The person voted off Dancing With the Stars is determined by a weighted combination of judge scores and viewer votes, with the lowest combined total eliminated each week. Judges award scores worth 50 percent of each couple's final result, and viewer votes supply the other 50 percent, according to ABC's published voting rules. Neither judges nor fans decide alone. That dual mechanism, plus ties, save formats, and voting-window rules, explains why the "obvious" bottom couple sometimes survives and a crowd favorite sometimes leaves.
The scoring mechanism, layer by layer
Each week, every couple performs and receives numeric scores from the judging panel. Those raw scores are converted into a percentage of the night's total points awarded, not used as raw numbers. A couple earning 24 out of 30 possible points holds roughly 80 percent of the judges' share. That percentage is then halved to represent the judges' 50 percent weight.
Viewer votes work the same way on the other side. Each couple's share of total votes cast becomes a percentage, then is halved. The two halves are added, and the couple with the lowest combined percentage is eliminated. This normalization step is the part most fans miss. A couple does not need the most votes in absolute terms. It needs enough relative support to avoid finishing last once both halves combine.
The consequence is counterintuitive. A couple with a large real-world vote margin can still fall if the judges' score gap is wide enough. Suppose one couple receives 40 percent of all votes but only 60 percent of the judges' share. Another receives 15 percent of votes but 95 percent of the judges' share. After halving and adding, the first couple scores 50 and the second scores 55. The popular couple goes home. This is why a low-scoring dancer with a devoted fan base is never fully safe.
Quantified context: how big the swings really are
The 50/50 split means each side can move a couple by a maximum of 50 combined percentage points. In practice, the judge half compresses differences because scores cluster near the top. On a typical night, the spread between the best and worst judge scores might run from 30 to 21 points, roughly 95 percent versus 66 percent of the judges' total. Halved, that is 47.5 versus 33, a 14.5-point gap. The vote half can easily overcome a gap that size if one couple's fan base is several times larger than another's.
Voting windows matter numerically too. Votes are cast during a live broadcast window that typically runs two to three hours, and online voting often closes shortly after the show ends in the Eastern and Central time zones. Late viewers in Mountain and Pacific time zones vote on a tape delay, but their votes count in the same pool. A surge in the final 15 minutes of voting can shift shares meaningfully because the pool is still growing.
Historically, eliminations begin in the second or third week of a season and continue weekly until a finale of three to four couples. That means a full season produces roughly 10 to 12 elimination events, each with its own combined-score calculation. Across those weeks, judge scores tend to rise as dancers improve, so the vote half becomes relatively more decisive late in the season when score gaps narrow.
Edge cases that change the outcome
Several format rules alter what "voted off" means in a given week.
Double eliminations: Some weeks remove two couples in one night, which compresses the margin between safe and eliminated positions. Judge saves and bottom-two formats: Some seasons place the lowest combined couples into a dance-off decided by judges, so the viewer vote alone does not send anyone home that week. Withdrawals and medical exits: A couple can leave without a vote, which is not an elimination and should not be counted as one. Non-elimination weeks: Occasional weeks announce no elimination, carrying all scores forward. Voting method changes: Some seasons allow unlimited online votes per account, others cap them, which changes how much a single dedicated fan can influence shares. Time-zone weighting: If producers weight regional votes to correct for population differences, raw vote totals do not map cleanly to shares.
A professional caveat worth stating plainly: the show does not publish raw vote totals or exact combined percentages. Any "results" circulated online beyond the announced elimination are estimates or fan reconstructions, not official data. Treat precise vote-margin claims with skepticism.
Misconceptions worth correcting
The most common error is treating the judge leaderboard as the elimination order. It is not. Judges' scores are only half the equation, and the normalization step means a couple can top the judges' board and still be eliminated if its vote share collapses. The reverse is equally true.
A second misconception is that the "worst dancer" always leaves. The mechanism rewards combined support, not technical skill. A technically weak but broadly liked couple can outlast a strong dancer with a smaller fan base.
