I started a thread about narrative gameplay in the actual play section. Here I'd like to get some feedback on the Tefr system I've been using to do that. The system itself is a modified percentile system, not unlike the old Chaosium offerings, and does not necessarily have to be used for narrative based play, though it is written to aid it.

The system itself is in the fantasy genre, couched firmly in a world which provides it with a logic to religion, history and magic. The premise for player characters is that they have a gods curse, which sets them apart from ordinary folk and that legally they must be registered with the Kerowan Guild to get work.

Imagine you are in a strange land,
a world steeped in history
and shaped by ancient strife,
a place where rumours of magic and
mythical creatures have a ring of truth to them.
You'd rather not get involved
in such matters; you want to stay
home with your family and friends.
But you have a problem;
they say there is something
wrong with you.
You have a god's curse upon you,
which makes you different:
stronger, harder, scarier.
Your family and friends no
longer want you near them.
Some would even want to harm you.
If you don't want to become
an outcast, your only choice is to
join the Kerowan Guild and follow
the only profession allowed to
those cursed by the gods:
become a sword for hire;
expendable and despised,
but better the jackal
than the hare.

The whole rule system is more or less complete, though the version of the basic system: Prelude to Rhapsody is undergoing a final edit. The more in-depth supplement, Knowledge is Power has not been edited yet.

Specifically in the Prelude to Rhapsody book, I'm looking for feedback on the way that the gods' gifts work within character generation -it's been significantly revised and I need to know if it is understandable and workable. I would also appreciate any thoughts on the combat system, particularly with regard to severe wounds -should there be mortal wounds in there? Perhaps I need some system that shifts up a gear triggered by the amount of damage. -this section has also been revised in this edit, and I need to know it works well.

I'm not looking to change the whole system, it's too late for that, so if comments could be constrained to improvements/simplification that would be helpful.

-It is a complete system, so useful feedback on the whole is also welcome.

http://www.tefr.com/System.html

Firstly, I've got to hand it to you that you've gone farther than I have with any of my designs-that takes a lot of devoted effort. Secondly, you obviously have a pretty firm vision of the world that you want to portray, what with the races, the magic system, and the nationalities described in the Ebook. I do have some questions and comments though.

Firstly, and this might just be a pet peeve of mine is that character creation seems as if it teeters back and forth between completely random and player driven (and not in a good way). What I mean by this is that a lot of your game seems to focus on having players portray characters that fit into the game world and that seem "real", which is a noble goal. But throughout character creation you have various pages of tables on which you role for random professions, backgrounds, and stats, nearly all of which have the proviso "if you don't like what you see, roll again". The problem with this is that vaguely-directed randomness doesn't gain the benefit of completely player-driven character creation (you don't get to play the exact type of character that you want to play), and it doesn't have the benefits of completely random character creation (you don't get the sense of a character who is a product of the game world, rather than design).

There are some character generation systems that merge the two together well (notably systems which use Lifepaths like The Burning Wheel and Traveler) but assuming that you want to stick with the same general character generation system that you have right now I would recommend either:

Mechanically balancing the various background, nationalities, professions and just having players choose all of these things for their characters, using a point-buy system for attributes. This has the benefit of preventing completely nonsensical results (like the ugly, low charisma Beastman courtesan who nonetheless has considerable wealth gained from his profession), and making sure everyone is on an even keel play wise but sacrifices some level of verisimilitude by making characters a product of design rather than chance.

or

Just making everything (or nearly everything) random. Perhaps allow a player to choose one background, and grant them bonus points to add to their attributes after they're rolled, but having everything else be random. Additional backgrounds, base stats, etc etc could all be random. For example:

Roll 2d6+5 for each stat in order plus 30 points to distribute amongst the 7 stats as desired. No more than 10 points to any single score. Choose one Background (the one your character currently fits into) then roll for one additional background (the one from your character's past that influenced them the most).

Just don't do things half-way and say "roll these things randomly, but not if you don't like the way they come out". Either give the players a choice in that aspect of their character, or don't.

Secondly, it seems as if there's some confusion in the text over the tone of the game. The history of Tefr makes it seem as if the game is supposed to be set in a fairly serious epic fantasy setting complete with a conflict between a Satanesque figure responsible for bringing death and pain into the world and the forces of an absent creator god and his progeny. Most of the "flavor text" at the beginning of each chapter seems to bear this home as well. The problem with this is that some of the rules included in the game might detract from this serious tone. For example, looking at the "Insanity table" (which, by the way seems unnecessary-why not just have the player or the GM choose a type of insanity that fits the situation or that seems dramatically interesting?) one of the possible results is "Believes that a supreme being has spoken to them and told them to prepare the world for the coming of the great spoggin" while another is "Believes that they are a badger". Neither of these is likely to be taken as anything but comedic, either by the character who has been afflicted or by the other players at the table. For a certain style of slapstick, gonzo gaming, such random results work really well but for a game system that intends to tell serious stories about cursed/blessed individuals sent to do the grunt work of a world that hates and fears them it can only detract from the atmosphere.

