Reference · Ratings
How DUPR moves after a match
DUPR compares the score you actually posted with the score it expected from everyone's ratings. It publishes the inputs and examples, not the formula, so this page explains the logic and leaves the arithmetic to DUPR's own app.
DUPR (Dynamic Universal Pickleball Rating) rates players on a 2.000–8.000 scale, with separate ratings for singles, gender doubles and mixed doubles. Since an algorithm update on July 8, 2025, a rating no longer simply rises with a win and falls with a loss. DUPR now calls its approach "performance vs. expectation": each match gets an expected score from the players' ratings, and the gap between that and the real score drives the change.
Why no calculator here: none of the DUPR pages used for this guide publishes the formula, its weights or the size of a typical move. Any tool that outputs an exact "new DUPR" from a score is a guess. DUPR's own app does it: Forecast (announced September 2025) shows win odds and who is favored, and Impact (described May 2026) lets you change a scoreline to see the predicted rating effect.
What goes into the change
| Input | What DUPR says |
|---|---|
| Everyone's current rating | Sets the expected score. In doubles, a team's rating is the average of its two players' ratings. |
| Performance vs. expectation | The biggest factor. Scoring more points than expected raises a rating; fewer lowers it, win or lose. |
| Point margin | DUPR says "every point matters." Its Impact example: an 11–2 win and an 11–9 win move a rating differently. |
| Type of match | Verified club, league and tournament results count more than self-posted rec games. The older "What is DUPR?" page puts it at twice the weight. |
| Recency | Recent matches count more toward the current rating. |
| Volume | More matches make a rating more stable, so each new result has less relative impact. |
| Format | Singles, gender doubles and mixed doubles are rated separately; players in the 50–65+ bracket also get an age-group rating. About 90% of DUPR's data is doubles. |
The sequence, step by step
- Before the match, DUPR projects an expected score from both sides' ratings (team averages in doubles).
- The score is entered, either self-posted in the app or uploaded by a club, league or tournament organizer.
- DUPR compares the actual points with the projection. Beating it moves a rating up; falling short moves it down.
- The size of the move depends on each player's history. DUPR's "How it works" page says each adjustment depends on how many matches a player has and how recent they are, and a 2025 DUPR webinar summary says higher reliability means smaller shifts per match.
- The rating updates shortly after upload. Upload order can technically matter, but DUPR's lead data scientist called the effect negligible.
Worked examples (DUPR's own numbers)
Every score below comes from a DUPR publication. The direction of each move is what DUPR states or what follows from its rule of more-points-than-expected; the size is not published, so none is shown.
- Expected (dashed ring): a 3.5 playing a 4.0 in singles is expected to score 5 points in a 5–11 game (DUPR, July 2025).
- Actual (yellow dot): the 3.5 loses 9–11, four points more than expected.
- Result: DUPR says the 3.5's rating goes up; its 2026 explainer adds that the 4.0 will likely see a small drop.
- Expected: a doubles team is expected to win 11–3, so the underdog's expected score is 3.
- Actual: the favorite wins 11–8; the underdog scored 8.
- Result: the favorite's rating goes down and the underdog's goes up (DUPR, July 2025).
| Scenario (published by DUPR) | Expected | Actual | Direction |
|---|---|---|---|
| Singles, 3.5 vs 4.0 | 5–11 | 9–11 loss | 3.5 up; 4.0 likely a small drop |
| Doubles favorite vs underdog | 11–3 | 11–8 win | Favorite down; underdog up |
| Favorite vs underdog (October 2025 post) | 11–5 | 11–9 win | Ratings move slightly closer; the next prediction might be 11–6 |
| Impact tool example (May 2026) | — | 11–2 vs 11–9 win | Different movements; margin matters |
Two general cases follow from DUPR's 2026 explainer: a 4.5 who narrowly beats a 3.0 has underperformed, and a favorite who wins 11–8 when 11–3 was expected sees a dip. DUPR describes a rating as a measurement of skill rather than a reward or penalty, and says one bad result is absorbed across a long history.
