World Tennis Number Explained for British Tennis Players

Two players meet in a club match. One has spent fifteen years winning the second-team singles and the other joined in March. Everyone in the clubhouse knows how that will go, and nobody enjoys it, least of all the winner. The whole purpose of a rating system is to stop that match being arranged in the first place, and Britain now has one that works across borders as well as across the road.

The number itself takes about a minute to understand and rather longer to stop arguing about. It runs downwards, it moves after matches you have already forgotten, and it occasionally rewards you for losing. All of that is deliberate.

What the World Tennis Number replaced

British tennis used its own rating system for decades. It ran on a two-part notation, from the recreational end at around 10.2 up to 1.1 at the top, and club players knew their rating the way golfers know their handicap. Fixtures, internal leagues and county team selection all leaned on it, and it did a reasonable job inside Britain.

Its limitation was that it stopped at the coast. A British 6.1 meant nothing to a Dutch club, an Australian tournament organiser or an American college coach, all of whom had their own incompatible measures. Ratings had become a Tower of Babel across the sport, with dozens of national systems doing the same job in mutually unintelligible units.

The International Tennis Federation's answer was a single global rating, adopted nationally rather than imposed, so that a player's number travels with them. Britain moved onto it, and the old rating system was retired. For players who held both, there was a period of translation, some grumbling, and a fair amount of suspicion that the new number had been set unfairly — an experience common to every rating changeover in every sport.

Two design decisions distinguish it from what came before. It is calculated from match scores rather than from a promotion-and-demotion ladder, and it applies the same scale to a twelve-year-old, a club veteran, a wheelchair player and a professional. That last point is a genuine change in how British tennis talks about standard.

• 40Beginner end of the scale
• 1Elite end of the scale
• 1Decimal places published

Fixed properties of the rating scale itself, not a measurement of any player or population.

Reading the scale from 40 down to 1

The counter-intuitive part is the direction. Lower is better, throughout. A player with a 9 is stronger than a player with a 22, and a player with a 2 is in a different sport from both of them. Newcomers to the system consistently get this backwards for the first few weeks.

The scale is continuous rather than banded. There are no rungs to be promoted between, no thresholds that trigger a change of category, and no annual reset. A number of 17.4 sits between 17.5 and 17.3 and means exactly what the arithmetic says it means. Publishing to one decimal place gives enough resolution to distinguish players who would have identical ratings on a coarser scale, which is the entire point when the system is being used to arrange fair matches.

Because the scale is global and unbanded, it compresses at the sharp end. The difference between 3.0 and 2.0 covers an enormous gulf in playing standard, while the difference between 25.0 and 24.0 is a much smaller step in absolute terms. This is a familiar property of rating scales and it catches people out when they compare gaps. A four-point difference near the recreational end is a competitive but playable match. A four-point difference near the top is not a match at all.

For most players in England the operational range is the middle and upper reaches of the scale, and the useful comparisons are local: the person you play on Thursdays, the second team you are trying to get into, the opponents in your division. Absolute position on a global scale is a curiosity. Relative position within your own tennis is the part that changes decisions.

What counts as a good number

There is no threshold at which a player becomes officially good, and the question is usually being asked for reassurance rather than information. The answer that helps is contextual. A number that gets you competitive matches in your local league is a good number. A number that would get you into your club's first team, if that is what you want, is a good number for you. A number that has come down over two seasons of steady competition is better news than any particular value, because it is evidence of the only thing a rating can really tell you about yourself.

The comparison worth avoiding is with the professional end of the scale. Ratings there compress into a very small range occupied by people who have played tennis full-time since childhood, and measuring a Sunday club player against them tells you nothing except that professionals are extremely good, which was not in doubt. The comparison worth making is with your own number a year ago, and with the numbers of the people you would like to be beating.

How a match result feeds the algorithm

The mechanism is a form of expected-result calculation, and once the shape of it is clear the system's behaviour stops being mysterious.

Before a match is processed, the algorithm looks at both numbers and forms an expectation of what the score should be. It then compares that expectation with what actually happened. If you did better than expected, your number improves and your opponent's worsens. If you did worse, the reverse. The size of the adjustment scales with the size of the discrepancy, so a result close to expectation barely moves anything and a genuine upset moves both numbers noticeably.

Two properties follow directly. The first is that the system is self-correcting: a number that is too generous will be corrected downwards by results that fall short of what it predicted, and one that is too harsh will be corrected upwards. Nobody has to appeal. The second is that it is symmetric — every match adjusts both players, which is how information propagates through a population without any central assessment of anybody's standard.

