Majority judgment asks voters to grade every candidate — Excellent, Very Good, Good, Acceptable, Poor, Reject — instead of ranking them or picking one. The winner is the candidate with the highest median grade. Paste the ballots below (one voter per line, giving each candidate a grade 0–5) and this computes each candidate's median, breaks ties the official way, and shows the full order. Everything runs in your browser; nothing is uploaded.
Write each candidate as Name:grade and separate them with commas. Grades run 0–5 (5 = Excellent, 4 = Very Good, 3 = Good, 2 = Acceptable, 1 = Poor, 0 = Reject). A candidate left off a ballot is graded 0 by that voter. Names are matched case-insensitively.
Majority judgment (MJ) was proposed by Michel Balinski and Rida Laraki. Every voter gives each candidate a grade from a shared, ordered scale — commonly Excellent, Very Good, Good, Acceptable, Poor, Reject. A candidate's score is the median of all the grades they received: the grade such that at least half of voters gave them that grade or better, and at least half gave that grade or worse. The candidate with the highest median grade wins. Because it uses the median rather than a sum or average, a majority judgment winner can't be dragged down by a handful of voters handing out zeros — the middle of the electorate decides.
Five judges grade two finalists 0–5. A gets 5, 5, 3, 2, 1; B gets 4, 4, 3, 3, 1. Sort each and take the middle grade: both have a median of 3 (Good) — a tie on the headline number. Now apply the tie-break: remove one grade of 3 from each. A is left with 5, 5, 2, 1 (new median 2); B is left with 4, 4, 3, 1 (new median 3). B's grade holds up better just below the median, so B wins. This is majority judgment's signature: two candidates can look equally "Good" on the median, and the method still separates them by who has firmer support around that grade — which the majority gauge summarises as B having more grades above its median than below.
Like STAR and score voting, majority judgment uses a graded ballot rather than a ranking, which sets this family apart from the Borda count, instant-runoff and Condorcet methods that take an order. The key difference within the graded family: score and STAR add scores up (a mean-like measure), while majority judgment uses the median. The median makes MJ far less sensitive to strategic exaggeration — pushing one candidate to 0 and another to 5 barely moves a median — which is the main argument its designers make for it. Want to see how a single ballot set behaves under several rules at once? Try the compare-all-methods tool.
What grade scale should I use? The classic MJ scale has six grades (Reject up to Excellent), which this tool maps to 0–5. Any consistent numeric scale works; just use the same one for every voter.
Why the lower median for an even electorate? Balinski and Laraki define the majority grade as the lower of the two middle grades, so that a majority (more than half) of voters always agree the candidate deserves at least that grade. This calculator follows that convention.
What if candidates tie all the way down? If two candidates have identical grade distributions the tie is genuine, and the tool reports them level and breaks the display order by name so the result is reproducible.
Is anything sent to a server? No. The whole count runs in your browser, so you can paste real ballots safely.