Who Went Home on Dancing with the Stars: Elimination Rules, Save Format, and Why Fan Votes Alone Can't Rescue a Low-Score Couple

The couple eliminated on any given Dancing with the Stars episode is the one with the lowest combined total of judges' scores and viewer votes — and in a three-couple final, the lowest vote-getter goes home even if they're leading on the scoreboard. The judges' scores are converted to a percentage of the night's maximum, the viewer votes to a percentage of the night's total, and the two percentages are added. That arithmetic is why a perfect 30 can still lose a finale and why the bottom-two format exists. Everything below explains who actually decides, how ties break, what changed when the show moved to Disney+, and which seasons broke the rule outright.

Who decides who goes home on Dancing with the Stars tonight

Two parties share the decision, and neither holds it alone. The judges — four in recent seasons — award each couple a score out of 10 per dance, so a single-dance maximum is 30 and a two-dance night tops out at 60. Those scores are published the moment the dance ends, which makes them feel like the verdict when they're really half the input.

The viewers supply the other half through voting during a live window that opens when the episode starts and closes during the broadcast. ABC and Disney+ cap votes per method: a fixed number per couple per voting method, with the exact limit displayed on screen each season because it has been adjusted before. The cap matters because it stops a single superfan from outvoting a large casual audience.

What you watch on results night is a production of that result, not a raw count. The show announces the safe couples in clumps, then reveals the bottom two or three, then eliminates one. That staging exists for television; the underlying ranking is arithmetic, and the show has been transparent about the formula since the early ABC seasons.

The practical takeaway: no couple survives on judges' scores alone, and none survives on votes alone either. The lowest combined percentage goes home. That single rule answers most of the questions viewers ask every Tuesday.

The percentage formula that actually eliminates a couple

The show converts both inputs to percentages before adding them, so the raw point total never matters on its own. Take a couple who scores 24 on a night where the highest score is 30: their judge share is 80 percent. If they take 10 percent of the night's total viewer votes, their combined total is 90 percent. A couple with a 30 and just 5 percent of the vote lands at 105 percent — lower. That's the whole mechanic.

The same arithmetic applies across a two-dance night where scores reach 60. A couple scoring 48 of 60 holds the same 80 percent share. Skipping the conversion and just adding "points plus votes" would let high-scoring couples run away with the season, which is why the show doesn't do it.

Consider a three-couple final. Couple A scores 30/30 and takes 30 percent of the vote: 130 combined. Couple B scores 27/30 (90 percent) and takes 40 percent of the vote: 130 combined. Couple C scores 30/30 and takes 30 percent: 130 combined. Three-way ties on paper are theoretically possible, and that's where tie-break policy and viewer totals as an integer count — not the rounded percentage — decide matters.

The formula has one known soft spot. Because viewers see the judges' scores before voting, a visibly underscored couple can attract sympathy votes. The percentage conversion doesn't neutralize that; it just gives the sympathy surge a defined weight. If the surge is 12 percent and the score deficit is 10 percentage points, the surge wins.

Bottom two, bottom three, and the judges' save, season by season

The format has shifted repeatedly, and those shifts explain plenty of fan frustration. In the early ABC era the bottom two were announced and the couple with the lowest combined total left. When the judges' save was introduced — the panel choosing whom to rescue from the bottom two — vote totals were overridden in specific weeks, and the show announced the save's use each time.

The judges' save has appeared and disappeared on a season-by-season basis. The show's official Rules page and each season's premiere episode state whether it's active, and that's the only place to confirm it for a given season. Reading a three-year-old recap and assuming the format still applies is the most common way fans get the mechanics wrong.

Disney+ era changes matter too. Voting moved largely to the Disney+ app and website, and the show displays the current per-couple vote cap on screen. If the cap is 10 per couple per method and you cast ten online, your eleventh doesn't register. The change tightened the ceiling on individual fan impact compared with phone-only voting days.

Here's the comparison fans actually need, stated plainly:

If a season uses the save, a couple who lost the popular vote can survive — that's the point of the device. If it doesn't, the combined percentage is final. Check the premiere before the first elimination night if the distinction matters to you.

Scores you can't argue with: every perfect-score elimination that happened

Perfect scores and elimination have coexisted on this show more than once, and the pattern is instructive. A couple who earns straight 10s on a night when several others do too holds a judge share near 100 percent but splits the audience's attention. If a lower-scoring, more relatable couple pulls the larger share of the remaining vote, the gap can be bridged.

The counterintuitive part is that the earlier the perfect score comes, the more dangerous it is. A 30 in week three, before the audience has formed attachments, produces a smaller vote share than the same 30 in a semifinal, when the remaining audience is more committed. Sympathy for an underscored couple also has more time to build.

