Bowling economy, average and strike rate explained

A bowler's card carries three summary numbers where a batter's carries two, and the extra one is the reason bowling statistics are harder to read. Economy, average and strike rate are built from the same three raw quantities, balls bowled, runs conceded and wickets taken, arranged in the three possible ways. Each arrangement answers a different question, and no single one of them is the measure of a bowler.

The three questions are worth stating plainly before the arithmetic:

• Economy rate: how many runs does this bowler concede per over?
• Bowling average: how many runs does it cost this bowler to take a wicket?
• Bowling strike rate: how many deliveries does this bowler need to take a wicket?

Runs per over. Runs per wicket. Balls per wicket. Every argument about how to rate a bowler is an argument about which of those three the situation actually cares about.

Economy rate

Economy is runs conceded divided by overs bowled. A bowler who concedes 32 in eight overs has an economy of 4.00.

The arithmetic has one trap, and it catches people constantly. Overs in cricket are counted in sixths, not tenths. A bowling figure of 9.3 means nine overs and three balls, which is nine and a half overs. Dividing runs by 9.3 gives the wrong answer. Convert the part-over to a fraction of six first: 9.3 becomes 9.5, 7.2 becomes 7.333, 4.4 becomes 4.667. Any economy figure calculated straight from the decimal on the card will be slightly too high, and the error grows as the spell gets shorter.

Economy is the only one of the three measures that does not involve wickets. It is a pure measure of containment. A bowler can have an outstanding economy rate and no wickets at all, and in certain roles that is exactly what the side wanted.

Its weakness is that it says nothing about progress. Runs are only half of what a fielding side is trying to control; the other half is the ten dismissals that end an innings. In a format where the innings ends by wickets rather than by overs, a bowler who concedes nothing and takes nothing has not moved the match forward.

Bowling average

Bowling average is runs conceded divided by wickets taken. A bowler who concedes 240 runs and takes eight wickets averages 30.

It is the direct mirror of the batting average, and it is read the opposite way round: low is good. It combines both of the things a fielding side spends and gains, runs given away and wickets acquired, into a single price. The average answers the question a captain asks when handing over the ball in a Test: what will this wicket cost me?

Two structural points about the bowling average matter.

It has no not-out problem. Unlike the batting average, there is no partial innings to exclude and no denominator distortion built into the definition. Every run conceded and every wicket taken enters at full weight. The bowling average is arithmetically the cleanest of the six headline batting and bowling figures.

It is unstable at low wicket counts. A bowler with two wickets from a heavy workload has an average built on a denominator of two, and one extra wicket moves it dramatically. Averages need volume before they mean anything, and a spectacular figure from a handful of matches is usually a sample-size artefact.

The bowling average also conceals the trade-off it is made from. Two bowlers can average the same 26 by completely different routes: one conceding few runs and taking wickets slowly, the other buying wickets at a rate but paying for them. The average deliberately treats those as equivalent. Whether they really are depends on the format, which is where the third measure comes in.

Bowling strike rate

Bowling strike rate is balls bowled divided by wickets taken. A bowler who sends down 300 deliveries and takes six wickets has a strike rate of 50, meaning a wicket every fifty balls, or roughly one every eight and a bit overs.

Note the direction. Batting strike rate is runs per 100 balls and higher is better. Bowling strike rate is balls per wicket and lower is better. They share a name and share nothing else, and readers new to a statistics page routinely read one as the other.

Strike rate is the measure of wicket-taking frequency stripped of cost. It does not care how many runs the bowler leaked while doing it. In a format where the constraint is getting twenty wickets rather than restricting runs, that indifference is the point.

The three measures are tied together by a simple identity that is worth carrying around:

Bowling average = economy rate × strike rate ÷ 6

Runs per wicket equals runs per over multiplied by overs per wicket, and overs per wicket is the strike rate divided by six. Any two of the three figures give you the third. It also means that a bowler can lower their average either by conceding less per over or by striking more often, and the two routes are entirely different skills.

Why economy is the death-overs metric

At the end of a limited-overs innings the batting side has decided to spend wickets. In the final few overs of a Twenty20 innings, or the last ten of a fifty-over innings, a batter who gets out is replaced by another batter with the same instruction, and the innings ends on a fixed ball regardless. The cost of a dismissal to the batting side has collapsed, because there is almost no innings left to lose.

That changes what the fielding side is trying to buy. Taking a wicket at the death is useful, mostly because it brings in someone yet to get their eye in, but it does not end the assault. Preventing runs does. A death bowler who concedes seven an over and takes two wickets has usually helped the side less than one who concedes six and takes none, because the four runs saved are certain and the wickets bought very little.

This is why death bowling is a distinct specialism with its own skills, all of them containment-first: the yorker, the wide yorker, the slower ball, the bouncer as a change-up, and above all the ability to hit the same length under pressure to a field set for a single, not a wicket. A death bowler's economy rate is the number that describes whether they can do the job. Their average, computed over a career spent bowling the most expensive overs in the game, will look poor next to a colleague who bowls in the powerplay, and that comparison is close to meaningless.

The same logic applies with less force to the middle overs of a fifty-over innings, where a bowler's job is often to keep the rate flat and force the batting side to take a risk against somebody else. Economy is the measure of a bowler being used to apply pressure rather than to break through.

