Variant Method For Rolling Ability Scores

download (2)

The way that ability scores are rolled in D&D has always seemed a little strange to me. For a normal character with no bonuses the scores rolled with D6 should range from 3 to 18. This puts the mid point at a 10 which makes sense. An ability score of 10 or 11 provides no bonuses and carries no penalties. It’s average. The issue that I have is the scores that fall in the below average range.

The 4d6 drop the lowest die method works okay, but it does open up the possibility, however unlikely, of a worst-case scenario of the player ending up with six scores that are 3’s. A good DM is going to let that player or any player who rolls up a character who is a below average stinker to re-roll for ability scores, but that wastes time during character creation. It doesn’t make sense to play a character with ability scores that are low across the board. It would be a challenge for the DM to create an adventure where that player wouldn’t die every twenty minutes. It doesn’t make sense in the game world either. Why would a sickly, weak idiot be out looking for adventure? They certainly wouldn’t be looking forward to a long career.

The point buy system tries to solve this problem, but it limits the available scores to a range between 8 and 15. There is no possibility for exceptional ability scores without racial increases. The best solution to solve the problem of low ability scores is to completely take them out of the picture. This can be done quickly and easily by changing the method used to roll ability scores. Just have the player roll a D12 one time and add a 6 to the roll. Repeat this six times and the player ends up with ability scores that all fall within a range of 7 to 18. Since scores between 7 and 9 still carry a penalty you can still make characters that have defects, but no score will be low enough to make that character useless.

Do you like this method? Do you think the players will like this method? Let me know what you think.

7 thoughts on “Variant Method For Rolling Ability Scores”

  1. I’m not sure another ability generating method is strictly necessary, to be honest.

    By and large, your scores are going to flatten out, so you’re really only looking at accommodating edge cases; my preference is to coordinate with PCs during chargen, and if they’ve rolled a truly abysmal score — or if they’ve got their hearts set on a particular character type/class that they don’t have the scores for — give ’em another shot. It’s interesting that in 5e, the system pushes you towards figuring out what you want to be, then rolling your abilities, which sort of bears out my approach.

    An alternative method might be simply to remove the cap on point buy, and open up lower ability scores to buy back points. That way if you’ve really got your heart set on 18 STR, you’d have to seriously consider the drawbacks of a 5 (or whatever) WIS score.

  2. I’ve always been a fan of knowing what class I want to play and then filling in my scores to allow that class. Rolling with my method eliminates the need for rerolls for terrible rolls. It doesn’t add any work or take away anything from the character creation. It just simplifies things.

  3. I had two players roll disappointing results on their SECOND tries, while the other two rolled very well, one of them being overpowered. I’ve been considering different options for the start of my next campaign, and probably spending way too much would-be prep time obsessing over numbers (numbers are fun for me). I’m using anydice.com by @catlikecoding to compare probabilities on different methods of character generation. I’m probably going with 2d6+6. Here it is compared with the standard roll. I added 1d12+6 to the program as well, and just to see what happens, 3d4+6:

    http://anydice.com/program/56da

    This is a really great way to directly compare different methods side by side. But it’s only showing you the possible outcomes of one roll. To get a better sense of what you can expect after rolling for all six abilities, you get these:

    http://anydice.com/program/56db Standard. Based on this graph, I changed the standard array at my table to 16, 14, 13, 12, 10, 9

    http://anydice.com/program/56dd Another method I considered was rolling seven times, keeping the best six

    http://anydice.com/program/56dc 2d6+6. Because there are only two dice, this yields a neat symmetrical graph

    http://anydice.com/program/56de 1d12+6. With only one die, you get a wild curve at the extremes.

    http://anydice.com/program/56df 3d4+6. Nevermind, this graph is symmetrical too and I don’t know what I’m talking about. This method is probably too powerful for my taste.

    And finally, if you are considering allowing 1s to be rerolled, I found this comment from the creator…
    “First, you could either reroll 1s a single time, sticking with a possible new 1, or eliminate the 1s by rerolling them until you get something better. Second, you could do this either before or after keeping the highest three out of four dice. See this program to compare these options: http://anydice.com/program/37c

    By the way, I can’t recommend anydice.com enough. It’s really powerful if you are savvy with it. If you’re inept like I am, the person who runs the site is very helpful and usually answers within a couple hours on Facebook and Twitter. It’s played a big part in the development of my house rules for character generation and hit dice.