A third error is assuming vote totals are the deciding number. Because both sides are converted to percentages, the shape of the vote distribution matters more than the raw count. A couple with 20 percent of votes in a six-couple field may be in real danger even if that sounds like a lot.
Finally, fans often conflate the dance-off result with the vote result. When judges choose who goes home from a bottom two, the viewer vote determined who reached the bottom, but the judges made the final call. Those are two different mechanisms stacked in one night.
Comparison of elimination formats
| Format | Who decides the exit | Judge weight | Viewer weight | Typical use |
|---|---|---|---|---|
| Standard combined vote | Lowest combined percentage | 50 percent | 50 percent | Most regular weeks |
| Bottom-two judge save | Judges pick from lowest couples | Decisive at final step | Sets the bottom two | Selected weeks |
| Double elimination | Two lowest combined | 50 percent | 50 percent | Mid-season compression |
| Non-elimination week | Nobody | Not applied | Not applied | Scheduling or twist weeks |
| Medical withdrawal | Contestant or doctors | Not a vote | Not a vote | Injury or illness |
| Finale | Combined vote crowns winner | 50 percent | 50 percent | Last episode |
The table makes one point clear: "voted off" is shorthand. In several formats the viewer vote narrows the field but does not deliver the final blow.
Timeline of a typical elimination week
| Stage | Timing | What happens |
|---|---|---|
| Performances | Live broadcast | Couples dance and receive judge scores |
| Voting opens | During broadcast | Fans cast votes online and by other methods |
| Voting closes | Shortly after broadcast | Vote pool freezes for the week |
| Score normalization | Post-broadcast | Judge and vote shares converted to percentages |
| Combined ranking | Post-broadcast | Halved shares added per couple |
| Results announcement | Results segment | Lowest combined couple revealed |
| Possible dance-off | Results segment | Judges may decide among bottom couples |
| Elimination confirmed | End of episode | Couple exits the competition |
The gap between voting close and results is where the mechanism does its quiet work. Nothing dramatic happens on screen, but normalization and weighting decide the outcome before the announcement.
Practical actions for viewers who want their vote to matter
Fans who want to influence eliminations can act on the mechanism directly.
Vote in every window, not just once. Since shares are relative, consistent voting across weeks compounds a couple's position. Vote early in the window rather than assuming a late surge is safe. The pool grows over time, and early votes count equally in the final share. Understand your platform's cap. If the rules allow a fixed number of votes per account, use all of them every week. Watch the judge-score gap, not just the leaderboard rank. A couple sitting third but far behind first is more vulnerable than a couple sitting fifth in a tight cluster. Do not trust unverified vote totals. Announced eliminations are the only official results. Recognize dance-off weeks. On those nights, reaching the bottom two is the danger, and judges, not votes, finish the job.
A simple threshold to keep in mind: if your favorite is more than about 10 combined percentage points behind the couple above them on the judges' half, the vote half must do heavy lifting. That is not a published cutoff, just a way to read the arithmetic. The show's own rules set no such threshold.
Limitations of what can be known
The public cannot reconstruct exact outcomes because raw vote counts, regional breakdowns, and any weighting adjustments are not released. Published accounts of "shocking" eliminations usually reflect a genuine mechanism, a split between judge and fan support, rather than a malfunction. Without official totals, claims that a specific couple "should have stayed" rest on assumptions about vote distribution that no outside observer can verify.
Format rules also change between seasons and countries. The U.S. version's 50/50 split and voting methods have varied over time, so a rule from one season may not describe another. Anyone citing a specific percentage split should check the season's official voting page rather than assume it carries forward.
The takeaway is structural. The person voted off Dancing With the Stars is the couple with the lowest combined percentage of judge scores and viewer votes under that week's format. Understanding normalization, the 50/50 weighting, and the occasional judge save explains nearly every surprising result. This article is general educational information about a television competition's scoring rules, not a statement about any individual contestant's circumstances; for official results, consult the broadcaster's published episode records.