they can add 2 to any one stat, apart from their lowest stat. They then have to take 2 away from their lowest stat.

Repeat the whole process again. Each player has 2 sets of stats. They pick 1 to keep and make a character with. The 5 sets not chosen (1 set per player) get set aside and used for replacement characters.

This should give some high stats as well as at least 1 low stat for each PC. Given that there is a Wild set of numbers in addition to the 6 stats, I expect the average stat to be slightly higher than roll 4d6 drop the lowest. On the other hand, the players won't get as much choice in where their stats end up compared to 4d6, drop the lowest, arrange as desired.

I'm going to give it a trial run next session with my group. If it results in characters that are too powerful (i.e. point buy 32+) I will probably also get them to take a further 2 points from any one stat, just to bring the power level back a touch.

Olaf the Stout

 

[Lwaxy]

For example, the player of the Fighter rolls much lower stats overall compared to everyone else and also rolls a low Strength and Constitution. So not only does he have much lower stats overall, but he is also only able to cover off one of his 2 key stats.

Olaf the stout

That's why you pick class after you rolled.