Explainer
World Test Championship points system: the arithmetic
Points per Test, why the table divides by points available, exactly where the over-rate deduction lands, and worked sums for projecting the standings.
By CricketTaken EditorialPublished Explainer19 min read
A World Test Championship table is one division sum repeated nine times. Points earned on the top, points available underneath, multiplied by a hundred. That is the whole of the World Test Championship points system, and nearly every argument about the standings turns out to be an argument about the number underneath the line rather than the number above it.
The parent question of what the competition is for, and whether a single final at a neutral ground settles two years of cricket honestly, is covered in the wider explanation of how the championship is built. This piece does something narrower and more useful. It follows a result from the moment the umpires shake hands to the moment it changes a figure on a table, and it shows the sums, because the sums are where the format's real decisions are hiding.
There are only four inputs and one operation. Once you can do the operation, a standings graphic stops being a thing you read and becomes a thing you can check.
What the World Test Championship points system pays for each result
Every Test in the competition is scored by itself. The series it sits in is irrelevant to the arithmetic. A dead rubber at the end of a five-match tour is worth precisely what the first Test of that tour was worth, and both are worth what a standalone two-match series is worth.
| Result | Points to each side | Share of the match's twelve |
|---|---|---|
| Win | 12 | 100% |
| Tie | 6 | 50% |
| Draw | 4 | 33.33% |
| Loss | 0 | 0% |
Twelve points are on the table in every championship Test. That number never changes and never depends on anything. It is the single most important quantity in the system, because it sets the denominator, and the denominator is what makes the table comparable at all.
A tie, in which both sides are dismissed and the scores finish level, pays half. It is among the rarest results cricket produces and you can more or less forget it exists when thinking about how a table moves.
The draw is the number that does the work. Four out of twelve is 33.33 per cent, and that figure is not a piece of trivia, it is the pivot the whole table turns on. Any side whose percentage currently sits below 33.33 improves it by drawing a Test. Any side above 33.33 has its percentage dragged downwards by the same result. The identical scorecard raises one team and lowers another, which is the first thing about the system that surprises people and the first thing worth internalising.
Notice what is missing from that table. There is no margin of victory, no bonus for an innings win, no credit for the runs. A side that wins by one wicket in the final over gets twelve. A side that wins by an innings and plenty gets twelve. Cricket's other great normalising device, the net run rate calculation used to separate teams in limited-overs tables, measures exactly the thing the championship refuses to look at. The choice is deliberate. Margin in a Test is heavily contaminated by declarations, weather and the state of the pitch on the fifth day, and a system that paid for margin would reward captains for grinding on after the result was already secure.
Why the World Test Championship points system divides rather than adds
Here is the constraint that forces everything else.
The competition does not build its own fixture list. Series are agreed between pairs of boards, usually years ahead, and the number of Tests in each is a commercial negotiation before it is a sporting one. How that calendar is assembled, and why some tours run to five matches while others run to two, is the subject of the way international fixtures actually get scheduled. The championship takes what the calendar produces and declares that six of those series count.
Which means the sides in the table are not playing the same amount of cricket. Not roughly the same, and not within a Test or two. One team may play twelve championship Tests in a cycle. Another may play twenty. If you rank on points won, the second team is competing on a scale that runs two thirds higher than the first. Every extra Test is another twelve points it can add and the other side cannot. The table would measure the itinerary and call it a championship.
That is not a small bias to be corrected at the edges. On those numbers it is the largest single term in the ranking, larger than the difference between a good team and a mediocre one.
So the table divides. Count the points a side has actually won. Count the points that were available across the matches it played, which is simply twelve times the number of Tests. Express the first as a percentage of the second. Now every side is scored on a scale from nought to a hundred, and twelve Tests and twenty Tests produce figures that mean the same thing.
The denominator is therefore the design decision everything else hangs from. Not the twelve points for a win. The twelve points for a match. Get that distinction wrong and every other part of the system reads oddly, which is why so many explanations of it stall at the over-rate penalty.
Two consequences follow immediately, and both are worth stating before the worked examples.
