Decision duels are a feature of David’s Sling, a novel by Marc Stiegler about technology, nuclear suppression and human rationality. They’re used as an organizational means of decision-making, not dissimilar to the double crux—they’re not quite debates or policy meetings or games, but they have elements of all three. This is a description of them as they appear in the novel, so that any useful marrow can be extracted.
Duels are best at resolving problems that seem political but are actually engineering problems.
This means that there are, in principle, crisp answers separate from the human element.
Good for: whether a budget is appropriate, which programs to fund, whether to continue a project or stop it, which avenues of research will be fruitful.
Bad for: who deserves a promotion or a leadership position, what an organization’s public-facing message should be, which solutions are more ethical.
Duels are always between two alternatives, which are stated outright.
Both sides are displayed on a screen for an audience, with each side taking up nearly half.
A grey section is left to run down the middle for third suggestions.
Duels that settle on third suggestions tend to produce the best policies.
In some duels third suggestions are prohibited, especially when the question is vulnerable to being redefined or slipped out of.
At the top of this screen are the words “LET ACCURACY TRIUMPH OVER VICTORY”
Winners are not recorded at the end of a decision duel, but whenever possible both sides are judged based on whether decision that results was the correct one.
Each alternative has a representative, called a slant moderator or, informally, a decision duelist.
Each may receive suggestions from the audience, and decides whether to use them.
Duelists chiefly create text boxes of various colors and draw lines between them.
There are no turns taken, and each duelist acts at their own pace.
At the start both sides are written as statements, and under these statements are a list of assumptions, placed in amber text boxes.
Assumptions can, and in many cases should, be multi-part.
Zooming in on these amber boxes shows an explanation for why the assumption is needed.
Most of the duel consists of the duelists each challenging these assumptions.
Under the assumptions are the opening arguments.
Any overly popular, bumper-stickerish slogans are usually listed first, even when the decision duelists don’t agree with them, just to get them out of the way.
Arguments can be colored in by the opposing side.
Purple means an argument exhibits a clear, labelable fallacy.
Red means the argument has another kind of flaw somewhere.
Zooming in on a red text box shows the opposing side’s explanation of the flaw.
Other colors are possible but unlisted.
Arguments are typically written and then reformulated after the opposite side’s criticism.
Duelists can invoke probabilities, spreadsheets of calculations, multiple iterations on an idea, and any other means of reaching as correct a solution as possible.
Decision Duels
(Crosspost from my more casual blog.)
Decision duels are a feature of David’s Sling, a novel by Marc Stiegler about technology, nuclear suppression and human rationality. They’re used as an organizational means of decision-making, not dissimilar to the double crux—they’re not quite debates or policy meetings or games, but they have elements of all three. This is a description of them as they appear in the novel, so that any useful marrow can be extracted.
Duels are best at resolving problems that seem political but are actually engineering problems.
This means that there are, in principle, crisp answers separate from the human element.
Good for: whether a budget is appropriate, which programs to fund, whether to continue a project or stop it, which avenues of research will be fruitful.
Bad for: who deserves a promotion or a leadership position, what an organization’s public-facing message should be, which solutions are more ethical.
Duels are always between two alternatives, which are stated outright.
Both sides are displayed on a screen for an audience, with each side taking up nearly half.
A grey section is left to run down the middle for third suggestions.
Duels that settle on third suggestions tend to produce the best policies.
In some duels third suggestions are prohibited, especially when the question is vulnerable to being redefined or slipped out of.
At the top of this screen are the words “LET ACCURACY TRIUMPH OVER VICTORY”
Winners are not recorded at the end of a decision duel, but whenever possible both sides are judged based on whether decision that results was the correct one.
Each alternative has a representative, called a slant moderator or, informally, a decision duelist.
Each may receive suggestions from the audience, and decides whether to use them.
Duelists chiefly create text boxes of various colors and draw lines between them.
There are no turns taken, and each duelist acts at their own pace.
At the start both sides are written as statements, and under these statements are a list of assumptions, placed in amber text boxes.
Assumptions can, and in many cases should, be multi-part.
Zooming in on these amber boxes shows an explanation for why the assumption is needed.
Most of the duel consists of the duelists each challenging these assumptions.
Under the assumptions are the opening arguments.
Any overly popular, bumper-stickerish slogans are usually listed first, even when the decision duelists don’t agree with them, just to get them out of the way.
Arguments can be colored in by the opposing side.
Purple means an argument exhibits a clear, labelable fallacy.
Red means the argument has another kind of flaw somewhere.
Zooming in on a red text box shows the opposing side’s explanation of the flaw.
Other colors are possible but unlisted.
Arguments are typically written and then reformulated after the opposite side’s criticism.
Duelists can invoke probabilities, spreadsheets of calculations, multiple iterations on an idea, and any other means of reaching as correct a solution as possible.
The duel continues until one side concedes.