But how many eggs were left out?

I recently spent 566 words of elaborate digressions, studiously not telling you why I think it isn't important how attribute points are allocated.

After that rather irresponsible abuse of verbiage, I thought it might be courteous of me to actually answer the unasked question.

But First, a Digression

The way that I have experienced games with structured character design and attribute point allocation, they generally fall into one of three categories:

• random point allocation which informs how characters are played
• random point allocation with little meaning
• non-random selective allocation

The first scenario is what you find in games like Traveller or Warhammer Fantasy  The random allocation might inform the circumstances of your characters birth, and determine the tone of how they are played.  These things would dynamically determine the shape of the entire story told at your table.

The second scenario -- "with little meaning" -- might read a bit flippant.  But "4d6 drop one, arrange to taste" will tend to average out at pretty comparable characters, and the resulting +1 or +2 on a D20 probably don't mean that much anyway.  These things are great for creating an illusion of fate, great for creating a sort of balance for "fairness" at the table, and a great starting point for organized play.

As for the third scenario, it is great for creating player choice (real or illusory) and great for systems that attempt to tune a wide variety of subsystems and character options for some very fine level of balance.

Of Course it is About Me

The vast majority of games that I've run and played have not had to be concerned with any of these things in the least.  I have the most experience with short one-session games.  Characters start at whatever ability level best suits the scenario.  Following that, I have run many short 3-5 session mini-campaigns.  In those, characters similarly had whatever attributes best suited the scenario.

I've never had to care much for a game's own sense of balance or themes, only what best fits a scenario.  Similarly, I assume that the conjectural players of this game might not care for its starting assumptions either.

Graeme Barber recently had a thread on non-standard starting characters, echoing zero-level characters which I first encountered in AD&D.

I expect people to use a game as a tool set.  If not to chop it up, as least to tune it for whatever their table needs.  My table has most often needed single-use characters that start out more competent and more proficient than those that default rules provide for starting campaign play.

Sometimes, in the case Graeme describes, characters need to be less proficient.

By writing "2dF down the line" I provide a rough outline of expected value ranges, and invite people to tune it to their own needs.

Counting a Few More Eggs

There are very few words that went into this original text, but they were whittled down with a ton of thought and feedback from others.  As short as it is, I tried to define all terms from first principles.

no expected mode of play

I didn't make any assumptions about players or game masters or modes of play.  I considered not including a formal description of the Ref, but in the end decided that as an "old school" styled game it should be there at least minimally.

It intentionally does not decribe how many characters players have, or if the ref's characters are in some way special.  Module S1 expected a bit of troupe play, and having multiple characters was normal part of most of the adolescent power-fantasy groups that I played with as a child.  Ars Magica did an astounding job of building troupe play into the entire structure of their game.

I only hint at it by suggestion and omission.

is everything a character or a clock?

I suggest that the Ref "does not always need to detail setting elements with attributes" it is actually hinting that sometimes they can.  It was pretty revolutionary the first time someone gave a door hit points, and even more clever when someone decided that they could give abstract concepts hit points.  But it wasn't until a later generation of games that designers really began to pontificate on this stuff.

only roll when you feel like it

I make a whimsical statement about when dice should be rolled, but this its self provides huge leeway irreverently trampling on a lot of strongly expressed opinions about when dice should or should not be rolled.  In the "old days" we'd do it to create dramatic tension, or to let players fail and be forced down some alternate path, or just to remind people that they're playing a game.  In the still old but not that old days it became commonly accepted wisdom that any time dice were touched it needed to be somehow spiritually significant.

I don't have strong objections to either position, but I do think that taking a position is outside of the spirit of allowing people to bring their own egg to the recipe.

all the bits and bobs

Classes, species, equipment, chases and special cases.  I've drafted a few in expansions that I'll share, but it isn't an accident that they're absent from the initial draft.

This All Could Have Been Shorter

OSRdF is, on its face, a sketch of something like Traveller and Warhammer.  Old-school, capricious, random.  A little deeper, it is just a sketch.  Something minimal, something similar to what the inestimable Brad Murray calls a Scaffold.  What Paul Czege might call an incitement.

It does not provide any expected mode of play, indeed it leaves as many undefined parameters as possible so that it can potentially branch into many modes of play.

It strips as much out as possible and reduces back to the bare essentials of what I personally consider Old School:  Character defined by six things, roll under them to do something, make the rest up as you go.  From that starting point, it can branch or evolve in any direction.