The other catch is trying to maintain consistency in skill levels (basic, intermediate, expert) when some effects are based on physicality, while others might be based on more abstract concepts. I don't want the idea I've seen in some games where different skills have different cost progressions. This whole project is aiming to create a system that's simple, that will blend into the background to allow more focus on adventure storytelling for small groups of players (and more complex storytelling between those groups).
These two elements are going to be the crux of the system, so they need to be right.
But that leads me to thinking about something else...the mechanisms need to inform the setting and the setting needs to inform the mechanisms in a positive feedback loop.
If this game were designed to be a gritty, "realistic" game of medieval war, sacrifice and honour, we might look more strongly at the way characters are injured, and the repercussions of healing (which might take days, weeks or months), as well as status and honour loss (which might leave a permanent stain on one's history). To make the setting more "playable", as well a fantastic and cinematic, we might add spells that reduce healing times from days to minutes (and from weeks/months to mere hours, or overnight at worst).
Some games take the tactic that characters become spectral beings apon their death, able to drift to a healing point for resurrection. This might be viewed like a MMORPG, and works really well if your game includes mythic setting elements such as an underworld, or if the game reflects a setting such as Valhalla (where warriors fight the final battle, and are continually reincarnated to keep up the fight). It makes far less sense to use these sort of rules and effects to mimic low fantasy gritty warfare, or something akin to the Vietnam War.
Basically, if we wanted these characters to be meaningless peons, we probably don't want combat to be overly dramatic because aren't really attached to them...we'd want a quick system where single strikes can tale someone down (then that same player might come back as another nameless horde member).
If we want the characters to be meaningful, with integrated backstory, we could take two tactics. Firstly, we could make combat bloody, dramatic, and climactic (in which case it would be quite dangerous, and a lot of people wouldn't want to take the risk). Or secondly, we could make the combat less lethal through the judicious use of healing and "resurrection" effects...which in turn modifies the nature of the setting and the expected narrative in it.
This is also one of the reasons why I'm incorporating "primary" and "secondary" characters to the game system. If players want to be reckless, they can wade into battle as their secondary character, get some wild swings in, practice their combat skills (or throwing skills with spell packet bags), then throw together a new econdary character with a bunch of templates when they are viciously cut down. If we want the "primary" characters to be epic heroes, then we need a suitably weak rank of secondary characters to compare them against. Perhaps we incorporate into the narrative and setting some kind of element where the primary characters are "infused by the power of the gods", maybe they are "throwbacks descended from a line of ancient immortals", or "wielders of eldritch artefacts". (This last option would actually work really well if I aligned this LARP system with the Goblin Labyrinth setting I developed a few years back).
Once a creative decision like this has been made, it's possible to filter out certain concepts and focus on the ideas that benefit both the setting and the mechanisms of play.
To make this post a bit more practical and grounded, less theoretical and nebulous, let's make a few specific choices and apply them to the game being designed.
This game is more about skirmishes between small groups rather than epic battles.
Being taken down in conflict leaves a character in a limbo state until they either recieve medical attention or the conflict is resolved.
On field medical attention takes a specific amount of time (determined by the skill of the user, the tools they are using, and any extra time they might take to improve success chances).
Characters who are down in a conflict may be deliberately removed from play (effectively a stab in the chest or a slash across the throat), this is a conscious effort and may have status/honour repercussions.
Characters not deliberately removed from play have a random chance of dying, recuperating with injuries, or walking away unscathed (and some kind of "stamina" ability might modify the results of this outcome). Secondary characters are less important in the grand scheme of things, they typically either die or "retire from a life of conflict due to their wounds".
Characters who are down in a conflict may be looted for their possessions, but this may come with a cost (loss of honour if this is ever found out, possibly being hunted down by the looted character or members of their warband/unit/party if the character doesn't survive).
Then we can consider whether characters remember things about the situation that took them out of action. I've played with several key players over the years who will take a character death very personally, and will enact vengeance and retribution against their assailant until they believe justice is done. It tends to be a minority of players who do this, and in an ideal world with mature players you shouldn't really need to design rules to circumvent this, but it's something to be wary of. I've seen the same sort of thing in tabletop games.
Now for a few more back-of-house game mechanisms.
Like an Australian Freeform, we will consider an optimal number of GMs to be one coordinating GM and an additional GM per five players.
Since this is intended to be a smallish game under most circumstances (lets say 20 participants), this would leave 10 active characters (probably split into 2 parties of 5), a separate antagonist party (of 5), 1 coordinating GM, 3 additional GMs (1 per party), and 2 players to take the role of key NPCs and non-combatants who have their own goals/agendas/information.
If we run a whole day event, it would make sense to split the day into morning and afternoon sessions. This way the ten active characters from the morning session can portray the support roles in the afternoon (and vice versa).
A player would not be able to portray the same primary character in the morning and afternoon, but they might be able to portray a primary character in one session and a secondary character during the other. Players who deliberately choose to play members of the antagonist party or key NPCs may earn Destiny Point bonuses...and it might even be mandatory for new players to portray henchmen, assistants, and characters like these during their first session or two as they learn the mechanisms of the game (not 100% sure about this though).
These last few points have reminded me of one of the interesting things that I have encountered in most forms of LARP rules. That is a focus on the out of game logistics. Perhaps it's because the interplay of reality and fantasy mingle closer in a LARP, perhaps it's because the potential for physical injury is greater, perhaps it's because these types of games require more space and are thus more likely to come into contact with the public...whatever the case, LARP rules contain more metagame aspects than most tabletop games. Social contract has been written into this style of play for a long time.