Shared resources depend on both individual demand and the rules people build together.
Game Theory.
What happens when the best choice depends on what everyone else does.
62 live experiment · 1 plannedA tempting individual move can leave both players worse off.
Who receives the upside—and who bears the loss—can change a decision.
A valuable shared goal can require confidence that others will join.
Each side wants the other to yield, but mutual escalation is costly.
A predictable choice can be exploited even when neither pure choice is best.
A shared convention can help people choose the same place without talking.
A contribution can benefit the whole group while costing the contributor.
An offer can be profitable and still be rejected.
A negotiated split depends on what each side can get without a deal.
Paying the second-highest bid changes the incentive to report your value.
The highest estimate may be the one with the largest upward error.
When buyers cannot observe quality, good products can leave the market.
People can follow earlier actions even when their own clue disagrees.
A shared project succeeds only after enough people contribute.
The first person can grow a shared pie but cannot force a fair return.
Passing grows the pot, but each player can take it first.
Everyone benefits if one person acts, but each hopes someone else will.
Try to predict what others will predict, not a number's intrinsic value.
Two sellers compete for customers spread along one street.
An extra shortcut can make selfish routing slower for everyone.
Moving toward the middle can win more nearby voters in a simple one-dimensional election.
If everyone expects a quiet venue, the venue may become crowded.
Spending more effort raises your chance of winning but everyone pays their effort cost.
Choose a side—but the smaller crowd wins.
No move wins against every response; predictability is exploitable.
Choose boxes, then distinguish prediction from causation.
Watch a small bid become an expensive contest.
A tiny undercut changes both travelers’ payments.
Take stones from one pile and try to take the last stone.
Race two coin patterns that have equal frequency but unequal winning chances.
Find a cycle of dice where no die is best against every opponent.
Compare fighting and yielding as the opponent mix changes.
Three consistent voters can produce a cycling majority.
Make proposals, keep tentative matches, and eliminate blocking pairs.
You both want to meet, but prefer different venues. Can you coordinate?
Choose how much to produce before seeing the market price.
Commit first, then watch a rival choose its best response.
Price a shared product and see what a one-token undercut changes.
Outlast a rival—but remember that every round costs you.
Allocate ten tokens across three contests. Win places, not just one landslide.
Route traffic and compare individual incentives with total delay.
Run two auction rules on the same values and compare one sale with many.
Choose a ranking and see whether an honest ballot produces your least preferred winner.
Find a credential threshold that separates two types with different costs.
Commit an investment before bargaining and inspect whose incentives cover its cost.
Ask for a share, face rejection, and watch delay shrink the available pie.
Divide a shared surplus by inspecting every joining order.
Change the quota and compare vote weight with pivotal power.
Change future incentives before a rival decides whether to enter.
Compare maintenance incentives before and after a warranty shares the loss.
Report a benefit, fund a project and inspect the pivot payments.
Assign a noise right and test whether a voluntary bargain changes the outcome.
Decide when to stop a descending auction clock before a scripted rival does.
Choose a bid for a twelve-unit prize when every participant pays regardless of winning.
Compare a hat puzzle with and without a public announcement that at least one hat is blue.
Price two quality tiers and observe which contract each private customer type chooses.
Follow or ignore a private recommendation in a collision-avoidance payoff game.
Design a publicly known signal policy while keeping the receiver’s Bayesian incentives visible.
Choose actions after costless messages from aligned or biased scripted senders.
Execute cycles of favourite available houses without using money.
Remove only strategies that are strictly worse against every active opponent action.
When someone acts on your behalf, whose interests are they serving?