Character Generation - Stat Draft Method

[Olaf the Stout]

When we started playing 4e I switched from using point-buy to using stat arrays. This means everyone's starting on an equal footing _and_ min-maxing is minimized.

Of course this leads mostly to cookie-cutter builds, but I'm fine with that. There's still been minor variations (e.g. switching primary & secondary stats).

I definitely like it better than players creating ineffective characters "for the sake of roleplaying". Unless you're playing a pretty untypical D&D game, combat is going to be important, so what's the point of creating a genius, weakling fighter?

It's detrimental to what I consider a fun game of D&D and in the worst case can ruin the game for everyone. It's not even realistic: If you don't have the slightest talent for something, there's no way you'll end up doing it for any appreciable length of time (unless someone else benefits in some way from your incompetence...).

Just to be clear, I'm not trying to get my players to create genius weakling fighters or dumb brute wizards. What I'm looking for is something that doesn't result in the same classes using the exact same stat every time as a dump stat.

So you wouldn't have a weakling fighter, but maybe he would be a bit more charismatic and a little less wise. Or a little more intelligent and slightly less dexterous.

Just something different to the cookie cutter stat builds that show up time and time again.

That, and not have one PC have awesome stats due to rolling well, while another player is Joe slightly above average (in some stats) due to some bad rolls.

Olaf the Stout

 

[Halivar]

I like the grid system. One change I'd make to it, though: next to the 3x3 grid I'd make a "scratch" column of 3 numbers, with 4d6-drop-2-plus-6. If any row does not have a number higher than a 12, replace the highest number on the row with the scratch value for that row.

It yielded some surprisingly organic characters:

• A fighter with Str 13, Dex 15, Con 16, Int 9, Wis 11, Cha 12. The stats definitely weight toward archery.
• A very lucky-rolling and very hearty wizard with Str 10, Dex 16, Con 18, Int 15, Wis 14, Cha 18 (no scratches used). I think this would be my optimal 1st-edition wizard right here.
• A paladin with Str 14, Dex 16, Con 14, Int 11, Wis 13, Cha 12. It would be the first paladin I'd ever rolled up to prefer light armor.

 

[garrowolf]

I use a +12 system. That seems to work pretty well. Everyone starts off with +12 total in bonus points. They spend them as they want with a max of +4 in any one attribute.
Then you roll for even or odd on any kind of dice you want. If it is odd then the first attribute is odd. Take the attribute numbers for that bonus level and take the odd one. Then next one is even, and so on.
It is extremely quick to write up. When a GM allowed me to do it in their games I always had my attributes the way I wanted them and much faster then anyone else.
If you want to run a higher power level game then up it to +16 or something.

 

[eamon]

I think it's worth emphasizing that cookie-cutter stat distributions on the one hand and power disparities on the other are opposites you'll unfortunately need to choose between to some extent. The "right" answer depends on what you're looking for.

What we've been using for years now is natural roll with point-buy fallback: i.e. you roll stats in order; reroll one stat then get to swap too. If you don't like the result, you may discard and use point buy. In practice, in 4e anyhow, this means most people end up using point buy since it's simply better than the vast majority of rolls; those that do keep the rolls tend to be numerically more powerful but also constrained in their choice of class. I don't think this is an ideal system, but it works.

However, I think part of the issue is that power differences are all about perception. I don't get the impression that significant power differences actually occur due to stat differences with a system like this very often. Of course people can get envious of super-stats, but that's not because the real power is that much higher, it's because the stats are known, and dammit I want those two 18's too... and even though such brilliant stats almost never occur, they wouldn't cause power problems compared to say a 17+16 array (which is achievable in 4e's point buy anyhow).

All of this is a roundabout way of saying that I think power-differences are overblown, and I think it's more fun in the long run to avoid cookie cutter stats rather than to avoid power differences (at least as long as you avoid the real crap stats).

So an ideal rolling system for me should:

• Allow some power differences, recognizing that perfect power balance comes at too great a cost
• Encourage randomness whereever it can, and as much as it can
• Avoid too-weak characters to ensure that no matter what, PC's are useful.

Looking at those issues, I prefer a system that does not allow arbitrary stat arrangement (or at least, not without some cost): this encourages weird stat orders. E.g. something like in-order rolling with one free swap and subsequent swaps reduce the higher ability score by one.


There's a fundamental problem with 3d6 or 4d6 drop lowest in that it's just too variable; particularly towards the low end. Getting only roughly average scores leads to weak characters, and it's quite likely with such a system. Conversely, having several very high stats is surprisingly likely: there's just too much variation, and that's unavoidable due to the fact that the dice are memory free: your next 18 is just as likely as the last. A character with otherwise mediocre stats is just as likely to roll a high stat as a character that already has a bunch of high stats, even though in the first case he really needs it to be playable and in the second he really shouldn't get it to avoid power disparities.

