Yarn is the working title of a home brew game system I’ve been tinkering with for a few years. It is a middle road between crunchy old school and modern narrative approaches. The main features are familiar to those who’ve played a lot of TTRPGs, but the mix is I hope unique.
Philosophy
Yarn will always be free. Content may emerge for which the creators deserve recompense, and anyone is free to use Yarn for any legal purpose.
Yarn isn’t intended to be a universal game engine. Each setting will be unique, and usually the rules will be customized to some extent for that use. Anyone should feel free to customize these rules in any way they wish.
I want Yarn to have a narrative focus while offering concrete tactical options. I want it to use even polyhedral dice, d2 through d12, for more than damage. I want to avoid linear rolls. I want it to be no more crunchy than 5E overall, while not sinking in to a math and reference quagmire. I want classes to be an option but not a requirement. I want simple systems that are re-used throughout. I want to reject the quick turn, and let each player’s turn to be a mini-game. I want to limit player rolls to one per turn. I want the story to be in words. and the game part to be how you use your resources, especially steps and any margin you can generate from your roll. I want turns to offer modes of action that allow specialists to do more.
Yarn 0.1
Volume I of the Favored is done. Yarn held up well, and I learned a lot. We ran it every other week for the most part, for about 10 months. 18 adventures with no rules modifications.
Yarn 0.2
Now that we’re wrapping my campaign in our shared Traveller setting, I’m shifting to Symbaroum using Yarn rules.
This file is the second major revision, learning from the mass experiment that was the Favored. This is currently being tested.
Update: This is Yarn as of July 2025. Bookmarked and ready for play. I will give the players a few options for genre and setting, or let them use the Imagining from the rules.
Update for August 2025: Mike suggested I consider a published setting for the Yarn 0.2 test. I plan to locate the game in Symbaroum, from Free League . This is the opposite of the Favored, because the fantasy tropes are well-established, albeit with a Swedish spin on them, while the Favored was a home-brew in system and setting. I will wait on Mistrathe Manor for now.
Update for December 2025: V 0.24 Two sessions in to Symbaroum. Added two rules: another character can offer a help die to your feat if they forego their feat for the round; and players can “buy back” one trouble per session each from the narrator by playing their drawbacks in interesting and entertaining ways.
Symbaroum is turning out to be a good test for Yarn. It’s been tough, as I’ve been relying on a translation system to use their game info and stat blocks. When I use it for Mistrathe Manor, it will be an easier fit, as MM has used both our Short and Core systems, and Yarn is their successor.
Update for March 2025: V 0.25 Changed trouble mechanic. Now, the narrator only gets a trouble if you use two or more fortune for a roll, or if you have two or more un-kept ones on the dice.
Update for April 2026: Yarn 0.27 has a few minor tweaks. I am beginning to compile notes and ideas for a third revision. I plan to freeze at 0.29, and continue with that revision until Symbaroum is concluded. Yarn 0.3 will be a return to original source material, Mistrathe Manor in person and possibly Eldaraen on Foundry.
Yarn 0.3
Some notes on Yarn 0.3: I think a finer scale may be preferable to zones, so I am adopting the pace from the Core system (2.5 ft or .75 m). It roots the action in what we physically understand as actual people: “I take four steps towards the door.” I also think leaning in to pairs of attributes will help loosen up turns, not just for getting dice, but for steps; a turn has as many steps as both base ranks added together. As with Yarn 0.2, you have a turn type (now called focus), but rather than a single attribute, the focuses are named. You take a fight turn, or a shoot turn, or a search turn, or a persuade turn, etc… and each focus has the two associated attributes, as well as a list of aligned actions. A class may also offer a unique focus, but this is still based on two attributes. If you do something during your phase that is mismatched to your focus, it costs double steps, and there is a cap equal to the mismatched base, limiting the steps available for those activities.
For example, if your focus is searching, you’d use awareness and intelligence, and if you hit somebody during your turn, that would be mismatched to your focus. It would cost double for each step required for the attack, and you couldn’t get more steps for attacks than your strength and dexterity ranks added together. This is to prevent a really smart but very feeble character from taking mental turn and thus being able to move at average speed, or a strong but low intelligence character suddenly have enough steps to solve a problem.
Due to the more granular scale, actions will cost more steps. Facing may be a thing, especially in a melee fight. Paces are intuitive in TotM, but by moving to this scale, if we go with tactical maps, the option to use hexes is there. On the occasions we do this, we will use 1 pace to the hex.
Also, with the pace as the standard, reach will be finer tuned. Your footprint is the full hex you occupy. So reach 0 is to yourself only, or someone sitting on your shoulders perhaps. Reach 1 is adjacent, for grappling, dancing, elbows and knees, standing shoulder to shoulder, being in a crowded elevator. Reach two hits someone who is one arm’s length away, with a punch or a knife. Reach 3 is for long range kicks and short swords. Reach 4 is for long swords. Long pole arms could be 6 or higher. Ranges longer than that would be for long chains and lassos and the like, but by then you’re bordering on a ranged attack.
