I have been pondering a new design which would be focused on supporting adventurer-scientists, from victorian naturalists to pulp archaeologists, modern tenure track scientists or post-apocalyptic holders of knowledge.  In any case, the central mechanic I was thinking about is one where each individual in a conflict (including a GM figure for the world) defines their own stakes.  In the stakes setting games I have seen to date, the participants have identified what they wish to _gain_ if they are successful.  This mechanic would flip it around and ask what each participant is willing to loose in order to be successful.

Now, there are a great many bugs in the central idea and it would require some deep thought.  Does anyone know of any other games out there where players define what they are willing to lose during the stakes-setting portion of conflict resolution?

I don't really know of any games that focus on what you're willing to loose to succeed, apart from stuff like Dogs in the Vineyard where you can choose to escalate a conflict in which you seem to be losing, so you can still win but have to pay steeper and steeper costs if you do end up losing. Since I've never given DitV a read-through this is just my impression of how it works based on what other people have told me.

I do think that if you're going this route for resolution (which could turn out to be incredibly rewarding in play), it seems like the only viable option is to have it work along the lines of
A: Loose the first round of conflict
B: Escalate
C1: Win conflict
C2: Lose the conflict and lose more stuff than if you'd lost in step A.

It'd be interesting to skip the above though, just to see how a system in which you simply set two losing-stakes against each other, though. Any thoughts on how you're going to handle the inherent difference in value between the stakes you're likely to be getting? What I'm getting to is: If Character A offers "If I lose this conflict, I don't get the girl" and Character B offers "If I lose this conflict, I die", it seems like Character B has a lot more to lose, and hence should get some kind of advantage.

I was considering that this system would need very strongly defined resources.  The severity of different bets would have to be tracked in a numerical fashion. 

Minor stakes might simply be "I spend an hour of minor effort" or "I am out of breath" which might have a level 1 stake.  "Spend a day of exhaustive effort", "Sprain" as a level 2.    This may be tied into certain traits, so that someone could loose set amounts of "Professional Reputation" or "Funding" traits.

Both the loser and the winner suffer their stakes.  Usually this means that the victors (with the high stakes) will suffer more to achieve their procedural goals then those goals are actually worth.  There would have to be some kind of balancing measure and some kind of method of recovery, but I think it could be an intriguing idea.

Anyone else see similar approaches in existing games? 

Vincent Baker's Poison'd does something similar: fights have three levels of escalation, based on what kind of fight it is (e.g., swords, pistols, broadsides, ship pursuit). If you're losing after round 1, you can opt to take round 1 consequences (usually mild) and lose. Or you can escalate to round 2. Round 2 consequences are always worse if you lose. Or you can escalate to round 3, which is terrible if you lose.

I suppose the key factor is that all of Vincent's designs (that I have seen to date) assume that the chance of victory was variable yet the cost of victory is zero.  By contrast, the approach I am suggesting is that the chance of victory is 100, if you are willing to pay the cost.  \

Any notable stake-setting games outside of Vincent's?

It could be said that in every functional conflict resolution system that use "stakes", they are both "winning stakes" and "losing stakes".

Example: in dogs in the vineyard, or in primetime adventures, let's say that my character is in conflict with a NPC to save the life of his wife.

Now, if he win, she is saved... but if he lose, she is dead!  So, it was a "winning" stake... or a "losing" stake?

I suggest that if your "stakes" are not both winning and losing ones, they are not stakes at all.  From what I get of your intention, I think you are talking about a resource cost for the character to get into the conflict and to increase his chances to win. It's difficult to say, without some other information.

By the way, the "game development" subforum rules state: "Some available document is absolutely required, preferably through a link. If I don't see that, then your thread gets bumped to the Inactive File, do not pass Go, et cetera.", so if you don't have some document to link to, probably you should ask Ron to move this thread in Actual Play adding some actual play account (maybe what gave you the idea of this design?)

1) Traditionally, what I have seen is that participants in a conflict each define what happens if they win and usually there is no cost associated with victory due to the element of chance.  By contrast, this approach is resource expenditure rather then fortune based and consists of defining exactly how much each participant will suffer during the conflict.  If D&D style conflicts were run in this way, each participant in a conflict could bid with hitpoints.  The participant who chooses to loose the most hitpoints would win a procedural victory and rescue the princess, while the one who bids for less damage would fail on their procedural attempt.

Example 1:
Goal A: Rescue the Princess
Goal B: Repel the invaders
Bid A1: 5 HP
Bid B1: 7 HP
Bid A2: 15 HP

If side B doesn't bid higher, then side A rescues the princess at a cost of 15HP.  Side B only pays 7HP but fails to repel the invaders.

2) You are quite right concerning the subforum rules.  I had read them and I was in a bit of a quandary.  Firstly, this stemmed from discussions of theory and was completely unrelated to actual play.  Secondly, this was a relatively preliminary design idea which I was trying to start researching.  I could have thrown in a few lines on my blog and linked to that as a "document", but that seemed counter to the spirit of the subforum rules.  That said, I could compile my explanation into a brief document and provide a link should Ron wish.  Thanks to the discussion thus far in this thread I have fleshed out my idea a bit more and it might be feasible to write it up. 

Jason,
  With Great Power has a mechanic where you can get more cards by putting one of your super powers at risk.
  I haven't had a chance to play it yet, but everyone I know who has played it, loves it.