Who went home on DWTS is decided by a weighted combination of judges' scores and viewer votes, with the lowest combined total eliminated—and in seasons with a bottom-two format, the judges choose between the last two. The show converts both inputs to a shared percentage scale, so a couple with mediocre scores but strong votes can survive while a high-scoring couple with a weak fan base can exit. The exact names change weekly and depend on the current season; the mechanics below explain why any given result happens.
How the combined score actually works
Dancing with the Stars uses a two-input elimination system. Judges award each couple a raw score, typically out of 30 in the standard three-judge era, though the panel size and maximum have changed across seasons. The show then converts every couple's judges' score into a percentage of the total points awarded that night, and converts every couple's votes into a percentage of total votes cast.
The two percentages are weighted, historically 50/50 in most seasons, and added. The lowest sum lands in the bottom. This percentage-of-total method matters because it normalizes across weeks: a 24 out of 30 is worth more on a night when the whole cast scores low than on a night when everyone scores high.
The key mechanism is that votes and scores are not compared in their raw units. They are compared as shares. That is why a couple sitting mid-table with judges can still be safe. If their fan base delivers a large vote share, the combined percentage can exceed that of a couple ranked higher by the panel.
One level deeper: because judges' points are capped and distributed across a fixed number of couples, the spread between the top and bottom judge scores is narrow. Vote shares are uncapped and can be lopsided. A couple that wins the popular vote decisively can overcome a substantial judges' deficit. That asymmetry is the engine behind most "shocking" eliminations.
Quantified context: why small score gaps rarely decide the result
Consider a six-couple night with a 30-point maximum. Suppose scores run 27, 26, 25, 24, 22, and 21, totaling 145. The top couple's judges' share is 27/145, about 18.6 percent. The bottom couple's share is 21/145, about 14.5 percent. The gap is roughly 4.1 percentage points.
Now add votes. If the bottom-scored couple takes 30 percent of the vote and the top-scored couple takes 8 percent, the combined totals flip hard. The bottom couple reaches about 44.5 (14.5 plus 30). The top couple reaches about 26.6 (18.6 plus 8). The judges' ranking becomes almost irrelevant.
This is the arithmetic behind fan-favorite survivals and frontrunner exits. The practical threshold: a judges' share gap of a few points is smaller than typical vote-share swings, so vote share usually dominates close calls. A couple needs a judges' cushion of roughly 8 to 10 percentage points to be safe against a weak vote, and that cushion is hard to build on a 30-point scale.
The scoring math also explains ties. When two couples tie on judges' points, they split the same judges' share, so the vote share alone separates them. Ties are resolved by votes, not by a judges' tiebreak, in the standard combined-score format.
The bottom-two format and the judges' save
Some seasons replace straight lowest-total elimination with a bottom-two reveal. The two couples with the lowest combined totals are announced, then the judges vote on which one to save. This changes the incentive structure. The audience vote still determines who lands in the bottom two. The judges then decide which of those two continues.
The mechanism: votes set the pool of at-risk couples, and the panel applies its own preference within that pool. A couple can survive the vote, land in the bottom two anyway because of a low judges' share, and then be saved by the panel. Conversely, a couple the judges score highly but the public ignores can still be sent home if they fall into the bottom two and the panel declines to save them.
The counterintuitive detail: in a bottom-two season, winning the popular vote does not guarantee safety if your judges' share is low enough to drop you into the last two. The save is a second gate. Straight lowest-total seasons have one gate; bottom-two seasons have two.
Timeline of a typical elimination night
| Stage | What happens | Why it matters |
|---|---|---|
| Performance show | Couples dance; judges assign raw scores | Sets each couple's judges' share for the week |
| Voting window | Viewer votes open, usually during and shortly after the show | Vote share is fixed before results are tallied |
| Results show | Combined percentages computed | Lowest total is identified from normalized shares |
| Bottom reveal | Lowest couples announced (one or two, by format) | Reveals who the vote and scores put at risk |
| Judges' save | Panel votes between bottom two, if that format is active | Second gate; can override the vote-ranked loser |
| Elimination | Couple exits; scores reset next week | Prior weeks do not carry over; each week is independent |
| Reset | New week, new shares | A saved couple starts even the following week |
Each week resets the percentages. A couple's prior vote margin does not carry forward as a buffer. This is why momentum and week-to-week fan mobilization matter more than cumulative standing.
