Jared spoke to me about Donjon Krawl last Wednesday, and I finally had a chance to read it over last night. Very interesting stuff. Anyway, some thoughts:

* I love the feel of old-school, wondrous donjon krawling, where you never really know what's coming around the corner.

* Has anyone else toyed with the idea of completely removing the DM, or moving the role to all players? I'm thinking of running a version where everyone comes to the table with a few encounters and ideas. Play starts with the, say, the oldest person at the table as DM, and he continues in that role until someone opts to use successes to drive the story, rather than conserve them for a later action. This might have the effect of players going with the flow of the story and saving successes for a scene where they want strong authorial control. Perhaps the current DM banks the successes he used to assume the DM-hood. As players challenge for control of the story, they need to achieve a number of successes greater than or equal to the current DM's successes. To prevent anyone from hosing the system to build up success dice, a majority vote is required to allow someone to make a skill check in a questionable situation. I can't, for example, keep making skill checks to tie my shoes in order to seize control of the story, unless the rest of the players are happy with that.

(There's an essay in my head about narrativist rules design v. gamist design that talks about writing rules from the POV of enforcing inter-PLAYER v. inter-CHARACTER conduct. Has anyone explored that topic?)

* The system of wealth and spending left me a little dizzy, and it seemed a little too complex for a game that focuses on donjons. I might use an alternative system that works like this: each character has X number of dice in an equipment pool. Whenever a character attempts an action, a player can take Y dice from that pool and use those as a tool to help with his action. For instance, Boldrek the stone troll is trapped behind a caved in passage. His player, Mark, thinks for a moment and decides that Boldrek has a handy shovel in his backpack that adds 2 dice to his attempt to dig his way out. Mark notes that 2 of his equipment dice are now tied up into a shovel, notes that on his character sheet, and makes his check. For the rest of the adventure, Boldrek has 2 fewer dice in his equipment pool but now has a shovel. After the adventure, Boldrek can drop his equipment and restore his dice. Cash is handled in a similar manner, but when you lose dice from that pool they're gone until you find treasure. Weapons and armor would draw from the equipment pool. Philomel the warrior might allocated 6 of his 7 equipment pool dice to a sword, shield, and armor before an adventure even begins. A few other rules might cover things that you can't pull out of your backpack, like armor, shields, big weapons, and siege engines.

* Thoughts on buying stuff: each market has a difficulty rating. To add dice to your backpack, make a test with your money pool against this difficulty. Each success adds 1 die to your backpack. All dice used to make this test are lost from your money pool.

* I love the idea of starting with a blank map and seeing how the adventure progresses.

Anyway, it looks quite interesting, and I'd like to try playing it soon.

- Mearls

I like the idea of just using Provision dice directly as the Rating of a piece of equipment. I wouldn't just allow it to be freely created however, thats a good bit more power than DK provides elsewhere. To be comparable there should still be some roll with failure being he looked through all his stuff and alas no shovel. The number of successes rolled could then tie into the value / ratings of the equipment. There are already rules for ascribing value to mundane gear in the loot section, this would probably be very easy to incorporate. In fact, in play its the sort of thing that most GMs would probably decide to do without explicit rules for it.

The Wealth rules do lead to some oddities. The worst of which is characters trying to give wealth to other characters. You cannot simply allow a die for die transfer because the difficulty of finding loot is directly tied to your current wealth. The obvious cheat is to give all of my current wealth to you then I can find mondo loot on corpses that otherwise I wouldn't find didly.

So transferring wealth between characters must be done with a die roll similiar to looting. Instead of a Loot Value the value is the number of dice I wish to transfer to you. The problem is 1) What does failure represent. When buying from a merchant it means the merchant refuses to deal with you. Obviously that doesn't work for PC to PC transactions. 2) If a Wealth 8 guy gives 3 dice to a Wealth 2 guy and only rolls 1 Success, the Wealth 2 guy goes up to Wealth 3...but what happened to the other 2 dice of Wealth?

Some oddities that could be smoothed out, but in general a small price to pay to never have to scan down equipment lists again.

Mearls,

I think that notion of competing/trading GM roles based on relative successes is a wonderful idea. I also think that the distinctions and subtleties among character vs. player, and how we expect/reward the one in hopes of influencing the other, deserves intensive scrutiny.

Best,
Ron

Mearls sort of beat me to the idea of removing the GM.
Actually my idea was that since the players are so powerful anyway, why not mostly eliminate the idea of "players" and instead have everybody be "GMs?"
Each GM earns "GM points" based on how much they influence the adventure. You can also earn or spend (I didn't work out the details) GM points to modify the Donjon quest- ie. you can add new Steps to the quest, or even change the big bad- "it turns out the main villain wasn't really the one behind everything." I referred to the latter as the Final Fantasy boss effect.
A step can only be passed after making a successful "scene change roll" whose difficulty is based on the number of encounters the GMs have made for that level. After each step is passed, the GM who changed the scene and the GM who made it up earn points.
But to prevent someone from taking over an adventure, a GM cannot do consecutive actions.
Also, a GM can demand another GM for a point if the second GM has made a particularly lame decision. The second GM has a chance to argue that the choice wasn't lame, or change it to make it not lame.
Even stranger (stupider?), I envisioned this kind of campaign being run like a simulation of games like "The Sims"- the pcs (remember, they aren't owned by anyone) exhibit autonomy when the GMs aren't controlling them and will basically use their skills more or less randomly to complete the quest. But in combat or other exciting scenes, the GMs can scramble to control various pcs.

Sorry if this is really, really fragmented and incoherent.

Kenway wrote:
Even stranger (stupider?), I envisioned this kind of campaign being run like a simulation of games like "The Sims"- the pcs (remember, they aren't owned by anyone) exhibit autonomy when the GMs aren't controlling them and will basically use their skills more or less randomly to complete the quest. But in combat or other exciting scenes, the GMs can scramble to control various pcs.

My only comment on this is not that these ideas aren't good ones...but that not EVERY indie game has to be bleeding edge avant garde.

In fact, DK is intended very much to appeal to Old School style players. The "author power" dial is very much regulatable in the game as it is, from none to tremendous.

You can as DM run the game and pretty much let the players be their own DM if you want with the rules as they exist now.

The rule is "winner narrates their success, loser narrates the rest"

Set the difficulty very low and load the players up with successes for them to spend how they like. Then as the "loser" use the "rest" to tie it all together coherently.

Or set the difficulty very high to make the players lose. Then with your couple of successes throw out some "facts" that you want acted on and let them "narrate the rest".

The DM can be everything from traditional DM to traffic cop in DK.

It really doesn't need any more radical changing of player roles than that.

I think the idea of "competing/trading GM roles based on relative successes" is done in Donjon explicitly. Relative success (or failure) determines how much the player gets to narrate.

I think a GM-less design is completely incompatible with the design of this game. (See Fungeon for GM-less dungeon crawling.)

Valamir makes some good points about Wealth, and Mearls makes excellent points, many of which could make the equipment system in Donjon much easier. (I like the idea of a relatively stable equipment pool that you draw from even before the adventure in order to define things you need.)