The denominator moves after every single Test, without exception. Win, lose, draw, tie: twelve points go underneath the line. That is why a defeat costs a side twice. Nothing is added on top and twelve is added below.
Playing more cricket is neither an advantage nor a disadvantage. It is a change in resolution. A side on twenty Tests has a denominator of 240 and its percentage moves in smaller steps. A side on twelve has a denominator of 144 and every result shifts it further. Same scale, different graininess, and the grain matters enormously in a close race.
Two sides with identical records, finishing in different places
The clearest way to see what series length does is to hold the series results fixed and vary the number of Tests inside them.
Everything that follows is invented. Harborne and Calder are constructed teams and every result attached to them was chosen to make the arithmetic legible.
Both sides win three series, draw one and lose two. On any traditional reading of a cycle, they have had the same campaign.
Harborne play all six series over two Tests, so twelve Tests in total.
| Harborne's series | Result | Points earned | Points available |
|---|---|---|---|
| One | Won 1-0, one draw | 16 | 24 |
| Two | Won 2-0 | 24 | 24 |
| Three | Drawn 0-0, two draws | 8 | 24 |
| Four | Won 1-0, one draw | 16 | 24 |
| Five | Lost 0-1, one draw | 4 | 24 |
| Six | Lost 0-2 | 0 | 24 |
| Total | 3 won, 1 drawn, 2 lost | 68 | 144 |
68 divided by 144 is 47.22 per cent.
Calder play two series of five Tests, two of three and two of two, so twenty Tests in total.
| Calder's series | Result | Points earned | Points available |
|---|---|---|---|
| One | Won 3-1, one draw | 40 | 60 |
| Two | Lost 1-3, one draw | 16 | 60 |
| Three | Won 2-1 | 24 | 36 |
| Four | Drawn 1-1, one draw | 16 | 36 |
| Five | Won 1-0, one draw | 16 | 24 |
| Six | Lost 0-1, one draw | 4 | 24 |
| Total | 3 won, 1 drawn, 2 lost | 116 | 240 |
116 divided by 240 is 48.33 per cent.
Identical series records. Calder finish above Harborne, by a margin that in a tight cycle is the difference between playing the final and watching it.
Now look at the match records underneath those series records, because they explain the result and they are not what most people would guess. Harborne won four Tests out of twelve, a third of them. Calder won eight out of twenty, two fifths. The series record concealed a real difference in match-winning, and the percentage found it.
Run the same comparison the other way and the effect reverses just as easily. A side that wins a two-Test series 1-0 with a draw banks 16 from 24, which is 66.67 per cent. A side that wins a five-Test series 1-0 with four draws banks 12 plus 16, so 28 from 60, which is 46.67 per cent. Both won a series 1-0. One of them gained twenty percentage points on the other for it.
That is the honest summary of what series length does. It does not systematically favour long series or short ones. It changes how much a given series result is worth, and the direction depends entirely on the shape of the result inside it. A short series magnifies whatever happened. A long one dilutes it.
Short series also produce a coarser set of possible outcomes, which is a subtler problem. In a two-Test series only six percentages are attainable at all: 0, 16.67, 33.33, 50, 66.67 and 100. There is nothing in between, because the only building blocks available are twelves and fours. A five-Test series offers a far finer grid. Sides on short programmes therefore move around the table in larger jumps, and their standing is more sensitive to a single session of cricket than the standing of a side grinding through twenty Tests.
From a result on the field to a figure in the standings
The order of operations is where explanations of this system usually go wrong, and there is exactly one step that causes the trouble.
- The match ends and Law 16 supplies the resultWin, tie, draw or no result. The championship changes nothing about how a Test is decided. It only decides what the result is then worth.
- Twelve points are placed on the boardEvery championship Test, in every series, at every ground. This figure is fixed before a ball is bowled and does not depend on anything that happens afterwards.
- The side's share of those twelve is worked outTwelve for the win, six for a tie, four for a draw, nothing for a defeat. Both sides are scored from the same twelve, so a drawn Test hands out eight of the twelve points on offer and a win hands out twelve.
- Any over-rate deduction comes off that shareIf the fielding side finished short of the minimum rate after allowances, points are docked from what it has just earned. Nothing is taken off the twelve that were available. This is the step that makes a penalty cost more than its face value.
