I have in mind a 19th century consulting detective game – players stalk the mist-shrouded streets of London with naught but a violin and a sturdy friend (with an even sturdier service revolver) by their side, crushing villainous murderers, and saving the reputations of  compromised society gentlewomen.

The mystery itself would run along the lines of Inspectres, with the players contributing ideas and coming up with the details of the plots and their various twists, clues, red-herrings, and denouments by themselves. Play will be on a scene-by-scene basis, with each player taking it in turns to lead on a scene. The lead player chooses the location and general purpose of their character in that scene (search for a clue, spy on a suspect, steal some vital piece of evidence), but the aim of all scenes is primarily to gain a Case Point – once they have a certain number of those they can move to the climax and confront the villain.

The central conflict resolution system will resolve around a dice bidding mechanism, leading up to a single roll in each scene that determines whether or not a Case Point is gained. Both players and GM will have a pool of dice. In each scene the players will narrate their characters actions, but success of failure is entirely at the whim of whoever has the last bid at that point in the narration. So say player A is leading on the scene where his character – Doctor Wilson – visits a dockside opium den to try and find the origin of the mysterious jade opium pipe. He starts the scene in control of success/failure. As he steps into the den the GM has him jumped:

- GM: “You are immediately set upon by a trio of unpleasant looking lascars, who surround you and grab your arms.”
- Player A has control of success/failure at this point, and he decides that the thugs will not get him: “They try to hold me, but I fall back heavily against the lascar standing behind me, breaking his nose - and then I spin away pulling my revolver out of my pocket.”
- The GM wants the thugs to succeed so he takes control by laying down 1 die from his pool. He can now decide whether any action succeeds or fails. “The lascar you injured steps back roaring wildly, but before you can point your gun at anyone, another has grabbed you by the wrist, and twisting your arm savagely forces you to drop it to the ground.”
- Player A doesn’t like the sound of this and decides to take back control, laying 2 dice in order to outbid the GM. “They got my gun by I didn’t spend three years with the pathans on the Northwest Frontier without learning a thing or two about street fighting. I knee the lascar standing next to me hard in the groin and twist my arm back, breaking free of his grip.”

The player who has doesn’t have control can call for a roll at any point.  When that happens both sides roll. Highest individual die wins – highest number of high scores if there is a draw, next highest die if there is still a draw after that. Winner gives the winning die/dice to the loser.

If the player wins he takes a Case Point, and then narrates an end to the scene. No new NPCs can be introduced at this point, but the player should be describing a resolution that gives some interesting openings for further investigation, plot twists, etc, which explain what ‘clue’ has been uncovered (“I escape from the den – something fishy going on there – they clearly don’t want us to know about the Jade Pipe”). If the GM wins then the player gets no Case Point, and in addition the GM can incapacitate him – meaning that the GM gets to choose where that PC starts his next scene (in bed with a bruised head, or in the basement of the opium den, trussed up and awaiting his fate?).

Questions: Any obvious holes in the dice bid mechanism? I think this could make for an interesting back-and-forth in each scene, though it would need some prescriptions on what could/couldn’t be narrated (no GM killing off the PCs the moment he has narrative control), and I think it is fast and requires almost no book-keeping.  The Case Points mean that in a sense the success or failure at the end of the scene is unconnected to anything that happens during the scene itself (Doctor Wilson could let himself get thoroughly trashed by the lascars and still walk out with a Case Point), but I’m hoping that because the players will be constructing the mystery themselves they will engage with creating a story, and use the Case Points as a framework.

This is a sort of "talking stick" or "conch shell" system, where players vie for control of the thing that lets them talk. When the talking stick is basically a metagame rule that has no direct connection to the fiction, you can end up with a game that feels rather hollow in play. You're not even vying for success or failure, just different kinds of authority. This is very similar to the card game "Once Upon a Time," where you steal some sort of "narrational" authority by playing a card and continuing the existing story based on what is on that card. At least that game constrains your narration to the card in some way.

Don't get me wrong. I, too, have toyed with game designs that are essentially very simple talking stick games. I'm just warning you about some of the pitfalls.

I'd argue that success or failure MUST be tied to the fiction in a very strong way. Furthermore, game currency should be tied to the things a player has his character do. These things can't just be random. They need to be modified in strong ways by the fiction's situation. For example, what modifies the die roll that determines if you get a Case Point? If nothing the character does in the fiction modifies this roll, then the character's actions are essentially meaningless, gamewise.

Vincent Baker has talked about this kind of thing, but in different language. I think this is his "right-pointing arrow" stuff. You can find it in his anyway blog and in some podcast interviews with Vincent.

Hi Adam,

Thanks for the feedback. I'd been wondering about this thread, and why it wasn't attracting any responses at all. I finally decided it was a toss-up between (a) a dud idea, and (b) my mis-spelling of 'victorian'.

Either way, I appreciate your comments. Yes this is essentially a duel over narration rights, in which character actions in some senses have very little mechanical effect - it doesn't matter if the PC shoots the lascars, runs away from them, or hurls an innocent widow into their path, the mechanical chances of winning the scene are exactly the same.

However I do think there is still a connection between the fiction and the 'winning' of a scene. One element I didn't explain in my original post - partly because I'm still not sure whether this would work well - is that players would have to tie every dice bid to a character 'aspect' which had to be narrated into play. Every time the player wants to raise his bid he has to invoke an aspect and work it into the fiction. e.g. in the example I gave, Dr Wilson uses his 'Experience on the Northwest Frontier' aspect to raise his bid when wrestling the Lascars.

I think this would tie the outcome of the dice roll more closely to the fiction: to win the roll the player would have to bid lots of dice, and through bidding he would work the character's involvement into the scene.

However since my intention was that the dice-bidding should result in lots of back-and-forth in each scene, and every player would be bidding multiple times, I think this would either mean that players were using the same aspects over and over - which could get boring - or else characters would need to have forbiddingly long lists of aspects.