Only matches recorded in the system count. In Britain that means sanctioned competition: tournaments, leagues and other events whose results are submitted through the official channels. A practice set with a friend, however competitive, produces no data, because nobody submits it. This is the same boundary that governs the domestic points system described in our guide to British tennis rankings, and it catches out club players who wonder why a summer of hard matches has changed nothing.

1. 1Result submittedThe score is entered through a sanctioned competition and reaches the rating system.
2. 2Expectation formedThe algorithm uses both players' current numbers to predict how the match should have gone.
3. 3Comparison madeThe actual game-level score is compared with that prediction.
4. 4Adjustment sizedThe gap between expectation and reality determines how far each number moves.
5. 5Both players updatedThe winner and the loser are both adjusted, in opposite directions, from the same comparison.
6. 6Confidence revisedThe system's certainty about each number changes with the volume and spread of recent results.

The logic of the calculation in sequence. The coefficients and weightings are the ITF's and are not published here.

Why the score matters, not just the win

This is the feature that generates the most complaint and the most benefit.

A binary win-or-lose system throws away almost all the information in a tennis match. Losing 6-4 6-4 to a much better player and losing 6-0 6-0 to them are completely different afternoons, and any measure that records both as "loss" is wasting evidence. Using the game score keeps that evidence.

The practical consequences are worth stating plainly, because they change how a match feels while you are in it. A player who is clearly going to lose still has something to play for: every game taken is data, and a competitive scoreline against a stronger opponent will help their number. A player who is clearly going to win cannot coast: an expected 6-1 6-1 that turns into 7-5 7-5 is a result that falls short of expectation, and their number reflects that.

There is a fairness dividend here that is easy to miss. Under the old promotion ladder, a weaker player entering a strong event had every incentive to withdraw rather than take a beating, because the beating served no purpose. Under a score-based rating, showing up and competing is rewarded on its own terms. Coaches encouraging juniors to play up an age group have a much easier argument to make.

The corresponding risk is gamesmanship, and the sport is not naive about it. Deliberately manipulating scorelines to move a rating is a form of cheating, and competition regulations treat it as such. Systems of this kind also detect implausible patterns statistically over time, since a player whose results consistently defy expectation in one direction is visible in the data.

The rolling period and how old results fade

A rating is a statement about now, so it cannot weigh a match from four years ago as heavily as one from last week.

The system therefore works over a rolling window, with recent results carrying more weight than older ones and the oldest eventually falling out of consideration altogether. That produces behaviour players notice: a number can drift slightly during a period with no matches at all, as the composition of the counted set changes underneath it.

It also means a number decays in confidence rather than in value when somebody stops playing. A player who takes a year off does not get worse on paper. What happens instead is that the system becomes less certain that the number still describes them, which is the honest position — a year off tennis usually does cost something, but nobody knows how much until they play again.

The design goal is responsiveness without volatility. Weight recency too heavily and a single bad Sunday sends a rating lurching. Weight it too lightly and a genuinely improving player spends two seasons carrying a number that flatters their opponents. Every rating system in every sport is a compromise between those two failures, and reasonable people disagree about where the balance should sit. Anyone who thinks their own number lags behind their improvement may well be right; they are also describing the price of stability for everybody else.

Development adjustments for junior players

Children improve between one match and the next, which is a problem for any system built on the assumption that a player's standard is roughly stable while it is being measured.

A rating derived purely from past results will always understate a fast-improving thirteen-year-old, because the results being averaged were produced by a weaker player than the one now walking onto court. Rating systems handle this with an adjustment applied to players in age groups where rapid development is expected, nudging the number in anticipation of improvement rather than waiting for the evidence to accumulate.

The effect is that a junior's number tends to move faster than an adult's from the same volume of matches, and that a junior who has not played for a while may find their number has shifted in their absence. Parents interpret this as an error. It is a deliberate correction for the fact that the child who returns is not the child who left.

The adjustment does not replace results; it works alongside them. A junior who is not in fact improving will have that reflected as soon as they compete, because the actual scores dominate. It is best understood as the system's default assumption in the absence of new information, and the assumption for a twelve-year-old is different from the assumption for a forty-year-old for good reason.

