From the get-go, "Knight Medieval" strives to set itself apart from other RPGs by having little to no numbers. To me, it feels the game straddles the line between classic RPGs (D&D, Pathfinder) and story games that focus on plot and inter-character relation (i.e. DramaSystem) and less on the numbers.
What I love the most about Knight Medieval is the little to no prep time needed for a player to make a character. You make three simple choices: Gender, knighthood, and what to buy with your starting 100 gold. (Optionally, you can write down what you bought on a blank sheet of paper--no character sheet required). That's it. The Gamemaster is given many hooks based on the knighthoods the players choose if they're at a loss of ideas. No need for GMs to memorize/familiarize the rules. No need to wade through pages for numbers, how to get derived statistics, or rulings.
Speaking of rulings, the heart of Knight Medieval is the diceless Logic System (dLS) that allows the game to have such an easy character creation and game play. The GM's logic on what the players want to do takes the place of dice rolling. Now this can be both good and bad--I can see players and GMs alike trying to bend this gray area to their advantage. But RPGs are a cooperative experience, and all participating should be there to have fun. If the GM makes the players fail everything they try to do, or if a player argues why he should succeed on anything he attempts, they're simply looking at it the wrong way. If the GM is unsure if something should happen, there's nothing to stop him from rolling a dice or flipping a coin to determine the outcome.
An example is a knight trying to scale a sheer wall. He can't climb up a flat, vertical surface, but if he bought a grappling hook in character creation this increases his chance of climbing, maybe the GM rewarding for his forethought and making him succeed. But what if he's trying climb the wall in order to sneak in undetected? That's when the gray area really shows up, and once again, can be determined with a dice roll.
Just by reading the Chivalric Code and the list of Knighthoods, I can see the developers really pushed themselves to cover as much ground as possible. The Code is awesome for its simplicity and flexibility--altering the Code to reflect another universe or one of your own making is an easy task. I can see the Jedi or samurai codes already, and makes original world building easy. This Code also unifies the knights so that there's a basis for each, yet each Knighthood provides enough variety to make sure not every player is playing a Paladin or White Knight (The knighthoods Bounty Hunter, Enslaved, Spy, and the Black Void come to mind). But Knight Medieval immediately turns the knightly code on itself. The Wizard Knighthood in this preview is a fine example. Under this knighthood, you are loyal to a lord and a wizard who gives you a task to complete, which he pays you for doing. The task can range from collecting herbs and cleaning his laboratory to dishonoring the church and sacrificing children. Obviously these last two actions contradict the teachings of the Chivalric Code and the Church. But what if you combined the Wizard and Stronghold knighthoods, and you have men-at-arms to pay to keep their loyalty? The tasks the wizard gives you helps pay for the men-at-arms. That's when drama and complications ensue and you have an interesting session/campaign before you that doesn't hinge on rolling natural 20's.
Overall, Knight Medieval looks to have a lot of promise. Personally I would've liked to have 2-3 more Knighthoods to get a better sense of variety, and a bit more explanation on the diceless Logic System. Despite this being a preview, I feel like I already understand the gist of the game, which is awesome in a preview. Getting rid of the dice and numbers is a bold move in a hobby based on dice and numbers, but I think it's a healthy step in a more abstract style of play that's easy to approach for newcomers and veterans. I'm looking forward to the full release and highly recommend looking over the preview.