Let me put it this way: can you imagine Conan the Barbarian, Aragorn, Rand Al'Thor, or Elric, in any medium getting hit on the head and believing that he's a servant of a divinity named "Spoggin" ? Or that he's a badger? Because a lot of players are going to look at your setting at first and probably expect to be able to play characters, that if not quite Conan or Elric will at least be able to do many of the same things: accomplish great deeds, get rich or die trying etc. The fumble table has the same problem-while some fumbles on the table are believable, others strain credulity like the 65-69 natural weapon fumble result of "strikes an ally if nearby, or self" so roughly 5% of the time when a Wolf fumbles it will end up biting either a packmate or itself (presumably it's own tail) ? I'd recommend looking over your fumble tables again and cleaning them out of results that seem nonsensical or that might contribute nothing more to the game than a chance for some cheap slapstick humor.

I really like the idea for your magic system (combining runes together to achieve various effects). It's interesting, fits the flavor of the setting (though I'd like to see the magical language incorporated into the setting's history more), and looks interesting in play. One possible concern that I'd have is that the effects you list seem a bit too cut and dry for a magic system that seems like it should be a good deal more flexible. In many ways, it reminds me of Ars Magica's system of magical grammar, where you choose a noun and a verb chosen from a list to determine what your spell can do such as "Alter Earth" or "Conjure Beast". I'd consider making effects more free form, providing general guidelines for what each symbol adds to a spell or allows you to effect. This cuts down on bookkeeping at the table, allows more flexibility and creativity from spellcasters, and allows you to save space in any hard-copy book you might want to print.

I'd also like to see rules for Sorcery, along with more information on what differentiates it from Enchanting.

Just some general ideas-I don't want you to feel as if I'm beating up on your design, as it's far farther than I've ever gotten with one of mine and looks like a very interesting setting and rules structure at it's core.

John, thanks for going through and reading the system so thoroughly. I know it's a bit of a behemoth.

Your thoughts are very useful to me as I have been working in a bit of a developmental vacuum on this for the last couple of years.

I Agree with you, to an extent, on the character creation; it is a little woolly in some ways. It has vacillated back and forth between very flexible and very rigid in it's various developmental phases, which has left it's mark.
It is supposed to work as a simple: roll three times on each thing, then make a choice of one of each and assemble the character, (perhaps I need to go through and kill the if you don't like what you see provisos).
I've always held the belief that random backgrounds can surprise players into playing something new, rather the stereotypes –perhaps I should trust their judgement more– I've attempted to balance the actual backgrounds so that something like peasant will give a lot of skills while a fighter will only have a few specialist skills.

I do like your suggestion of rolling 2D6+5 and having 30 points to play with, I can see the merits of that. Not least that it will cut down on character generation overhead. I think I may well implement that.

Yes, the tone of the text -it is the third revision, but it still retains a lot of the text from the second revision, which was much more flippant. I've just been revising the insanity table specifically for the reasons you raised. That said, I've never played any roleplaying system that didn't occasionally have the odd humorous element, and some of the best can be found presented perfectly straight-faced in the text. I'll need to give the fumbles a bit of thought, I'm trying to solve bigger issues with the criticals and serious wounds at the moment. I want to eliminate 'sudden death' options in the serious wounds, and just rely on actual HP damage to do the job.
As a rule, I've found that players like the threat of the world background, the seriousness of the underlying risk, but take their laughs while they can  -why else did Tolkien have hobbits?

The Enchanting magic did start out as a completely freeform grammar based system, but the players just couldn't get the hang of that, so I ended up writing out the effects for every combination, it's bulky but works (no messing about). It's all in the Knowledge is Power book, including the full system for Sorcery.
I'd love to have got Enchanting working in a streamlined fully flexible language (actually the the language of the gods and elementals is based on the language of the Enchanting symbols -Netroalic). Perhaps something to re-visit for the next edition.

Difference between Enchanting and Sorcery: Enchanting is done by getting elemental spirits to do the magic (hence the need to communicate with them), Sorcery is done in the head.

Cheers

Simon