 

[Dordledum]

The grid method... Fantastic really. You have choices, but you can't have it all. I would employ the grid for my step 1 in the future, and then follow the rest of my steps.

Aluvial

That's pretty sweet actually, might implement that when I start GM-ing again.

 

[Loonook]

4d6 drop lowest reroll 1s on 3 chosen stats.

With 4d6dlrr1s all stats I pulled an average of 34.7 points, with 2 hitting below 25, and 3 hitting above 45 over 25 sets. The mode is around 35 so definitely heroic but not overly so. Offering a minimum bonus of a certain amount would also help, though I do not know what specifically I would suggest.

Slainte,

-Loonook.

 

[AeroDm]

My group did stat draft for several years as our preferred method during high school. We thought we were very creative for coming up with something so unique... Our method was to do 4d6, drop the lowest with the GM rolling a set as well. We then rolled a d20 to determine initial selection order and proceeded as you suggested up thread, shifting one position each round.

I think it is a strong stat generation method. It ensures balance and fosters a cooperative team environment from the outset. When people roll stats, there is always someone secretly hoping whoever is up rolls poorly. Here, people are congratulating their co-players for rolling well. I also liked tossing a set as the GM because if I rolled something good, I'd remind the player who got it the entire campaign ("That's my 18 in your dex, you know..."). Fun times.

 

[Olaf the Stout]

Hmm. I think that's kind of contradictory, but we'll see. How about this:

The player chooses a class, and then chooses 1 primary ability and 1 secondary ability.

For the primary ability, she rolls 14 + 1d4.
For the secondary ability, she rolls 12 + 1d6.
For the other four abilities, she rolls 4d6, drop the lowest, in order. That is, Str, Dex, Con, Int, Wis, Chr, skipping over her primary and secondary ability.

So if she wanted a fighter, she would roll 14+1d4 for Strength, giving a Str score between 15 and 18. Then 12 + 1d6 for Con giving a Con score from 13 to 18. Then 4d6 drop lowest for Dex, Int, Wis, Chr in order, giving a completely random set of numbers.

This guarantees solid, decent scores for the two most important abilities for the class. However, the rest of the character is organically rolled. There are no "dump stats". You might end up with a smart fighter, or a wise fighter, or a charismatic fighter, but you can't predict it.

Well, we ran through a trial run of the stat draft last night, as well as the method described above.

The stat draft worked pretty well. The stat array was created by each player rolling 4d6, drop the lowest for each attribute, as well as a Wild attribute.

They then rolled to determine starting order. After everyone had picked once the person who picked first that round then dropped to the end of the line. So the picking order went like this

Rd1: 1 2 3 4 5
Rd2: 2 3 4 5 1
Rd3: 3 4 5 1 2
Rd4: 4 5 1 2 3
Rd5: 5 1 2 3 4
Rd7: 1 2 3 4 5

After they had got their stats they could add +2 to any one stat and -2 to their lowest stat. The highest point buy value from this method was 41 and the lowest was 32, which I think is a pretty reasonable spread. The -2 to the lowest stat also meant that all the PC's had a negative stat (which I was happy to see).

We then tried the method quoted above to see how that would work. To compensate for 2 stats guaranteed to be relatively high, I also added in a -2 to the lowest stat.

This produced some hilarious results, mainly due to some extraordinary rolling for the 4d6 drop the lowest part. A couple of PC's ended up with point buy in the 50's!

I rolled up a Cleric and got a 7 for Con, which was taken down to a 5 thanks to the -2 to the lowest stat. Even with that 5, the Cleric still worked out to a 36 point buy. The Rogue had a 17 Strength thanks to a great roll on 4d6, drop the lowest.

Overall it was clear that this method ended up with more powerful characters, even taking into account that the players just rolled really high for stats for the second method. I thought the -2 to the lowest stat would compensate it a little, but it didn't really do it enough.

I was quite happy with the stat draft though. Even though the point buy value equivalent was in the 30's for all but 1 PC, my resident powergamer did (correctly) point out that they still weren't as good as their point buy equivalents as the stat allocations weren't optimised.

Said powergamer still isn't convinced about the stat draft though. I'll just have to try and win him over before next session, when we roll up characters for real.

Olaf the Stout

 

[GSHamster]

We then tried the method quoted above to see how that would work. To compensate for 2 stats guaranteed to be relatively high, I also added in a -2 to the lowest stat.

You could try 3d6 in order. (And no penalties.) That would produce more normal curves. But the cost is that someone might roll low on several stats.

 

[Olaf the Stout]

You could try 3d6 in order. (And no penalties.) That would produce more normal curves. But the cost is that someone might roll low on several stats.

Yeah, I thought about that, but wondered if it might be too harsh. I'll have to give it a test run myself to see how it goes.

I do quite like the primary stat, secondary stat and then the rest in order. It makes for non-sterotypical characters that still have their key stats covered off.

My main worry, as you said, was someone rolling horrible for a couple of stats.

Olaf the Stout