- The earned figure joins the running numeratorIt rises on a win, a tie or a draw, and it stands still on a defeat. It can be reduced only by a deduction.
- The available figure joins the running denominatorTwelve, every time, regardless of everything. Twelve Tests played means 144 available. Twenty means 240.
- The percentage is recomputed and the table re-sortedNumerator over denominator, multiplied by a hundred, usually shown to two decimal places because cycles have been close enough for the second one to matter.
The sequence applied to every championship Test. The fourth step is the one most explanations misplace. The deduction is taken out of what the side earned in that match, not out of a season total and not out of the points that were available.
Two things about that sequence deserve spelling out.
The percentage is not an average of the percentages of individual matches, and it is not an average of series percentages either. It is one fraction with one numerator and one denominator, both accumulated across the whole cycle. That distinction sounds pedantic and is not: averaging series percentages would give a two-Test series the same weight as a five-Test one, which is the mistake the competition spent its first cycle making.
The recomputation happens after every match, not after every series. A side's position can therefore fall in the middle of a series it is about to win, which is why standings graphics during a long tour sometimes look wrong to people tracking the series scoreline instead.
Where the over-rate deduction lands, and why it costs double
This is the part almost every explanation of the points system either gets wrong or leaves out, and it is not a technicality. It has moved sides in the table.
Test cricket sets a minimum over rate for the match, calculated with allowances for wickets falling, injuries, reviews, drinks and other stoppages. A fielding side that finishes below the minimum is short by a whole number of overs. Under championship regulations the sanction is not only the usual fine. Championship points are docked. The rate at which they are docked has changed between cycles, and the allowances have been revised more than once, so the figure in force is the one in the current playing conditions rather than the one an older article quotes. How the shortfall itself is counted is a subject of its own, covered in the way over-rate penalties are calculated and applied.
The part that matters here is structural, and it does not change between cycles.
The deduction reduces the numerator. It does not touch the denominator.
Work through an invented case. A side wins a Test and is found three overs short. It earned twelve and is docked three, so nine points go into the numerator. Twelve points still go into the denominator, because twelve points were available in that match and nothing that happened afterwards changes what was on offer. The win has been converted from 100 per cent into 75 per cent.
- 12Points available in the match
- 12Points earned before the deduction
- 9Points entering the numerator after it
- 75Percentage yield of that win
Invented case, using a three-over shortfall and a deduction of one point per over purely to show where the arithmetic lands. The deduction rate in force for any cycle is published in the ICC playing conditions. What does not vary is which side of the fraction the deduction hits.
Compare that with the alternative most people assume is happening, which is that a deduction comes off a season total at the end. If it did, three points off a side that had earned 150 from 240 would take it from 62.50 to 61.25 per cent, a loss of 1.25 points. The arithmetic is the same either way, which is exactly why the distinction is worth making: the penalty is not an administrative fine sitting alongside the competition, it is a permanent reduction in the value of a result that has already been achieved, and it is charged against a denominator that has already been set.
Scale it up. A side two overs short in four separate Tests across an eighteen-Test programme loses eight points from a denominator of 216. That is 3.70 percentage points. In a cycle where the top two are separated by a couple of points, a side can miss a final by a margin smaller than its own accumulated deductions, and the table will never show it, because the table shows only the net figure.
The drawn Test is where the arithmetic gets genuinely unforgiving. A draw pays four. A shortfall large enough to wipe out four points therefore leaves the side with nothing from a match in which twelve points were available, which is scored exactly as a defeat would be. Whether a deduction is allowed to carry further than that, into the running total, is a matter for the playing conditions rather than something to assume. The structural point stands regardless: the denominator is indifferent to your over rate.
There is a second-order effect worth noticing. Because the penalty attaches to the match rather than to the cycle, its percentage cost depends on the size of the denominator it lands in. Three points off a side with twelve Tests behind it costs more than three points off a side with twenty, in exactly the proportion 240 to 144. A side on a short programme is punished harder by the same offence. Nobody designed that. It falls out of the fraction.