There is a knock-on effect on adult club tennis that catches people out every season. An improving fifteen-year-old carrying a number set by results from six months ago will sometimes arrive in a club league and dismantle opponents whose numbers suggest an even contest. Nothing has been manipulated. The rating is doing what any estimate does when the thing it is estimating is changing quickly, and it corrects within a few matches. Captains who have been on the wrong end of it tend to remember the junior's trajectory rather than the number for the rest of the season, which is exactly the right response.

Separate numbers for singles and doubles

A player holds two numbers, one for each format, and they are calculated independently.

This is more useful than it sounds. Doubles is a different sport played on the same rectangle: the serve matters more, the return matters differently, the court is wider, movement is lateral and cooperative rather than baseline-to-baseline, and a player whose ground strokes are ordinary can be genuinely good at it. British club tennis is dominated by doubles, and a rating system that ignored the distinction would have described most of the country's playing population badly.

The two numbers are frequently a long way apart, and the gap is informative. A player with a much stronger doubles number is typically a good volleyer with a reliable serve and limited patience for long rallies. One with a much stronger singles number tends to be a fit baseliner who has never learned to poach. Captains picking a team learn more from the two numbers together than from either alone, which is a large part of why the leagues have adopted the system so readily.

One quirk of doubles ratings is worth knowing. Because two players share a result, the algorithm has to attribute a team's performance to individuals, and it does so from the numbers each of them brought to the court rather than from any judgement about who played well. A weaker player carried to a good win by a strong partner gains from it; a strong player let down by a weak one loses. Over a season with varied partners this averages out honestly. Over a season played exclusively with the same partner it does not, and a player who only ever partners somebody considerably better should treat their doubles number with a degree of scepticism.

Mixed doubles and other formats are handled within the doubles framework rather than as a third scale. The practical guidance for anybody entering competition is to check which of your numbers an event or league is using, since a tournament using the doubles number for a doubles draw will produce quite different acceptance and seeding outcomes from one using singles.

How many matches you need before the number means anything

There is a difference between having a number and having a number worth trusting, and the system distinguishes between the two.

A relatively small volume of counted results is enough to generate a provisional figure, and the ITF publishes the current threshold, which is the kind of detail worth checking rather than taking from memory. Reliability is the more useful question. A number built from four matches against three opponents from the same club describes a small and possibly unrepresentative sample. A number built from twenty-five matches across leagues and tournaments, against opponents of varying standard, is a genuinely good estimate of how well somebody plays.

Two things improve reliability faster than volume alone. The first is variety of opposition: beating the same person repeatedly adds little, because the system already knows how that match goes. The second is a spread of standards, which anchors your number against the wider population rather than against a local pocket of it. A club whose members only ever play each other can develop an entire internal scale that is systematically out of step with the world beyond it, and the cure is inter-club competition — which is one of the underrated arguments for playing in the structures described in our guide to adult tennis leagues in England.

For a player new to competition, the honest advice is to get matches in and ignore the number for the first season. It will be jumpy at the start, because the system is still locating you, and reading meaning into early movement is a reliable way to become annoyed about arithmetic.

Using your number to find fair matches

The system's primary purpose is matchmaking, and this is where an ordinary player gets the most out of it.

The basic move is to look for opponents within a modest range of your own number. Close numbers predict close matches, which are the ones worth playing: both players are extended, both learn something, and neither spends ninety minutes being embarrassed. Organisers of local competition increasingly build draws and divisions around rating bands for exactly this reason, and the informal recreational competition described in our page on starting tennis as an adult in England leans on the same idea.

A useful second move is deliberate stretching. Playing somebody a few points stronger is where improvement comes from, and under a score-based rating the cost of doing so is low. There is no ladder to fall down and no rating penalty for a competitive defeat.

The third use is honest self-assessment before entering a tournament. If the acceptance list from a previous edition shows numbers well clear of yours, you have your answer about whether it is the right event. That is a different question from whether your ranking will get you accepted, which is covered on the rankings page, and running both checks before entering is the mark of somebody who has planned a season rather than filled in a form.

1. 1RegisterObtain an LTA number and the membership category that competitive entry requires.
2. 2Play sanctioned matchesEnter leagues or tournaments whose results are submitted to the rating system.
3. 3Receive a provisional numberOnce enough counted results exist a figure is generated, initially with low confidence.
4. 4Broaden the oppositionPlay across clubs and standards so the number is anchored to more than one local pocket.
5. 5Use it to choose eventsCompare your number with the field in competitions you are considering entering.
6. 6Review across a seasonJudge movement over months of matches rather than reacting to individual results.

