Discourse on the Methods

As a DM, the first thing you will do when starting a new campaign is to create your world. You’ll at least sketch out some details of the environment where the PCs will begin: the starting location (will it be a village, town, or warren of caves?), the balance of good and evil (are threats rare in a peaceful society, are the PCs starting in a hostile wilderness, or are the PCs effectively underground resistance fighters toiling against the dominant evil?), initial NPCs, and adventure opportunities that will present themselves in your campaign (vile dungeons, dark towers, dank jungles, invading armies). All this is good preparation, but it’s all stuff that you will likely prepare all on your own. The first thing you will do with your players is help them create their characters. In AD&D, the first part of creating characters is generating ability scores.

Ability scores are very important, as they determine which classes a player can choose from to select for his/her character. The Dungeon Masters Guide (DMG) details the four possible Methods for determining ability scores for player characters. You, as the DM, have to determine which Method will be used in your game. Each is subtly different, and while each can produce any possible set of ability scores, they are different enough that they produce consistently distinct averages. As such, you need to choose which Method will work best in your campaign/which Method your players will use.

Why it matters

Holding a session where everyone makes characters together really helps the gaming group come together and bond with one another. Players can celebrate high rolls, enjoy background as it develops, and may even contribute ideas to one another as the characters are being formed. Along those lines, having one method that everyone uses publicly to generate his/her character just makes sense, as then it’s fair and consistent to each player. Players who generate characters together tend to care more about the group than those who create their characters using who-knows-what methods in secrecy. When players generate characters outside of the group, there can be suspicion: was that 18/76 Strength really rolled legitimately? So, if you can, set aside the first session to generate characters and provide setting and other introductory information about your campaign.

Pre-generated characters vs. player-made characters

You may have ideas on the sort of campaign you wish to run, and your story hooks already hinge on having a psionic bard in the party. Do you really want to leave it to chance? Wouldn’t making up the characters yourself be easier?

I strongly urge caution here. While pre-generating characters (that is, the DM making characters for the players) can save time, I really don’t recommend forcing “pre-gen” characters on players outside of a convention game. For a convention game, which is typically played in only one session, from two to ten hours long, you want the players to hit the ground running and start immediately, because you don’t have the time to work on characters, setting, backstory or relationships. But if you are planning on running an ongoing campaign, you really do want to allow for those things, and players almost always want a hand in determining what they are going to play for the next several weeks or months of game sessions. Sure, you can do some pre-generated characters so that you can offer them to a player who just doesn’t want to be bothered making his/her own character, but most people love to generate characters, particularly as a group.

So, you’ve arranged for all your players to join you and make characters as a group, and the first part of making characters is rolling up the ability scores. Which method will you use?


Not a Method

First, a note on things that actually aren’t official methods for generating ability scores. For many of us who came to AD&D after playing other editions of D&D (for me it was the D&D Basic Set), we have to break our habits of using other ways of generating ability scores from these other sources. In AD&D there is no rolling 3D6 once for each ability, to generate the abilities in order (STR, INT, WIS, DEX, CON, CHR, COM), as was done in D&D (but see Method Four for something similar). There’s also no option for allocating points instead of rolling dice, in order to get artificially balanced scores, like you might find in later editions. Nor is there any allowance in AD&D for trading numbers between ability scores once you’ve rolled them. They roll what they roll, but your players do get some choice in how they roll the abilities for their characters…and some of them can be of benefit to you as the DM, too.


The Four Methods

Of course, you can have your players generate their ability scores however you like, and there could be as many ways of generating numbers from 3 – 18 as there are DMs. You could play a hand of blackjack for each ability, pick numbered balls from a hat, throw darts at a dartboard…there are many, many ways you could concoct. But there are the four official Methods described in the DMG, and they are sufficiently varied as it is. Why not try them? They are described on page 11 of the DMG, but here they are as well. Note: for the purposes of this article, I’m assuming that you will use all seven characteristics. You can, of course, choose not to use comeliness in your campaign, either because you don’t have the Unearthed Arcana (UA) or just don’t like comeliness.


Method One:

The player rolls 4D6 seven times (once for each ability). Each time, he/she adds the three highest dice results for an ability score, discarding the fourth, lowest die. Then, after the seven scores are recorded, the player assigns them to the seven abilities in any order. Perhaps not surprisingly, this Method consistently generates the highest average ability scores of any of the four Methods, and thus has the greatest likelihood of allowing players to play some of the harder-to-achieve classes, like paladins or monks. In my experience, this is the method that most DMs use. It has two advantages: generating consistently higher numbers on average (because a low number is usually discarded) and doing fewer dice rolls to generate a character’s ability scores. The player rolls only once for each ability, so the ability portion of character generation can go that much faster.