How a side can lose a series and still climb the table
This is the single most useful thing to understand about the standings, and once it is clear, most of the confusion about mid-cycle movement disappears.
A series improves your percentage if the percentage you scored inside that series is higher than the percentage you walked into it with. The result of the series is irrelevant. The break-even point is not 50 per cent, and it is not a drawn series. It is your own current figure.
An invented case. Meridian have played ten Tests: four wins, two draws, four defeats. That is 48 plus 8, so 56 points from 120 available, which is 46.67 per cent.
Meridian then tour, and lose a five-Test series 2-1 with two draws. Two wins and two draws is 24 plus 8, so 32 points from 60 available. That series was worth 53.33 per cent.
Their new position is 88 from 180, which is 48.89 per cent. They lost the series and went up, because 53.33 is better than the 46.67 they arrived with.
Now the mirror image, which is the one that produces genuine indignation online. Ashfield, another invented side, sit on 70 per cent. They win a two-Test series 1-0 with a draw, which is 16 from 24, or 66.67 per cent. Their percentage falls. They won the series and dropped, because a 1-0 win with a draw is a worse rate than the one they were already sustaining.
This is not a flaw and it is not the table malfunctioning. It is what any rate-based measure does. A batter averaging seventy who scores sixty has had a good day and a lower average. The championship table works the same way, and the discomfort comes from the fact that we are used to reading series results as binary outcomes rather than as rates.
The practical consequence is that the value of a fifth-day rearguard depends on where the side is standing. For a team down at 30 per cent, batting out a draw is worth 33.33 and is a straightforward gain. For a team at 65 per cent chasing a final, the same draw is a small loss dressed up as a moral victory, and the arithmetic pushes their captain towards the risk of a declaration instead. The reasoning behind when a captain declares and what he is actually buying changes shape once a percentage is involved, because the alternative to a possible defeat is no longer a neutral outcome.
Abandoned matches, forfeits and the difference that matters
Weather creates two completely different situations, and they are treated in opposite ways.
A match that is played and ends in a draw because rain took two days is still a draw. It pays four out of twelve. There is no adjustment for the lost time, no reduction in the points available, nothing at all. Both sides take 33.33 per cent from a game neither had a fair chance to win. That is harsh, and it is also the only workable rule, because any attempt to discount a match by the amount of play lost would require the competition to judge what would otherwise have happened.
A match in which no play at all is possible is a different animal. Under the competition regulations such a match is removed from the calculation, so neither the points earned nor the points available move, and the side's percentage is exactly what it was before. This is the one place where the denominator does not advance. It is the right answer: a side that turned up and was rained out entirely has neither achieved nor failed at anything, and adding twelve to its denominator would punish it for the weather.
Forfeits and matches not played for other reasons are handled separately again, with points awarded to the side that was ready to play. Those situations are rare enough that the sensible course is to read the competition regulations for the cycle in question rather than to reason from first principles, because the treatment has been adjusted as the ICC has met cases it did not anticipate.
The distinction to hold on to is simple. A result, however unsatisfactory, goes into both sides of the fraction. A match that never happened goes into neither.
How the first cycle's arithmetic broke, and what replaced it
The original design did not use a percentage, and understanding why it did not is the fastest route to understanding why the current system must.
In the first cycle, every series was worth 120 points in total, split evenly between its matches. A two-Test series paid 60 for a win in each match. A three-Test series paid 40. A four-Test series paid 30. A five-Test series paid 24. A draw was worth a third of the match allocation and a tie a half, so the fractions did not always land on whole numbers and had to be rounded.
The intention was sound and identical to the intention behind the current system: no side should benefit from having more cricket scheduled. The method chosen was to make every series worth the same. Since every team played six series, every team's denominator was 720, and the table could be ranked on raw points because the denominators were equal by construction.
Then look at what the equal-series rule did to individual results.
Winning a five-Test series 3-1 with a draw paid 72 plus 8, which is 80. Winning a two-Test series 1-0 with a draw paid 60 plus 20, which is also 80. Four Tests of dominance against a full-strength opponent over six weeks was worth exactly what one win and one draw against anybody was worth.