A specific failure case worth knowing: a couple who lands in the bottom two despite a perfect score is usually the victim of a large-field night — more couples means each one's vote share shrinks, and a differentiated dance style can lose casual voters even while thrilling judges. The show's results-night recap will confirm the bottom-two pairing, but the vote split is only announced when the show chooses to disclose it.

The lesson for anyone reading recaps: "perfect score, eliminated" is not a scandal. It's the formula doing what it's designed to do. Judge share maxes near 100 percent; it can't climb further. Vote share can fall to near zero. The average decides.

Viewer votes alone can't save a low-scoring couple, and here's the math

This is where fan anger most often misreads the format. Suppose a couple scores 21 of 30 — a 70 percent judge share — and takes 45 percent of the vote. Combined: 115. A couple scoring 27 (90 percent) with 30 percent of the vote lands at 120. The 27 wins. The 45-percent vote share was impressive and irrelevant.

Flip the inputs and the arithmetic changes. A couple at 70 percent judge share needs a vote share 20 points above the high-scorer's to beat them. In a ten-couple field, that's very hard; in a four-couple final, it's plausible. Vote share matters most when the judge gap is narrow and the field is small.

Run the numbers on a two-couple final: Couple A scores 30 (100 percent judge share) and takes 40 percent of the vote: 140. Couple B scores 27 (90 percent) and takes 60 percent: 150. Couple B wins with a lower score. That exact shape is why stars campaign for votes — not because scores are meaningless, but because their ceiling is 100 percent and the vote share can be far more variable.

The practical calculation is simple. Whatever a couple's judge share, they need enough vote share to close the gap between them and the leader. If the gap in judge share is 15 points, they need to beat the leader's vote share by 15 points. The show never confirms vote shares publicly after elimination nights, so any "we know who the public voted for" claim is an inference, not data.

Why the winner isn't always the best dancer, in one worked example

The finale format makes the year's most misleading assumption visible. A three-couple final with scores at 30, 29, and 28 converts to judge shares near 100, 97, and 93 percent — all bunched at the top. The vote shares then decide. A 33 percent vote share for the 93-percent couple beats a 30 percent vote share for the 100-percent couple, 126 to 130? No — 93 plus 33 is 126, and 100 plus 30 is 130. The 130 wins. The bunching is real but small.

Size of final field | Maximum judge-share gap | Minimum vote-share gap needed to flip 3 couples | about 7 points | about 7 points 4 couples | about 10 points | about 10 points 10 couples | about 30 points | about 30 points

The table is the useful part: the smaller the final field, the easier it is for a lower-scoring winner, and the larger the field, the harder it is. Season after season, a three-couple finale that crowns the 29-scorer over the 30-scorer is a legitimate outcome, not an upset. The bunching at the top means a swing of a few vote-share points flips the result.

Compare that with an elimination in week five, when ten or more couples remain. The judge-share gaps between top and bottom are wide — often 30 percentage points — and the vote share required to close them is enormous. That's why early eliminations almost always reflect the lowest score or the least-known name, while late-season eliminations feel like audience choices.

The show's official Rules page is the only authoritative place to confirm how vote shares are calculated and whether finals are weighted differently in a given season. Anything else is pattern inference from aired outcomes.

What to watch on results night if you want to predict the bottom

Three cues on results night correlate with elimination more reliably than scores do. The first is the safe-clump order: couples called safe early are usually mid-pack in vote share, and the ones held until the last clump are the ones the show suspects will be in danger. The second is the bottom-two reveal order — the couple announced last is very often the eliminated one, because the staging builds to it. The third is the editing: a couple who gets a full package and a fresh camera close-up on results night is rarely the one leaving.

These are production habits, not arithmetic, and they fail in double-elimination weeks, when the show has to reveal two couples in quick succession. If the show has a judges' save, watch the panel's faces when the bottom two are named — the save's existence is confirmed by the premiere, but the panel's reaction tells you whether it will be used. As of the November 2026 reference point, confirm whether the save is active for the current season on the official Rules page, since it has flip-flopped across modern seasons.

For the two-dance nights, the judge share is calculated on the combined score. A couple who scores 24 on dance one and 27 on dance two holds a 51-of-60 combined, which converts to 85 percent. That's within striking distance of a 26-and-26 couple at 87 percent, and the vote share then decides.

The safest prediction method for a viewer is to ignore the leaderboard and watch who gets the "we'll see you next week" camera treatment on the results-night couch. The arithmetic behind elimination is fixed; the announcement order is theater, and it leaks.