New players: the initialization period
DUPR's "What is DUPR?" page, as read in October 2026, describes a 7-day initialization starting with a new player's first rated match. During it, matches change only the new player's rating. The player leaves initialization after earning 3 points or when the 7 days end, receiving a "best guess" rating if short of 3. Points per match depend on how many others are rated:
| Who else in the match has a known rating | Points |
|---|---|
| Everyone (new player is the only unknown) | 1 |
| Opponents rated, teammate unrated | 0.5 |
| Teammate rated, one opponent unrated | 0.25 |
| Only one player in the match rated | 0.125 |
| Nobody rated | 0.0625 |
Caution: the same page still describes a "modified Elo algorithm" in which wins raise and losses lower a rating, which the July 2025 update replaced. Treat its details as possibly out of date. DUPR's "How it works" page also lists a certified DUPR Coach assessment or an AI video rating as ways to get a starting rating.
The Reliability Score
Next to the rating, each profile shows a Reliability Score from 1 to 100%, separately for singles and doubles. DUPR treats 60% or higher as reliable; its analytics consultant called 60% a "passing grade" chosen from a stability analysis. It rises with more, more recent and more varied matches and fades without play.
| Item | DUPR's published figure |
|---|---|
| Path to 60% assumes | Partners and opponents within 0.5 of you and themselves 60%+; at least 2 unique doubles partners; at least 6 unique opposing teams |
| Path to 100% assumes | Same 0.5 and 60% conditions; at least 4 unique partners; at least 12 unique opposing teams |
| Decay examples (no new results) | 100% to about 60% in roughly 6 months; 60% to about 30% in 3 months; 30% to just under 10% in 6 months |
| Update timing (May 2024) | Weekly, on Tuesdays |
Sources differ on timing: in June 2024 DUPR said the score was not yet part of rating movement, with the stated goal that less reliable ratings move more; the August 2025 webinar summary says higher reliability means smaller shifts. Neither gives the size of the effect.
How DUPR's rule has changed
| Date | What DUPR published |
|---|---|
| May–June 2024 | Reliability Score replaces the older "halflife" match count. |
| January 2025 | The Dink, reporting DUPR data scientist Sarah Carpenter's podcast comments: winning raises and losing lowers a rating; earlier, a close loss could raise one. |
| July 8, 2025 | Performance vs. expectation: wins can lower and losses can raise a rating. |
| September 24, 2025 | Forecast: pre-match win odds and favorite/underdog. |
| October 30, 2025 | Review-request form for players who think a rating is wrong. |
| May 1, 2026 | Impact: predicted rating change for a simulated scoreline. |
Checklist
- Know which rating an event uses: DUPR is not USA Pickleball's tournament rating (UTPR). See ratings compared.
- Enter every result; verified club, league and tournament matches weigh more.
- Play varied partners and opponents (DUPR's 60% path: 2 partners, 6 teams).
- Judge a match by points against expectation, not by the win alone.
- Check Forecast or Impact in the app before reading meaning into a decimal.
- Think a rating is wrong? Use DUPR's review-request form.
Related: pickleball skill ratings, ladder leagues, scoring explained.
Sources
- DUPR — How it works (scale, factors, doubles team average, Reliability Score)
- DUPR — How your DUPR rating is calculated (August 7, 2026)
- DUPR — DUPR algo update: win or lose, your rating can go up (July 8, 2025)
- DUPR — Request a review of your rating (October 30, 2025)
- DUPR — Cracking the DUPR code (webinar summary, August 15, 2025)
- DUPR — Introducing DUPR Forecast (September 24, 2025)
- DUPR — DUPR Impact: see how every match affects your rating (May 1, 2026)
- DUPR — What is DUPR? (initialization, match-type weighting)
- DUPR — Introducing the DUPR Reliability Score (May 28, 2024)
- DUPR — Paths to reliability: 60% and 100% (July 5, 2024)
- DUPR — Reliability Score: a deep dive with Scott Mendelssohn (June 30, 2024)
- Pickleball.com — DUPR updates algorithm (July 8, 2025)
- The Dink — Inside DUPR's quest for accurate ratings (Luke Burton, January 9, 2025)
- USA Pickleball — What are Tournament Player Ratings (UTPR)
Sources checked October 2026. Rules, fees and schedules change — confirm with the venue or governing body.