Worse, winning four Tests out of five paid 96, while winning both matches of a two-Test series paid the full 120. The system was not merely indifferent to the length of a series, it actively rewarded the short one, because a clean sweep is far easier to achieve over two matches than over five. A competition intended to add meaning to Test cricket was paying a premium for playing less of it.
There was a second failure inside long series. If the pot is fixed, the fourth Test of a five-match series that has already been decided is worth the same slice as the first, which sounds elegant and in practice meant the biggest series in the sport contained the least valuable individual matches in the competition. The most watched cricket had the lowest points density.
Then the pandemic removed the assumption the whole scheme rested on. Sides stopped completing six series. Tours were cancelled outright. With unequal numbers of series played, a denominator of 720 was fiction, and the raw total became meaningless overnight. The percentage arrived as an emergency repair, mid-cycle, because it was the only method capable of ranking teams who had played wildly different amounts of cricket.
What happened next is the part that matters, and it is usually described as a single change when it was really two.
The flat twelve-point allocation, introduced for the following cycle, fixed the valuation problem. Every Test became worth the same, so a five-Test series was worth more than a two-Test series, exactly as intuition says it should be. That change, on its own, destroyed the equal denominators the first cycle had engineered. Sides playing twenty Tests now had 240 available and sides playing twelve had 144.
Which means the percentage stopped being an emergency measure and became structurally necessary. You cannot have flat per-match points and a raw-total table at the same time. Choose to value every Test equally, and you are forced to divide. The two reforms are not independent changes that happened to arrive together, they are one decision with two halves.
Projecting the table forward, which is what people actually want to do
Mid-cycle, the interesting question is never what the percentage is. It is what the percentage can become. Three sums answer it, and all three can be done on a phone.
Take an invented side. Twelve Tests played, 80 points earned, so 80 divided by 144, giving 55.56 per cent. Six championship Tests remain, so its final denominator will be 216.
The ceiling. Win everything left. That is 80 plus 72, so 152 from 216, or 70.37 per cent. No sequence of results can take this side above that number, and comparing the ceiling with a rival's current figure is the quickest way to know whether a race is still live.
The floor. Lose everything left. That is 80 from 216, or 37.04 per cent. The floor is more informative than it looks, because it says how much a side is risking by continuing to play. A team that has finished its programme cannot fall.
The requirement. To finish on a target percentage, multiply the target by the final denominator, then subtract what is already earned. For 60 per cent: 0.60 times 216 is 129.6, so 130 points, less the 80 already banked, leaves 50 needed from the 72 remaining. Four wins and a draw is 52 and clears it. Three wins and three draws is 48 and does not.
Invented side on 80 points from twelve Tests, with six remaining and every non-win assumed drawn. The line is straight because each win instead of a draw is worth exactly eight extra points out of a fixed final denominator of 216.
Show the numbers
| Item | Value |
|---|---|
| 0 wins | 48.15% |
| 1 win | 51.85% |
| 2 wins | 55.56% |
| 3 wins | 59.26% |
| 4 wins | 62.96% |
| 5 wins | 66.67% |
| 6 wins | 70.37% |
That line is straight, and its slope is the most useful single number in the whole exercise. Each win instead of a draw adds eight points to a denominator of 216, which is 3.70 percentage points. Once you know the slope for a given programme length, you can convert any gap in the table into results without doing further arithmetic. A rival four points ahead is a little over one result away. Nine points ahead is two and a half.
The general form is worth committing to memory, because it works for any side in any cycle. A win instead of a draw is worth eight points. A draw instead of a defeat is worth four. Divide by the final denominator and you have the percentage swing. Everything else is bookkeeping.
One warning about projections. The denominator you are dividing by depends on the number of Tests a side will actually play, and that figure is not always as fixed as it looks. Series get rescheduled, matches get added, and a side's final denominator can move after you have done the sum. Any projection is conditional on the fixture list surviving, which in Test cricket is not a safe assumption.
What separates two sides who finish level
Percentages carried to two decimal places rarely tie, but the regulations do not leave it to chance. The competition regulations set out an ordered sequence of tie-breaks, and that sequence has been adjusted between cycles, so the version that applies is the one in the regulations for the cycle in force rather than whatever was true last time.
