Run a poll on a hidden population: choose a sampling method, set the sample size, correct the bias it injects, and publish an estimate that lands on the truth.
Each poll hides a population's true value. Pick a sampling method (each has a known bias and a cost), choose how many respondents to sample on a budget, then run it. Subtract the method's known bias to recover an unbiased estimate, watch your margin of error, and publish. You score when the true value falls inside your margin.
Poll Position is a survey-sampling game about bias, sample size and margin of error.
Choosing a sampling method with a known bias and cost, then a sample size against a budget, is the real design problem behind every survey — and sampling bias is what decides whether the answer means anything.
Pricing the methods is the original mechanism.
It forces the trade every pollster actually makes. A larger sample narrows the margin of error, which is the part people quote, but it does nothing whatever about bias — a badly drawn sample of ten thousand is worse than a well drawn sample of five hundred, and no amount of size will rescue it.
Poll Position is a survey-sampling game about bias, sample size and margin of error. It's an original free browser game built for PlayPendium.