Method Two:

The player rolls 3D6 fourteen times, recording all of the results, then keeps the top seven scores. These are then assigned to the seven abilities in any order. In a way, this is an alternative, almost-inverted approach to Method One: instead of rolling more dice a minimum number of times, you roll the minimum dice more often. The result is characters with slightly lower total ability scores (on average) than Method One, though the rate of getting a single high primary ability for a desired class are almost the same. This Method has some advantages. First, the resulting characters tend to be more “normal” or “exceptional in only one ability” rather than “extraordinarily gifted across all the abilities,”  which is desirable in a more low-fantasy campaign. As a side benefit, this also makes the rare cases where you achieve a hard-to-get set of ability scores like the minimums for a Bard all the more exciting. Second, like Method One, it also gives the player a greater sense of control, in choosing which scores to assign to which ability.


Method Three:

The player rolls 3D6 six times for each ability, in order (six rolls each for STR, INT, WIS, DEX, CON, CHR, COM, for a total of 42 rolls), and records each of the results. The player then takes the highest results in each set of six and assigns it to the corresponding ability score. This Method results in characters with almost as high ability scores, on average, as in Method One, but it arrives at those scores in a much more “random” way…that is to say, the player is going to end up with a set of ability scores that he/she can’t switch around in order to qualify for a particular class, so that available classes will be entirely based on the rolls. This means that the classes with particularly high minimum ability scores (Paladin, Bard, Thief-acrobat, any character with a shot at psionics) will be even more rare than they are with other Methods, and there might be a slight increase in the “default” classes with low minimum requirements (Fighter, Thief, Cleric, Magic User). This Method thus has the advantages of greater randomness (for some lifelong players this is definitely an advantage) and even more of a low-fantasy feel, while still having a good likelihood of getting above average scores for suitably “danger-worthy” characters. DMs who want illusionists, monks, bards, cavaliers, paladins and psionics to be very rare in the world but who still want their players to have a good chance of survival in the world would do well to select this method.


Method Four:

The player generates a complete set of ability scores, rolling 3D6 for each of the seven abilities, in order (STR, INT, WIS, DEX, CON, CHR, COM), twelve times, for a total of 84 rolls. This results in twelve complete sets of abilities. Player chooses the one set of ability scores he/she wants, and discards the rest. This method is the least forgiving of all. While still not as brutal as the roll-once-for-each-ability-in-order method of previous editions of D&D, this still is very reliant on fickle dice. This method is the most likely to result in characters with wildly varying ability scores, and so makes the four basic classes (Cleric, Fighter, Magic-user, Thief) extremely likely to result for most players. This is the “hardcore” method, and yet also the Method closest to the one used in various D&D video games like the Baldur’s Gate series (before you gripe, I know, I know: that was AD&D2), so in that respect it should be familiar to many players: you roll all of the ability scores as a set multiple times, then choose the set that you like best. This Method is particularly suited to DMs and players who want to play in a low-fantasy setting where death is a constant companion and PC mortality is likely if combat is pursued regularly (because of the rarity of having all three physical/combat abilities higher than normal/average). This means that this Method is desirable for DMs who are either going to focus on role-playing and/or puzzle-solving at least as much as on combat, or who want to see how far the PCs can get before getting a total party kill (“Six rooms! That’s a new record!”). An additional advantage to this Method is that the DM can take the eleven sets of ability scores that weren’t used from each player and use any or all of them as the basis for NPCs that the party will meet in its adventures! A good DM can even make creating those NPCs from the players’ cast-off ability scores an interesting part of that initial session of character creation and campaign introduction:

The DM grabs a set of discarded ability scores from Jane, circles it, and says “You all know Henrik here, who grew up in the village of Rushford alongside Tibbot (a PC) and Marlin (another PC). He’s the second son of Arvek, the merchant who still runs goods between Rushford and New Bayside. Henrik lived with his mother in Rushford until he was old enough to work the caravan. He was weak (STR5), somewhat slow-witted (INT7), and a bit clumsy (DEX7), but he always had a long-term plan (WIS15). Henrik worked the caravan route with his father for years, saved up his earnings, and opened the Traveler’s Rest Inn and Tavern three years ago. It’s a regular gathering place for the locals as well as a regular stop for several trade routes, and Henrik knows and recognizes each of you.”