The criteria that have been used run along predictable lines: the number of series won, then performance in away matches expressed as a percentage of the points available on the road, and finally position in the ICC Test team rankings on a specified date. The last of those is worth pausing on, because the rankings are a separate system built on a separate formula that includes matches outside the championship entirely. The way the ICC rankings formula weights opponents and results has nothing to do with the percentage table, which means that in the last resort a championship place could be decided partly by cricket the championship does not count. It has not happened. It could.
The away-matches criterion is the one that says most about the competition's values. Of everything the percentage cannot see, home advantage is the largest, and the tie-break is the only place in the entire system where the format admits it exists.
Reading the standings without being misled
Four checks, all of them arithmetic, none of them longer than a few seconds.
Convert every gap back into results before reacting to it. A percentage gap means nothing until it is expressed in wins. Multiply the gap by the final denominator, then divide by eight to find how many win-instead-of-draw swings are inside it. Gaps that look enormous in percentage terms are routinely one result wide, and gaps that look tight are sometimes beyond recovery. Percentages compress, and the compression is worst exactly where the table is most interesting.
Look at the denominator before the percentage. A side that has played three Tests has a denominator of 36, and a single result moves its figure by up to 33 percentage points. Early-cycle tables are not information. A percentage only starts to mean something once a side has five or six Tests behind it, and the sides in a table often reach that point months apart.
Count the points still available, not just the matches remaining. A side five points behind with eight Tests left is in a completely different position from a side five points behind with two left, and the second may already be mathematically finished without the table saying so. Ceiling and floor are the two numbers a standings graphic never shows and always should.
Check whether deductions have been applied. Over-rate penalties are announced quietly in disciplinary notices after matches and go straight into the numerator with no annotation in the table. A side docked repeatedly across a cycle carries the cost permanently, and the standings display only the net result. If a percentage looks lower than the win-loss record suggests, deductions are usually the reason.
Do those four and the table becomes what it was built to be, which is a running account of results converted onto a common scale. The current standings and the fixtures still to come sit on the championship section of this site, and everything else about the format, the final and the arguments around it is in the cricket archive.
One last thing, the sentence to keep if you keep nothing else. The championship does not rank teams on how much they have won. It ranks them on what proportion of what was in front of them they took. Those are different questions, and the second one is a great deal harder to flatter.
Common questions
How are World Test Championship points calculated?
Each Test is scored on its own and carries the same allocation regardless of the series it belongs to: twelve points for a win, six each for a tie, four each for a draw and none for a defeat. Those points are added to a running total, and a separate running total counts twelve points for every Test played, whatever the result. The table ranks teams on the first figure divided by the second, expressed as a percentage.
Why is the World Test Championship table a percentage instead of total points?
Because teams play different numbers of Tests. Series lengths are agreed between boards rather than set by the competition, so one side may play twelve championship Tests in a cycle and another twenty. A raw total would rank the side with the fuller schedule higher for no cricketing reason, so the table divides points won by points available and compares the fractions instead.
How many points is a draw worth in the World Test Championship?
Four, which is a third of a win and a third of the twelve points that were available in that match. A drawn Test therefore scores 33.33 per cent, so it lifts a side sitting below that figure and drags down a side sitting above it. That single fact explains most of the movement in the middle of a table.
Do slow over rates change a team's World Test Championship percentage?
Yes, and by more than the raw deduction suggests. Points docked for a slow over rate come out of the points a side has earned, while the twelve points that were available in that match stay in the denominator untouched. A win with a three-over shortfall is scored as nine out of twelve rather than twelve out of twelve, so the match is worth 75 per cent instead of 100.
What happens to World Test Championship points if a Test is abandoned?
A match in which no play at all is possible is taken out of the calculation entirely, so neither the points earned nor the points available move and no side is punished for the weather. A rain-affected match that is played and ends in a draw is a different case: it pays the ordinary four points out of twelve. The competition regulations for the cycle in force set out the exact treatment, including forfeits.
Filed under Cricket·cricket · test cricket · icc · tournament format · explainer