You can go two ways with this, depending on how "leaky" you want the mystical elements to be. Is it sunshine, pink bunnies, and technicolour in the Fairy Woods even before you meet a fairy, or is the world mundane with the fantastical elements playing more of a "This can't be real, I don't believe my eyes" kind of role in decisively mundane surroundings?
Fantastic things in mundane places
Simply put, don't try to transition. State the facts of the fantastic elements as your PCs perceive them, and nothing more.
State the fantastic as it happens. The way the facts violate the expectations of the rest of the established setting will do your work for you. Describe it in the same terms, themes, and "colour" that you do ordinary life in the setting, which will ground it in reality. Its unreal elements, narrated matter-of-factly, will actually sound more fantastical for the contrast.
The sweat and grime of the trail is getting to you around noon, when you stop to drink at a stream off the road. Refreshed, you notice a dark cave mouth on the far bank. Noxious fumes drift from it, and you become aware of an uncanny stillness which no birdsong disturbs. Then, an unearthly rumbling sigh issues from the crack in the earth.
(the PCs investigate and disturb something)
An enormous, stinking lizard drags itself from the depths. Smoke billows from its nostrils and its eyes are orbs of flame where a beast should have whites and pupils. It pulls the rest of its carcass free of the cave, revealing tattered wings like those of a bat and a wickedly barbed tail …
Describing the fantastic in mundane terms maintains the possibility of plausible deniability, and plausible deniability of the fantastic can keep your players grounded in the mundane. "Was that a dragon we fought? Well, we'll definitely brag that it was a dragon, but maybe that was just a really big lizard… It was a weird foreign country like [country only two borders away] after all, who knows what kind of barbaric beasts they normally have…"
Of the two options this is really my preference for maintaining my own grounding in the mundane. It just seems to work better, possibly because it forces me to recontextualise the fantastic occurrences in mundane terms, keeping my own "camera perspective" firmly on the mundane side of reality.
Fantastic lands border the mundane ones
If you want a world (or parts of the world) to have entire regions that are not mundane, you need some tricks to indicate when the characters are transitioning between their familiar, reliable world to the lands that are "not in Kansas anymore". Much like a movie/TV character's musical theme can subtly telegraph when they're having a spotlight scene, you want some tools that signal that the surroundings are changing.
You can do this by straight-up telling them, when the altered nature of the places is perceptible. Just extend the matter-of-fact statements about weirdness to the environment. Where there are (e.g.) uncanny, eerie places that the characters would "sense", just say so:
This place is eerie and uncanny. It feels wrong in your bones.
or for (e.g.) a nice-fantastical locale:
There is a lightness to the air that seeps into your skin, and everything feels brighter.
You want your players to know that the place is unusual, so tell them (via addressing their characters) that it is so, in plain language.
For a more subtle signal, you can use certain kinds of language to distinguish the mundane realms and the fantastic realms. A fairy wood might be "fresh", "bright", and devoid of the language that describes the grime of the mundane realms. You can narrate more slowly, in a more relaxed manner, giving it a bit of a dreamlike quality.
You can use different metaphors. A fantastic realm might get narrative descriptions that frame everything in terms of clean edges and elements in motion:
The flaming leaves of the autumn forest burn bright as the sun sets behind the knife edge of the mountains. The stars come out like hard diamonds, and the bright silver of the moon pours over the forest.
A different fantastic realm might get metaphors about swords, blood, and smoke. Another might be bones and death. Whatever metaphors you like, you can divide them up and say "these are for the mundane places, and these are for the fantastic places", and players will pick up on the distinct feel of different places, even if they can't put their finger on why.
Best Answer
The Riddle of Steel was designed by Jake Norwood, who practices historical European combat styles, studied and taught at the Association for Renaissance Martial Arts (ARMA), and is a founder of the Historical European Martial Arts (HEMA) Alliance. He designed TRoS to have an authentic feel similar to fights from his practical experience, with the same kind of micro-tactical gambles, decision points, flow, and failure modes.
Here’s a quick rundown of how it fits your needs:
Yes: Wounds are individual and have immediate location- and attack-type-specific effects on your continued ability to fight.
Yes: A broad choice of common and weapon-specific manoeuvres, attacks are always aimed at a part of the body.