An aside on dice:

Here’s a fun and time-saving approach for dice afficionados who want to try Method Four (even Two or Three with a little modification), just want to quickly make random NPCs before a session, or find they need to make a set of ability scores quickly during the adventure. As a dice afficionado, you will likely have a can, jar, box, or bag overflowing with dice. Sift through your dice and find seven sets of 3D6; each set should be easily distinguishable from the others, like a different material (wood, bone, metal), size, or color. Specify which set will correspond to each ability score, then pick up all 21 dice at once and toss them simultaneously in one big roll. The result will be one complete set of ability scores! You should then be able to quickly pick out each set of 3D6 for each ability by their different visual characteristics. For generating a character using Method Four, you’ll do this twelve times. It makes for quite a “presence attack” – dropping all those dice at once can be startling to someone who isn’t paying attention – in a kitchen, at a convention, or at the back of a game store; it’s fun to do; it speeds up the large amounts of rolling involved in making characters otherwise; and it somewhat justifies your having all those dice!


Selecting your Method

In brief: Method One is the fastest of the four Methods, the most “heroic fantasy,” and the most likely to allow players to choose any of the classes for play. Each Method after that involves incrementally more work and more reliance on luck to determine which classes can be played, with more and more likelihood of a low-fantasy, grim and gritty “realistic” flavor. Which Method you choose to use may have subtle but long-lasting effects on the campaign, its setting, and the player experience. Consider your ideas for the campaign and also what you know about your players as you decide which Method you will use. Test generate some characters in advance using the Method(s) you are seriously considering, to both make sure that the results are what you expect/desire, and to make sure that you are solid on how the Method works before your players need you to explain it to them. Make your decision and then go help your players make some characters and have fun with your campaign!



One thought on “Discourse on the Methods

  1. “Players who generate characters together tend to care more about the group than those who create their characters using who-knows-what methods in secrecy. When players generate characters outside of the group, there can be suspicion: was that 18/76 Strength really rolled legitimately?”

    I can offer a different view, from the perspective of players who come to the activity to get acquainted with other people. Freeform role-playing games like Advanced Dungeons & Dragons are constructed from a social fiction more so than an RPG simulation game because the game play is steeped in abstraction and privileges rulings over rules. And rulings, like opinions do through the process of experience, change as the one making those rulings remains in contact with the players. This is fertile ground for using the activity of play to get acquainted with other players, as I wrote in an article here:


    While I am not disagreeing with your assertion, I would like to add to it. If players bring an attitude of playing together (PHB pg 107 Successful Adventures) and combine it with individual player skill, something modern gamers misnomer as Meta Gaming, then there is no need to have players go through character generation together. I have all players roll their characters and equip them before starting them in a Blue Book game in a one-on-one or two-on-one introductory session.

    No game materials are present at character generation. Player skill infers to me that players already have an idea of proper equipment and, therefore, do not need lists. I also play with new-to-the-hobby players and handing them rulebooks when I am going to be ruling is not only a mixed message but also intimidating. The glue that holds it all together in AD&D 1e is trust in the DM.

    In later editions necessary elements like trust, rulings, player skill, and abstraction get replaced with heaps of rules and optional rules; mountains of warnings concerning rule changes; categorizing and codifying player decisions against skill checks; and 6-second rounds of blow-by-blow accounting. I play AD&D 1e specifically to allow my player’s creativity a place to exist and to fill in the gaps the game purposefully leaves open.

    The contrivance of a party needing a type of class – or players casting lots to see who plays what – is not necessary. Aside from the dictates of the dice, limiting the available choice of class or race based on ability, there really needs not be a player conspiracy.

    I usually have players roll up 2 characters by casting 3D6 in a row, keeping characters with a minimum score of 15 in at least one ability. Of the two characters thus generated, the player chooses which one they want, usually choosing the one that best suits the class they prefer rather than the one with the highest statistical average on paper. I am sure this 15 minimum is mentioned in one of the game books.

    Like I wrote, I am just offering another perspective that meshes great with AD&D 1e. I am not disagreeing with you, which would be pointless based on the fact we operate “as one Dungeon Master equal to another.” (DMG pg 7).

Leave a Reply, all comments must be approved to show

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Discover more from The Evil Dungeon Master

Subscribe now to keep reading and get access to the full archive.

Continue reading