The practical sequence for a British player. Thresholds and membership requirements are set by the LTA and the ITF.

What improving your number actually takes

Players who understand the mechanism start asking how to move it, and the honest answer runs against the instinct most of them arrive with.

The instinct is to protect the number: to enter events you will win, avoid opponents who will beat you, and treat every match as a defence of a position. That works under a promotion ladder. It does very little here, because beating opponents the system already expects you to beat produces almost no movement. The information content of a predictable result is close to zero, and the adjustment reflects that.

What moves a number is exceeding expectation, and the reliable way to do that over time is to get better at tennis. There is no shortcut hiding in the arithmetic. What the arithmetic does do is tell you where the evidence of improvement will show up first, and it is not usually in wins. It is in margins. A player who has been losing 6-2 6-3 to the same opponent and starts losing 6-4 7-5 has improved measurably, and the rating registers that before the win arrives. Anyone who has coached an adult improver knows how demoralising the pre-win period is, and a measure that acknowledges progress during it is genuinely valuable.

The second lever is volume against varied opposition. Playing more matches, against people from outside your usual circle, gives the system more to work with and moves your number towards an accurate estimate faster. If your number is currently flattering you, that is uncomfortable. It is also the point.

The third is patience with the timescale. Meaningful movement is a matter of seasons rather than weeks for most adults, because adult improvement is genuinely slow and the rating is not going to invent progress that has not happened. Juniors move faster, partly because they are improving faster and partly because of the development adjustment described above.

The behaviour to avoid is chasing the number itself — declining fixtures, retiring from matches going badly, or picking opponents to farm. Beyond the competition-rules problem with some of that, it produces a rating that no longer describes you, which defeats the purpose of having one.

How British clubs and leagues use the number

Adoption at club level has been quicker than most administrative changes in the sport, because captains and organisers had an obvious problem that the number solves.

Team selection is the clearest case. A captain choosing an order for a league fixture has traditionally relied on a mixture of memory, reputation and internal politics. Two numbers per player make that conversation shorter and more defensible, particularly when a newer member turns out to be considerably better than the person whose place they are taking. County structures use the same information alongside results in the competitions set out in our guide to county tennis leagues.

Internal competition has changed too. Club box leagues and ladders can be seeded from ratings rather than from last year's finishing order, which stops the same four people occupying the top box indefinitely. Some clubs run rating-banded events specifically so that improving players get a competitive draw instead of an annual first-round exit from the club championship.

The administrative side sits alongside membership. Competitive entry in Britain runs through the LTA's membership and competition system, described in our page on LTA Advantage membership, and a player's results flow into the rating from there. For anybody following the professional game rather than playing, the wider tennis coverage on this site sits in the tennis section, and the rest of the practical material for players in England is collected on the England guides hub.

There are limits to what clubs should do with it. Using a rating as an entry requirement for social tennis, or as a basis for excluding people from sessions, converts a matchmaking tool into a status hierarchy. Plenty of clubs have made that mistake and had to unmake it.

Myths about how the number moves

Several persistent beliefs are simply untrue, and they cause real decisions to be made badly.

The first is that losing always makes the number worse. It does not. A defeat that is closer than the algorithm expected improves it. The corollary catches people out more often: an unconvincing win against a much weaker opponent can move a number in the unwanted direction, which is why "I won but my number got worse" is a common and entirely correct complaint.

The second is that the number can be protected by not playing. It cannot, in any meaningful sense. Confidence in it decays, results drop out of the window, and the player returns to competition with a figure that describes somebody they no longer are.

The third is that one bad match ruins it. Individual results move a number by small amounts unless they are wildly out of line with expectation, and even then the system is designed to be corrected by subsequent evidence. Judge it over a season.

The fourth is that it is a ranking. It is not, and conflating the two is the single most common error in British club tennis. A ranking sorts players into a list by accumulated points and governs who gets into oversubscribed draws. A rating estimates standard and governs whether a match will be worth playing. They routinely disagree about the same person, and both are right about their own question.

The fifth is that the algorithm can be reverse-engineered and gamed by a club player with a spreadsheet. The published mechanism is descriptive; the coefficients are not public, and the manipulation strategies people describe in bars mostly amount to losing matches on purpose, which is against competition rules and, more to the point, a peculiar way to spend a Saturday. The number is a tool for arranging better tennis. Treating it as a score to be maximised is a category error that the system was specifically designed to discourage.