Edge cases: withdrawals, double eliminations, and schedule shifts
Not every exit is a vote. Couples withdraw for injury or personal reasons, and those departures bypass the combined-score mechanism entirely. A withdrawal can also change the percentage denominators mid-season, altering every remaining couple's share.
Double eliminations compress two exits into one night, which intensifies the vote-share swing because more couples are competing for the same vote pool. A double elimination late in a season with few couples left can remove a strong scorer and a weak scorer in the same results show.
Schedule shifts matter too. A week with a shortened voting window reduces total votes, which makes each vote a larger share and amplifies organized fan campaigns. A week with a themed double dose of dances can split a couple's rehearsal time and depress judges' scores without changing their vote share.
Red flags that distort results:
Vote windows that open before all performances air, favoring early slots. Judges' scores bunched tightly, which inflates the relative weight of votes. A couple announced safe before the bottom reveal, then placed in the bottom two, signaling a low judges' share. Withdrawals that shrink the cast and shift every remaining percentage. Double eliminations announced without a judges' save option.
Misconceptions about who goes home
The most common misconception is that the lowest judges' score means elimination. It does not. The lowest combined percentage does, and votes can lift a low-scoring couple clear of the bottom.
A second misconception is that the popular vote winner is always safe. In straight lowest-total seasons that is usually true, but in bottom-two seasons a low judges' share can still place the vote leader in the last two. The save then decides.
A third is that scores carry over across weeks. They do not. Each week is scored and voted independently, and percentages reset. A couple's season-long average is a talking point, not an elimination input.
A fourth is that the judges' save is a formality. It is a real second decision. Panels have saved couples the vote ranked last and have declined to save couples with strong scores. The save reflects the panel's judgment within the bottom two, not the audience's ranking.
The professional caveat: published vote totals are not released, so any claim about exact vote shares is inference from the format and the announced results. Analysts can reason about relative shares from who lands in the bottom, but the precise percentages stay internal.
Comparison of elimination formats
| Format | Inputs that decide exit | Who can be saved | Main risk for a strong scorer |
|---|---|---|---|
| Straight lowest combined total | Judges' share plus vote share | No save; lowest exits | Weak fan base drives exit despite high scores |
| Bottom two plus judges' save | Vote sets bottom two; panel picks survivor | Judges save one of two | Low judges' share drops them into the bottom two |
| Double elimination | Two lowest combined totals exit | Usually no save | One strong and one weak couple can both exit |
| Withdrawal | None; departure is not a vote | Not applicable | Denominator shifts alter remaining shares |
| Tied combined totals | Vote share separates tied couples | Format-dependent | A tie on scores is broken by votes, not the panel |
The formats differ mainly in how many gates stand between a couple and elimination. Straight lowest-total has one gate. Bottom-two has two. Double elimination removes the second gate and doubles the exit count.
Practical actions for reading results and rooting effectively
Viewers who want to understand or influence results can act on the mechanics directly.
Vote during the full window, not just the performance slot, because total votes set the denominator for every share. Recognize that a judges' share gap of a few points is small relative to vote swings, so vote volume is the lever that moves outcomes. In bottom-two seasons, track the judges' share as well as the vote, because a low share can put a popular couple at risk of the save. Treat a couple's placement in the bottom two as a signal about their combined share, not just their scores. Watch for format announcements, since a double elimination or a shortened window changes the math before any dance airs. Discount claims of exact vote percentages, because totals are not published and any figure is an estimate.
The practical threshold for a fan campaign: concentrate votes on one couple rather than spreading them, because share is what counts and a divided vote lowers every recipient's percentage. A couple needs either a strong judges' cushion or a decisive vote share to be safe, and most weeks the vote share is the easier lever to move.
Limitations: what the public can and cannot know
The combined-score mechanism is stable and well documented in format terms, but the actual vote totals are not released. That limits any external analysis to inference. We can explain why a result happened in structural terms; we cannot verify the precise percentages that produced it.
Season-to-season changes in panel size, maximum score, and elimination format also mean that a rule from one season may not apply to another. The 50/50 weighting and the bottom-two save have both varied across eras of the show. Anyone applying the math should confirm the current season's format before drawing conclusions.
Finally, the mechanism explains outcomes but does not predict them. Vote behavior depends on audience mobilization, running order, and week-specific factors that are not observable in advance. The math tells you what a given combination of shares produces; it does not tell you what those shares will be.
This article is general educational information about how a television competition's scoring and voting format works. It is not a prediction of any specific season's results, and it does not promise outcomes for any couple or fan campaign. For current-season eliminations, consult the show's official results announcements.