Why Did My DUPR Go Down After I Won?
Because DUPR no longer treats a win as automatic evidence that your rating should rise. Since July 8, 2025, the system has compared your actual score with the score it expected from the matchup. Win by less than expected and your rating can fall.
This is not a glitch—it is the current algorithm
Before July 2025, DUPR used a simpler directional rule: winners went up and losers went down. DUPR changed that behavior on July 8, 2025.
The current model asks a different question:
Did you score more or fewer points than a player or team at your rating was expected to score against this opponent?
That makes the final margin meaningful. A win still advances you in a bracket or gives you the match result, but the rating system is using the score as evidence about playing strength.
The official DUPR example
| Before the match | What happened | Rating signal |
|---|---|---|
| Favored team expected to win 11–3 | Favored team wins 11–8 | Winner scored fewer points than expected relative to opponent → can move down |
| Underdog expected to lose 3–11 | Underdog loses 8–11 | Underdog scored more than expected → can move up |
The counterintuitive part disappears once you stop treating DUPR like league standings. It is trying to estimate skill from the scoreline.
You can also lose and gain DUPR
DUPR gives a singles example in which a 3.5 player faces a 4.0. If the model expects roughly a 5–11 result but the 3.5 player loses only 9–11, the lower-rated player substantially exceeded expectation.
The 3.5 can therefore move up after losing, while the 4.0 can move down despite winning.
Why did my partner's rating change more than mine?
Even when two doubles partners move in the same direction, they do not have to move by the same amount.
DUPR says each player's adjustment also depends on that player's own recent match history. A highly established player with many recent results usually supplies less new information to the model than someone with fewer or older results.
So one result can move the less-established player more, even though both partners played the same match.
Your partner gaining 0.06 while you gain 0.02 is therefore not evidence that DUPR credited one partner with individual shot performance in an ordinary score-based match. The publicly disclosed explanation is that each player's rating history and recency affect the size of the adjustment.
Four reasons a win may not help as much as you expected
| Reason | Why it matters |
|---|---|
| You won by less than expected | The actual score underperformed the model's expected score |
| You have a deep, recent match history | One additional match tends to move an established rating less |
| The result was self-posted recreational play | DUPR says self-posted results carry less weight than club, league or tournament results |
| The match was close to expectation | A result that provides little new information usually creates less movement |
Does that mean you should avoid lower-rated players?
No. The algorithm does not say that playing down automatically hurts your DUPR.
If you perform at or above the expected level, the score can support your rating. The risk comes from underperforming the expectation—not merely from the opponent having a lower number.
The same logic means playing a stronger opponent is not a guaranteed rating hack. You still have to outperform what the model expected from the matchup.
How to see the expected impact before you play
DUPR introduced DUPR Impact in 2026 inside its Forecast feature. The tool lets players simulate a matchup, adjust possible scorelines and see the predicted rating change associated with those outcomes.
That is the most useful way to make the performance-versus-expectation model tangible. An 11–2 and an 11–9 win over the same opponent are different pieces of evidence, and Impact is designed to show that difference before the match.
Forecasts are still predictions rather than a promise that every later account update will be identical, but they are a better guide than old “a win always adds points” folklore.
Why older articles and videos get this wrong
If you find advice saying “your DUPR can never drop after a win,” check the publication date.
DUPR itself says that was the behavior before the July 8, 2025 algorithm update. The company changed the model specifically so that performance relative to expectation—not the binary match result—determines direction.
This is one reason Pickleball Cosmos dates its ratings coverage and links to the governing algorithm documentation. DUPR evolves.
What should you do after an unexpected rating drop?
- Look at the score, not just the W. Ask whether the margin was weaker than expected for the matchup.
- Check DUPR Forecast / Impact. It can show how different scorelines were expected to affect the rating.
- Compare reliability. Newer or stale ratings can move more because the model has less current evidence.
- Check result type. Structured club, league and tournament results carry more influence than self-posted recreational matches.
- Do not judge the system from one decimal move. DUPR is designed to update from a body of results, with recent history affecting stability.
Can your rating move a lot from one match?
It can, particularly when your rating is not yet well established. DUPR says players with fewer or older matches can experience larger adjustments because each new result provides more information about current level.
Players with many recent results generally see smaller changes. This is closely related to the logic behind DUPR's Reliability Score.
Full guide: how DUPR and Reliability Score work →
The answer to remember
A win is a match result. A DUPR change is an estimate update. They are not the same thing.
If you win but score fewer points than DUPR expected from a player at your level, your rating can go down. If you lose but outperform expectation, it can go up. That is intentional behavior under the current algorithm, not an error.
Primary sources
- DUPR — July 8, 2025 algorithm update
- DUPR Support — How is the DUPR rating calculated?
- DUPR Support — Which matches impact my rating more?
- DUPR Support — Why partners can move different amounts
- DUPR — DUPR Impact, 2026
Current algorithm guidance checked August 12, 2026. Advice published before the July 8, 2025 performance-versus-expectation change may describe an older DUPR model.