    1. Here’s my concern with the 2d6+6 method. With the 1d12+6 method you have a 1 in 12 chance of rolling any number over an infinite amount of rolls. With the 2d6 method there is only a 1/36 chance of rolling the extremes like an 8 or an 18. The ability score probabilities all clump together around 7. There is a 1 in 6 chance of rolling a 7, but there is only a 1 in 36 chance of rolling a 12.

  4. I relish randomness in my games. The first thing I did with D&D5 was create my 2-page summary of the rules, which I do with every RPG I read. The second thing I did was create random character generation tables. If I’m generating attributes randomly, I want the entire character to be random. Everyone who has played in my games loves this method – particularly the random backgrounds, which are definitely the best part of D&D5.

    Given that perspective, I have to say that this dice question seems like a non-issue to me. I think that d12+6 is a great way to ensure that the players have at least average characters. But the essential point of rolling dice is the gambling aspect: the balance of risk vs. reward. Removing the risk (or simply lessening the risk) begs the question: why use dice at all?

    A number of additional points:

    The assumption that a character with all six attributes as 3s is undesirable is flawed. First of all, rolling all 1s on 18 rolls (or 24 rolls, if you’re talking the now-standard 4d6 method) is so fantastically improbable that any gamer who actually did that would scream with delight at the amazing once-in-a-lifetime, no-one-will-believe-this aspect. Second, playing a character with incredibly low stats is FUN, particularly as the probabilities make it that the opportunity to play such characters (randomly generated) are very rare. Only power gamers would balk at ever playing a character with way below average attributes, most likely because they don’t see the fun in role-playing such a character. Why did a sickly weakling become an adventurer? What a great opportunity for a fun backstory!

    Also, we gotta admit it: D&D (in any edition, with the possible exception of 4th) is extremely deadly for beginning characters. You say in your article “It would be a challenge for the DM to create an adventure where that player (sic) wouldn’t die every twenty minutes.” But the thing is, the fact that dice are used in the playing of the game makes it very likely that characters will die, no matter what their stats are. D&D5 seeks to ameliorate this by dropping the XP requirement for level 2 down to 300, so that a lower percentage of characters will die before gaining 2nd level and increasing their chances of survival. But it’s still a pretty high mortality rate, unless the DM is molly-coddling the players. So (unless the DM is such a softie), players shouldn’t have the expectation that their new characters will survive long, *no matter what the character’s attributes are.* They should have hope, but not the expectation. Given that each beginning character they generate is more likely to die than survive, the cost of having a low-attribute character is even less (and the delight, on the off-chance that the ‘underpowered’ character is the one who gets lucky with the INIT and ATK rolls and manages to survive, even greater).

    An example: INIT is incredibly important. A character with 18 DEX is going to get a +4 to INIT, while a character with a 3 DEX is going to get -4 to INIT. That seems pretty big, but the fact is that they are equally likely to have a 5 for initiative in combat, because with the 1d20, any number is equally likely to be rolled. So Barry Badass, with his all 18s, walks into a cave with Larry Loser, who has all 3s, and they run into 1d6 goblins. The DM, Barry’s and Larry’s players all roll for INIT, and the goblins (rolling a 16) get 18, Larry (player rolled 17) gets a 13, and Barry (player rolled 8) gets 12. If the goblins are lucky on their attack rolls, both characters get slaughtered.

    Finally, I’d like to respond to the idea that re-rolling extends character creation. Maybe there are new players who don’t know the high mortality rate, or maybe there’re players who are fed up after the last 10 characters have died and the DM wants to help them feel good and so allow them to re-roll low stats (what you referred to as a “good DM”); this doesn’t take long at all, unless the players can’t do rudimentary addition, and in that case, the DM could tell them the result each roll. Seriously, the dice rolling is the fastest part of character generation. It’s the rest of chargen that takes a long time; the picking of spells, the writing-down of equipment (or even buying it, if you allow the players to buy gear as part of chargen), the reading of all the class abilities, choosing the background….each of these take ten times as long as rolling dice.

    So, to summarize: d12+6 is a great method to generate stats for a supers campaign where there are few below-average people. Otherwise, I don’t think it’s necessary.

Join the Party