Yes: The rules for multiple opponents is a trivial extension of the basic combat mechanic. Your weapon skill is represented by a pool of dice divided between attack and defense – when attacking and defending against multiple opponents, you simply have to split your pool further to attack and defend against each opponent. You need to be very good (that is, have a large pool) to expect to successfully defend against more than one opponent, let alone also attack them all. Outnumbering an opponent can be a huge advantage, but skilled fighters can take on a number of much less skilled opponents—if they fight smart and make no fatal mistakes.
Yes: Wounds take weeks and months to recover, much longer (if at all) without hospitalisation.
Yes: Sabres are good at thrusting, cutting, and parrying, making them evenly matched. A knife is difficult to parry with while a hammer is great at striking and has a reach advantage as well, making it unlikely for the knife-holder to succeed in the opposed roll to parry the swing. However to reverse is also true: a hammer is not great (but not bad) for blocks or parries and a knife is good at thrusting, so the hammer-holder needs to pair the hammer with a shield or end the fight quickly before a fast knife can get inside their reach.
Weapon type also directly determines injury types, and size modifies injury severity.
And here’s a more detailed account of how the system works relative to your needs:
The system is remarkably simple, easy to learn, but hard to master. It rewards competent tacticians who know their character’s abilities over players who assume that better numbers necessarily mean they’ll win.
Your Weapon Choice and How You Use It Matters
Different weapons are stronger and weaker on defense and offense, and for doing different manoeuvres. It represents this elegantly by making your combat roll be against a target number for your weapon with a weapon-specific modifier for the manoeuvre, and comparing how well you rolled against how well your opponent rolled against their own weapon/manoeuvre. For example, hammers make a high success result on a swing easy, making them effective against anything that is poor or even just average at blocking or parrying. As a result, a knife wielder parrying against a warhammer needs a very high target number, while the hammerer will have an easy target number for the strike: the high TN for the knife/parry and low TN for the hammer/swing means the swing will likely be more successful than the parry, beating the margin of success of the parry and smashing through it.
Your skill with a weapon gives you a number of dice in a pool. During one “exchange” (which is defined as about a second, or “roughly how much time it takes to attack with a weapon and begin to recover”) you choose how many dice to commit to attacking and how many to hold for defenses against any incoming attacks. On top of that decision, your pool of dice only refreshes every second exchange, so you have to choose how many dice to use in your first attack/defence and how many to hold back for the second clash—assuming you're both still standing for the second exchange, because one good hit can end a fight. Dice on the attack and defense rolls that come up equal or higher than the target number that player needs are counted as successes: if the attacker gets more successes, their manoeuvre succeeds.
The more difficult or risky manoeuvres subtract dice from your pool before you roll (this is the aforementioned manoeuvre modifier); which manoeuvres count as difficult or risky, and how many dice they cost, depends on the weapon. Most basic manoeuvres with most weapons (such as thrusts, swings, blocks, and parries) have no “activation cost” to subtract from your dice pool. More complex manoeuvres, or ones that aren’t really suited to the weapon subtract from your dice pool. (For example, slashing with a pole-axe has no cost, but attempting a hook costs 1 die, and riposting with a pole-axe costs a hefty 3 dice.) This means that showy or expert-level moves only have much chance of success when executed by a PC with dice to spare in their pool—i.e., PCs who are very skilled with their chosen weapon. Successful manoeuvres also affect the state of combat apart from injuries dealt: for example, a good feint lowers the opponent’s pool next exchange, and a very successful riposte can disarm your opponent. A character starts with all the basic manoeuvres for their weapon, and can learn more advanced moves as their skill level increases.
Injuries Hurt a Lot
Injuries are dealt by location aimed for and the difference between attack and defence rolls, then reduced in severity by the armour on that location (modified by how effective a weapon is against the armour, of course). This is looked up on a table, which isn't fast, but isn't slow either because they're not big tables. The injury math is simple enough if not fast, but the speed of the math and tables is almost irrelevant because most fights are decided after the first injury or two.
All wounds either reduce your combat pool briefly (“shock”) or for the lifetime of the wound (“pain”), both of which makes you more likely to be lethally injured; give you a “blood loss” result that puts you on a time limit until you fall unconscious (and likely bleed out), or kills you dead. Most results are a combination of those, with the exact combination reflecting the location (e.g. head/neck hits are more severe than a similar strike to the leg) and the nature of the blow (a mace results in less blood but more shock and pain that reduce your pool temporarily, while a sword blow will do slightly less shock and pain but cause more blood loss that will soon drop you unconscious), with greater margins of attack success dealing more severe injuries. Injuries also have special effects such as reduced movement, concussions, broken bones, lost eyes, severed hands, and such, which may or may not ever heal fully. Getting in a fight and risking being hit is not to be done lightly!