The one night fans misread: what a double elimination changes

A double elimination removes two couples in one episode, and it does so without revealing the full ranking. The format in those weeks is to name the bottom three or four, then eliminate downward until the count is met. That's why fans sometimes conclude the "wrong" couple left — the show never disclosed the middle of the ranking.

The arithmetic doesn't change on a double-elimination night: the two lowest combined percentages go home. What changes is the staging and, occasionally, the vote cap, since a night with two eliminations sometimes extends the voting window or changes the per-couple vote allowance. Check the on-screen display that week rather than assuming.

Ties become more likely on double nights because more couples share the bottom band. Be aware that, once the bottom two are announced, the show's tie-break rules apply; if the two couples' combined percentages are identical, the show has historically favored the viewer vote as the tiebreaker, though that's a policy to confirm on the official Rules page for any given season rather than something the arithmetic guarantees.

Edge case worth flagging: a couple who is saved by the judges' save on a normal night, then faces the same bottom two on a double-elimination night, is at risk twice in two weeks. The save doesn't carry a credit. The arithmetic resets each week, and the same deficit can be fatal twice.

Where to confirm the active rules for any season

Every piece of this article converges on one habit: confirm the active rules before the first elimination night of a season. The premiere episode states whether the judges' save is in use, sometimes displays the per-couple vote cap, and announces the format for any double-elimination weeks. The official Rules page, maintained by ABC and Disney+, is the authoritative source for the voting window and the vote cap; no fan wiki or recap site overrides it.

Season-specific details change. The vote cap has moved, the save has appeared and disappeared, and the number of judges has shifted, which changes the per-dance maximum from 30 to 40 in four-judge seasons. None of that changes the elimination formula — judge share plus vote share, lowest total leaves — but each piece is a fact to verify for the season you're watching, not a constant.

For anything the show hasn't disclosed, don't treat a recap's speculation as data. Vote shares are rarely published, so a claim like "the audience clearly voted for X" is an inference from the staging, not a number. The show's own disclosure — the on-screen vote cap, the premiere's format announcement, and the finale's winner reveal — is the documentation trail.

When in doubt, treat the official Rules page as the source of record and treat a season's aired outcome as the only confirmed vote data. The arithmetic above is what the show has stated publicly; the exact active settings for the season you're following are a short check away.

FAQ

Does the judges' save still exist in modern seasons?

It's a season-by-season setting, and the premiere episode states whether the panel can rescue a couple from the bottom two. That save, when active, can override the lowest combined percentage — so the answer to "who went home" in a save week isn't purely the arithmetic. Confirm it on the official Rules page before assuming it's in play, since it has been withdrawn and reinstated across eras.

Can a couple win with the lowest judges' score?

Yes, if the judge-share gap is small enough and their vote share is large enough. In a three-couple final, the maximum judge-share gap is roughly 7 percentage points, so a winning couple whose score is a couple of points behind only needs a modest vote-share lead. The bunching at the top of a final makes that outcome realistic rather than shocking.

Is the bottom two always the eliminated couple?

No — in double-elimination weeks three or four couples can be named, and only the lowest total leaves. And when the judges' save is active, the bottom two can survive to dance another week if the panel rescues them. The announced bottom two is a staging device; the elimination itself is the lowest combined percentage after any save.

How many votes can I cast per couple?

The cap is announced on screen each season, and the current number isn't fixed across seasons. It's a per-couple, per-method limit, so a phone vote, an online vote, and a text vote don't always count the same way. Follow the on-screen display during the voting window or the official Rules page.

What happens if two couples tie in combined percentage?

When the percentages are close or identical, the show applies its tie-break policy, and the viewer-vote total is the historical tiebreaker. Ties are rare because the percentage conversion makes exact matches unlikely, but the tie-break policy is the place to check for any season in which a tie occurs.

Why do perfect scores sometimes get eliminated?

Because a perfect score gives a couple the maximum judge share of about 100 percent, and that ceiling can't rise any further. A low vote share can drop their combined total below a couple with a lower score and a stronger audience following. The "shock" eliminates are the formula working, not a malfunction.

Are vote totals ever published?

The show rarely discloses raw vote totals, and when it does, it's usually on results night to illustrate the bottom two's gap. If the totals aren't announced, any "who really won the vote" claim is an inference from the staging and the scores, not a confirmed figure.

Is the finale format the same as a regular elimination night?

The formula — judge share plus vote share, lowest total loses — carries through, but the finale field is small and the judge shares bunch at the top, so the vote becomes decisive. The reliable way to know the finale's exact format is the episode's own rules announcement, not the earlier-season recap.