The rules in three numbers
On the WTA Tour, 18 results at most, or 19 with the WTA Finals (2024 rules, to be rechecked for 2026).
The protected ranking
A player injured for a long time loses all his points within a year. So that he can return to the big tournaments, he can request a protected ranking: a frozen ranking used to enter draws.
At the ATP
At least 6 months out through injury. The protected ranking is the average of the player’s ranking over the first 3 months of the injury. It can be used for 9 tournaments, or 12 after an absence of a year or more. It does not allow the player to be seeded.
At the WTA
The ranking is frozen at the one held when the injury occurred. It can be used for 8 tournaments within the year, or 12 after more than a year out.
On Proofset, a player returning after more than 21 days without a match gets the Comeback tag. His protected ranking may be good, but he has not played for a long time.
Reading a ranking line
Here is an extract from a ranking, with fictional players. Click a column heading to see what it means.
| Rank | Change | Player | Points | Tournaments played | Points to defend |
|---|---|---|---|---|---|
| 14 | ▲ 3 | Player A | 2 870 | 21 | 90 |
| 15 | ▼ 2 | Player B | 2 805 | 19 | 1 000 |
| 16 | = | Player C | 2 640 | 24 | 250 |
| 17 | ▲ 11 | Player D | 2 455 | 16 | 45 |
An example reading: player B has almost as many points as player A, but he has 1 000 points to defend this week. If he loses early, he may drop out of the top 20 next Monday. Player D is climbing fast with few tournaments played: a sign of recent form that the rank does not show yet.
The ranking and the Race
The ranking
Points from the last 52 weeks. It is used to enter draws and to determine the seeds (lesson 1.3).
The Race
Only the points won since 1 January. It starts again from zero each season and determines the 8 qualifiers for the ATP Finals and the WTA Finals in November.
In January the Race is almost empty while the ranking is full. In November the two converge. Mid-season, a gap between the two shows whether a player is having a better year than the previous one.
Ranking or ELO?
The ranking
It adds up tournament points over a year. It changes once a week. Beating a top-10 player or a player ranked 300th earns the same points if it happens in the same round.
Proofset ELO
A rating that changes after every match, by more or less depending on the opponent’s strength. Beating a highly rated player raises it a lot; losing to a low-rated player lowers it a lot.
What Proofset measured (benchmark of 11/10/2026): once the ELO is known, the current ranking adds nothing more for predicting the winner. On the other hand, a player’s career-high ranking carries information that the current ELO lacks: what a player has already proved counts.
The ELO calculation, live
An ELO rating converts into a probability of winning. Only the gap between the two ratings matters: a 100-point gap gives the higher-rated player about 64 %, 200 points about 76 %. Set the two ratings, then see what the match result changes.
Standard ELO formula, with a coefficient K = 32 for the example. Proofset ELO follows the same principle with its own settings. The probability shown on a Proofset match card is that of its ELO.
Key takeaway: beating a stronger player raises the rating a lot, beating a weaker one raises it only a little. The ATP or WTA ranking, by contrast, gives the same points whoever the beaten opponent is.
Why it matters for a bet
- The ranking reacts slowly. A player who has been in great form for two months may still be poorly ranked, and vice versa.
- A return from injury can mislead. The protected ranking shows a good rank after months without a match: look at the Comeback tag.
- Proofset picks the favourite by ELO, not by ranking, because ELO follows every match and takes the opponents’ strength into account.
Test yourself
8 questions. You need 7 correct answers to pass the lesson.
0 / 8