Back in January I replied to a Reddit post asking about Point Buy vs. Rolling 3d6 for determining ability scores. For the sake of brevity I kept my reply short and targeted toward how dice can model rarity since it had not been covered elsewhere in the post. But now the subject of rolling to create characters is on my mind and I’d like to explore it more in depth here.

Before continuing I do want to say that though I highly prefer dice rolling I’m not trying to be pedantic. If it doesn’t work for you or your table do it differently. We used the 4d6 method in the old days and sometimes fudged if it meant someone would be happier with their character.

For the Players

Today it’s popular to come to the table with a character concept you’ve all ready put some thought into and have some attachment to. The point buy method of ability score generation underscores this sort of play by allowing the player to assign ability scores according to their character concept. With this style of play come certain implicit expectations which should be made explicit in order to understand what kind of game you’re playing and what kind of game rolling ability scores may produce.

Among these implicit expectations seems to be that the player has created a character immune to the common depredations of the game world. Before the game has begun the player has created a hero immune (at least in their own mind) to an ignoble death at the hands of the first monsters encountered. The character may be very difficult to ambush, capable of taking on several enemies single-handedly or able to throw seemingly limitless spells. In brief the story of how awesome the character is has been told in the imagination of the player long before a single die has been cast.

In years past rolling dice to determine ability scores, class and even race was a bit of a foregone conclusion¹. To modern eyes it may seem hopelessly random and it’s nearly certain you won’t have your ability scores perfectly distributed. The experience of character creation is however very different. Systems in which the players roll for scores tend toward faster character generation and in my experience a more meaningful session zero.

A more meaningful session zero is worth a moment of focus. In a point buy character creation game the only interest one player may take in another player’s character creation is whether that other player has optimized in some meaningful way. In a roll system if a player rolls especially high there’s cause of jubilation among the group because a powerful character benefits everyone. Conversely low score rolls are an occasion for sympathy.

Not surprisingly the random rolling player must also come to the table without a preconceived character which has implications for how the resulting game can be played. These games tend toward the dangerous. The players of these games tend also to be less attached to their characters—at least initially. It’s common in such games for life to be fleeting and characters to die. Perhaps not quite as fleeting as grognards like myself reminisce about but certainly talking to that first group of orcs before rolling initiative would be wise.

Many in the hobby both young and old would of course prefer to start off ideal (point buy) rather than random (rolling). Even though a point buy system enshrines the preference in the rules it’s certainly not new as can be seen by the several methods of manipulating score generation in the AD&D DMG². But in my opinion the preference for ideal is a preference for easy.

It’s of course easier to begin with maximized offense, defense or with the ability to access all spells and powers available to your chosen class. But which is better? To perfectly place your scores the exact way everyone else would (optimizing) or to roll a wizard with an 11 Intelligence who has to struggle and grasp for every spell? Who is more epic after 15 levels of play? The character that began as ideal or the one that started off imperfect and had to build themselves up?

Dice (may) Model the World

In the AD&D Dungeon Masters Guide there is a section on dice and probability. It’s pretty dry stuff but it can be summed up. If you roll more dice you are more likely to roll an average number and less likely to roll extreme numbers. That means the very low (3) and the very high (18) are much more rare than the middle-ish (10, let’s say).

This is important because that system (AD&D) also has score minimums for the player to be certain classes and certain races. This effectively allows the dice plus the requirements to determine the rarity of certain classes and races in the game world.

This is one of several ways that old school games achieved “balance”. An AD&D Paladin may be objectively more powerful than an AD&D Fighter. But rolling a Paladin is much more rare.

Again this may not be to your personal preference. The modern RPG tends to be high-fantasy which lends to point buys and an anything goes attitude toward player characters. But have you ever asked what kind of world the game and the dice are modelling?

5E models a world in which anyone can be anything with no restriction. That’s true while being neither good nor bad. However if anyone can be anything then is anyone special or unique? You may argue everyone is special and unique and in the real world I might agree. In a game however, this philosophy robs your decision of any mechanical consequence. In my opinion this philosophy makes for a horribly bland game world.

In a 3d6 system there may be a requirement for a player to roll a Dexterity and Intelligence score of 11 (example pulled purely from my ass imagination) in order to create an Elf. That requirement would tend to inform the players that Elves are uncommon in the world. Additionally it would mean rolling one is a special event that unlocks special advantages. Which it does in D&D/AD&D since elves may be Fighter/Magic-Users.

Since we as game-masters and players can decide which dice method we’d like to use or choose not to roll at all we must ask, “Do we want dice to model our world?”

Certainly we accept that dice are modelling some portion of the world since we allow them to determine success and failure in combat, skill tests and other random outcomes. Shouldn’t we then consider that allowing the dice to model character creation may have some benefit?

Clearly I would argue that letting the dice model character creation has benefits. It can break players out of falling back on what they always play. Prevent players from making character decisions based purely on mechanical advantage; though they may still to some extent. Importantly dice couple with minimum score requirements can be used to distinguish between the common and uncommon in a game world.

Your mileage may of course vary.

Fudge no more

The revelation that dice can be used to model the rarity of class and race options is a recent one for me. While it’s an idea I quite like I’m not one to slavishly follow R.A.W. By way of example an AD&D Paladin is required to have, “…strength of not less than 12, a minimum intelligence of 9, a wisdom of 13 or more, a minimum constitution of 9, and not less than 17 charisma.”³ The 17 for Charisma was a rare to roll even taking into account that we typically rolled 4d6 and allowed the player to place their scores.

Imagine for a moment a game is going to start in a distant rural area with small churches and land rich but cash poor nobility. It would be the rare individual who would have the time and money to train in both martial combat and religion. Additionally the mentors this individual would need might not be present. In this case I might be inclined to leave the stat requirements as is in order to keep Paladins possible but rare.

Now imagine a different game taking place in a large wealthy city where nobles and faiths maintain private security forces and compete for prestige through public works, displays of culture, wealth and power. In this region religious colleges might aggressively recruit and train warriors making Paladins somewhat more commonplace. That could of course be communicated to the players in several ways but a relaxing of the score requirements would allow the random rolling party to increase the availability of the Paladin class.

Ultimately point buy yields an anything, anytime, anywhere sort or game or perhaps it’s more accurate to say that an anything, anytime, anywhere sort or game lends itself to point-buy score generation. This rings true without being either good or bad. Random score generation on the other hand seems to yield or perhaps is a product of systems that embrace randomness as part of the game model.

In either event I find myself preferring the latter to the former.

by Sage Jim

Foot Notes

1. To be perfectly honest we colored outside of the lines a lot in those days. A rules-as-written game is much more a modern idea of how a game should be played
2. Gary Gygax, Generation of Ability Scores: Dungeon Masters Guide (Lake Geneva: TSR Games, 1979), 11.
3. Gary Gygax, The Paladin: AD&D Player’s Handbook (Lake Geneva: TSR Games, 1978), 22.