About TallyLab
TallyLab started as a tool for a book club exploring how voting systems shape outcomes. The question we kept asking: what if we'd used a different method? This site lets you find out.
Plurality
The most common system in the English-speaking world: each voter picks one candidate, the candidate with the most votes wins — even without a majority. Simple, familiar, and deeply flawed.
The central problem is vote-splitting. When two similar candidates run, they divide the votes of voters who like both, often handing the win to a third candidate that fewer people actually prefer. This dynamic pushes systems toward two-party dominance (Duverger's Law) and creates the "lesser of two evils" dilemma.
Real-world example: in the 2000 US presidential election, Ralph Nader received 97,000 votes in Florida. George W. Bush won the state by 537. Whether Nader "cost" Gore the election is debated, but it's a canonical illustration of vote-splitting under plurality.
Approval Voting
Voters approve as many candidates as they like; the candidate with the most approvals wins. It's a minimal change from plurality — same ballot structure, just remove the "pick only one" rule — but the effects are substantial.
Approval voting largely eliminates the spoiler effect. You can vote for your genuine first choice and a more viable alternative without wasting either vote. Studies suggest it tends to elect broadly acceptable, centrist candidates rather than those with a passionate minority base.
It's been used by the United Nations for secretary-general selection, and several US cities and states have adopted or are considering it. Critics argue it loses information about how much voters prefer one candidate over another — which is where ranked and quadratic methods come in.
Ranked Choice (RCV)
Voters rank all candidates in order of preference. If no candidate has a majority of first-choice votes, the last-place candidate is eliminated and their votes transfer to each voter's next choice. This repeats until someone has a majority.
RCV (also called Instant Runoff Voting) guarantees the winner has majority support in the final round and removes the fear of "wasting" your vote on a long-shot candidate. It's used in Australia's federal elections, Ireland's presidential elections, and increasingly in US cities and states — Maine became the first US state to use it for federal elections in 2018.
The main critique: ballot exhaustion (voters who ranked only eliminated candidates don't contribute to the final result), and it's possible — though rare — for a candidate to be hurt by gaining more support (the "monotonicity" problem). It's also more complex to explain to voters and to count.
Quadratic Voting
Each voter gets 100 voice credits. Casting v votes for a candidate costs v² credits — so 1 vote costs 1 credit, 2 votes cost 4, 5 votes cost 25, 10 votes cost 100. Negative votes are allowed. This forces voters to reveal how intensely they feel, not just which direction.
The quadratic cost curve is the key insight: it's cheap to express a mild preference, but expensive to express a strong one. This prevents any voter from dominating the outcome through sheer intensity on a single issue. In theory it aggregates not just majority preference but the total welfare of the electorate.
Proposed by economist Glen Weyl and explored in the book Radical Markets, QV has been used experimentally in Colorado's state legislature for budget prioritization and in various corporate and DAO governance settings. It's an active area of research in mechanism design. The negative vote option — controversial but mathematically principled — is what makes it genuinely different from everything else on this list.