To make a playable space fit on a table, I’m using 1/2 inch hexes, and smaller pawns, rather than minis. Because of the scale of hexes, this will actually be the same scale as 1 inch hexes that represent 5 ft. This may seem potentially cramped, but characters are not often adjacent it turns out, so there’s room to pick up pieces. Zones of variable size can be used for large scale maps, and for purely TotM, paces will be used, when distance is the primary consideration.
We all like miniatures, but here the map and markers will be tools, just like dice and character sheets, and the story is still told in words. It’s going to increase in complexity, though it will have a lot of return for that crunch, and once learned, will require very little referring to the rules.
A player’s turn will be longer. I am looking at ways to reduce latency. It’s ok if it’s detailed, but I don’t want it to be math-heavy, and I want to stick to simple re-iterative ideas like sigma and margin. If in testing a long player turn exceeds five minutes, I’ll know I’m on the wrong track.
Update: I am working on a zoom-able scale that will permit dual maps, so that one scale can be zones while the other a hex grid zoomed in on one or more zones. That way you can have a fire fight and a melee fight at the same time. A zone is as many steps to enter as the paces it is across. I think standard size could be 5 paces across. An average person could then move two zones per turn, or one zone and attack. A larger scale is possible, 10 paces across, for those long range fire fights.
Update:
Unified Roll
For Yarn 0.3 I am implementing a unified roll for your entire turn. You roll dice and chooses your kept dice as before, but if you are actively opposed or supported by non-player forces , you leave the unkept dice as rolled, and the process is extended into a mini-game of dice selection and denial.
From your unkept dice the narrator chooses two dice for the Adversary (the forces active against you that turn). As they choose each die, it is flipped to the opposite face, inverting the value. On a d8, a 7 becomes a 2, etc…
To keep you from choosing a terrible focus just to sabotage the Adversary’s dice, eligible dice must have at least as many faces as the Adversary’s level. A level 7 adversary can’t choose a d6 or smaller. When they go to choose a die, if there isn’t an eligible die available, they get a ‘ghost die’ instead that is a flat value equal to their level. If no unkept dice are eligible, the Adversary gets two ghost dice. Dice that were not chosen, including ineligible dice, are left as rolled, unkept.
This gives you a decision when you have a high roll on an eligible die. You could keep it for your feat, or leave it unkept, knowing it will invert to a low die for the Adversary’s feat. Ghost dice create a special challenge. A lower level Adversary won’t have big ghost dice, and so it’s easier to exclude them from getting access to unkept dice, but a high level Adversary could be getting 10 on each ghost die, and this could be incentive to leave eligible dice to reduce their margin.
The combined non-player forces active on your behalf on your turn are the Ally. The narrator chooses the Ally’s dice from whatever is left after the Adversary has chosen. These are not inverted, and there are no ghost dice for the Ally; however, if not enough unkept dice remain, you can donate from your kept dice enough to give them up to two kept dice. Donated die rolls can’t exceed the Ally’s level, but unkept dice are not restricted by level for the Ally.
A high level Adversary won’t ‘see’ smaller size dice, and so you may be able to ‘launder’ dice to your Ally. For example, if the Adversary is level 9, then d8s are not eligible, even if they rolled 8. You could be certain those dice made it through to be available for the Ally. So the Adversary gets a high roll for their feat, but if you and your Ally both succeed it may have been worth it.
In some cases, the narrator may see an eligible die with a high number showing, and claim it for their feat, in spite of the inversion, just to keep that die out of the Ally’s hands.
Because your roll may be needed for NPCs that are active during your turn, with the new rules you will now declare your intended feat, as well as the target(s) if any, and roll for it before the events part of your turn. Thus, you will have margin and steps in hand, and can commit as many of either or both as you want to increase your initiative for the turn. This also means you will know before you begin moving whether or not you succeeded. You have steps and margin in hand as you take your active phase.
You can abort a feat right before your phase, if you failed your roll for example, but there is an incentive not to. If you abort your feat, you lose half your steps. Thus you may want to go ahead and fail so that you can still have most of your steps for movement and routine actions.
This will reduce the decisions the narrator has to make substantially once dice are chosen, and simplify NPCs a lot, but with the dice choices, the early turn will take longer, so I have moved the popcorn choice to right before your events. Thus, you choose the next player, who gets to see your margin and the initiative, and so they can start preparing for their turn, choosing a focus and feat, determining how many dice they get, while you are moving and acting and being targeted, etc… during your events. Thus, when you are done they are ready to declare and roll. This should reduce latency quite a bit.
This mini game seems novel and strategic, and the potential for the entire turn is decided in the process. It will save a lot of time during events, but testing will show if it adds back that much time in the dice selection game. As long as it’s break even with Yarn 0.2’s turn time, I will be happy, because the narrator’s work load will be reduced while keeping the player-facing parts of the game rich and detailed.