Should a combatant be wounded and survive the fight, recovering is slow for anything worse than minor cuts and bruises. The severity of a wound (“pain”, the number of dice it removes from your pool that isn't temporary “shock”) is the minimum number of weeks of rest and medical attention required. A wound's pain rating drops by one if a healing roll is successful, with one roll made each week of rest and care. Multiple wounds recover in parallel, and can leave lasting effects if they were really bad.
Rules Mirror the Choices of Combat
The net effect of all the combat rules is to give the player a lot of simple but meaningful decisions. How many dice to commit and reserve, and what manoeuvres to use given how they and their opponent is armed and armoured – these translate in very direct ways into choices about whether to fight conservatively or go for a desperate gamble, whether to try to alpha-strike at the risk of having no defense if they survive, whether to try to wear them down with nicks and scratches or to hold back waiting for them to make a mistake and leave them open to a decisive blow, whether to defend and run from a mismatch (e.g. unarmed fencer versus a mailed knight) or to see if careful tactics can win the day, whether to fight quick and light or heavy and slow or in between. The basic rules give a lot of freedom to choose how to fight.
The choices the rules ask of players during a fight parallel the choices their PCs would be making at the duelling level: feint or strike, go on the defense or press the attack, and so on. This makes it feel very authentic. And since all the numbers (pool sizes, target numbers) are precalculated during character creation, as soon as those decisions are made it is a simple matter of rolling and comparing dice to a target number on your sheet to resolve the outcome. That makes it feel both tactically detailed and fast-paced. Those same elements of choice also make it a combat system that heavily reflects players' tactical skill rather than mostly relying on the PC’s stats—a smart, experienced player can easily take down an opponent that has better numbers and equipment if the opposing player/GM makes poor tactical choices and simply doesn’t make good use of their advantages. Even a knight can fall to an unarmoured knife-fighter if the knight gets overconfident and leaves themselves open to a killing blow.
Lethal Combat and Quick Chargen
As I said, it's a very lethal system. The game does have a set of rules to make heroes more survivable by adding dice to their pool when they are following their heroic destinies and driving motivations, which is designed to discourage fighting except for when it really “matters” to the character in a personal-plot sense. However, you can easily leave the heroic goals rules unused and just let the combats be life-or-death every time to make it very bloody and gritty. Even with the heroic rules, though, it's still a very gritty and bloody game, where a PC can die from one combat mistake.
On the plus side, characters are fairly quick to make. They consist of a few stats, an occupation-based skill package, and combat skills. You have a few points in each of those areas to customise the numbers; for example, you can spend all your combat skills points on a large dice pool in one weapon style such as Sword and Shield, or you can divide your points to have more styles and be a more flexible fighter but have a smaller combat pool for each weapon style. The most time-consuming part of chargen is precalculating and noting down combat numbers, but that doesn't take very long either and is part of why resolving manoeuvres is so fast.
The only problem with The Riddle of Steel is that it’s out of print. Copies can be had from second-hand dealers online for a range of prices, which is how I got mine. There’s a PDF floating around the internet that is a nearly-illegible scan and not worth your time—sadly there was never an official PDF version. There were a number of official PDF-only supplements that you can’t get legally anymore, but they’re out there if you know where to look.
For more about The Riddle of Steel there is a review and brief discussion at RPGGeek written by a long-time player and one of the current maintainers of trosfans.com. One of the few things still available from the defunct publisher's half-working website is the Riddle of Steel Quickstart (partway down the page, or try this TROSQS.zip direct download), which is a much-simplified but complete game that should give you enough of a feel for the full game. There is a review of the Quickstart rules themselves at RPGNet, and this thread discussing the differences between TRoS and the Quickstart may be helpful.
I've recently learned that the IP holder of the core book has authorised the owner of the trosfans forum to sell the core book PDF. To inquire about acquiring the PDF, send an email to ian [dot] plumb [at] griffingrove [dot] com [dot] au (with the email obfuscation replaced with the appropriate punctuation, of course).