Why strike rate is the Test metric

A Test match is won by taking twenty wickets. Nothing else ends the opposition's innings, and a side that cannot take twenty wickets in the time available cannot win however few runs it concedes.

That makes wicket-taking frequency the primary currency and time the binding constraint. There are only so many overs in a match, and a bowler who takes a wicket every forty-five balls delivers substantially more of them into a five-day window than one who takes a wicket every seventy, even if the second bowler's average is a little better because they concede less.

The distinction shows up most clearly in the difference between the two archetypes of Test bowler. One is genuinely fast or genuinely dangerous, buys wickets at a high frequency, and concedes more per over doing it: a low strike rate and a moderate economy. The other is relentlessly accurate, dries up an end, concedes very little, and takes wickets more slowly: an excellent economy and a longer strike rate. Their averages can be almost identical. Their value to a Test attack is not.

A balanced attack usually wants both, and the reason is practical. The accurate bowler builds the pressure that makes a batter take a risk against the dangerous one. A four-man attack of pure strike bowlers leaks runs from both ends and gives the batters somewhere to score; an attack of pure containers can bowl all day without ever looking like getting anyone out. Reading strike rate and economy separately is how you tell which kind of bowler you are looking at, and the average, which merges them, is exactly the number that hides it.

Test economy still matters, but it matters as a supporting figure. Its function is not to save runs for their own sake but to deny the batters release, keep the scoreboard from doing the batting side's work for it, and let the captain keep attacking fields in place. A bowler whose economy is very high is one the captain cannot bowl for long, and a bowler who cannot be bowled for long does not take many wickets whatever their strike rate says.

How the three measures trade off by format

Format	Scarce resource	Primary measure	Secondary
Test	Wickets, and the time to take them	Strike rate	Average, then economy
One-day international	Both balls and wickets	Average and economy together	Strike rate in the middle overs
Twenty20	Balls	Economy	Strike rate, especially in the powerplay
Death overs, any white-ball format	Balls, almost entirely	Economy	Dot ball percentage
Powerplay, white-ball	Early wickets against attacking batting	Strike rate	Economy

The pattern is consistent. The measure that matters is the one denominated in whatever the innings is about to run out of. In a Test the innings ends when the wickets are gone, so count wickets against balls. In a T20 the innings ends when the balls are gone, so count runs against overs. The fifty-over game genuinely sits in between, which is why one-day bowling is the hardest to rate on a single figure and why good analysis of it almost always splits the innings into phases.

One consequence worth stating: economy rates and averages are not comparable across formats at all. An economy that would be poor in a Test is excellent in a T20, and a bowling average that would be routine in a T20 would be exceptional in a Test. Any table that ranks bowlers across formats on a raw figure is producing nonsense.

Which runs are charged to the bowler

All three measures depend on runs conceded, and runs conceded is not the same as runs scored while the bowler was operating. The rule is one of responsibility: the bowler is charged with the runs their delivery was at fault for.

Charged to the bowler:

• Runs scored off the bat, including boundaries.
• Wides. One run for the wide itself, plus any runs run or any boundary that results from it. The delivery was out of the batter's reach as they stood, which is the bowler's error, and the ball is bowled again.
• No balls. One run for the no ball, plus any runs scored off the bat from that delivery. An overstep, a dangerous full toss above waist height, too many bouncers in the over, or a variety of technical breaches all produce a no ball, and all are the bowler's responsibility. The ball is bowled again, and in white-ball cricket a free hit usually follows.
• Overthrows arising from the delivery, which are added to the runs from that ball.

Not charged to the bowler:

• Byes. The ball passed the bat and the keeper without touching either, and the batters ran. The delivery was legal and the batter did not hit it. The runs go on the team total and to nobody's analysis.
• Leg byes. The ball struck the batter's body rather than the bat, and the batters ran. Legal delivery, no contact with the bat, so the bowler is not charged. Leg byes are only permitted if the batter was attempting a stroke or trying to avoid the ball.
• Penalty runs. Five-run penalties for offences such as the ball striking a fielding helmet left on the ground, or deliberate damage to the pitch. These sit against the fielding side and against no individual.

This is the reason the bowlers' runs-conceded column on a scorecard almost never adds up to the team total. Subtract the sum of the bowling figures from the total and the difference is exactly the byes, leg byes and penalties.

The same rule governs maidens. A maiden is an over from which no runs are charged to the bowler. Byes and leg byes conceded during the over do not spoil it, because they are not the bowler's runs. A wide or a no ball does spoil it, because they are. A maiden in which a wicket also falls is a wicket maiden, and some cards give it its own column.

What the three measures still hide

Even read together, the three numbers leave a great deal out.

They do not record who the bowler bowled to. Wickets taken against the top order and wickets taken against the tail are counted identically, and a bowler used mainly against tail-enders will carry figures that flatter them.

They do not record when the bowler bowled. A new-ball spell, a first-change spell into a set batter, and four overs at the death are three completely different jobs producing three completely different sets of numbers from the same bowler.

They do not record the conditions. A green seaming surface, a fourth-day turner and a flat white-ball pitch produce figures that sit in the same column without any adjustment.

They do not record pressure created. A bowler who builds three maidens and forces a rash shot at the other end receives no credit for the wicket. Dot ball percentage captures a little of this; nothing in the three headline figures does.

They do not record dropped catches. A bowler who beat the edge four times and had two catches put down appears identical to one who bowled poorly, and the wicket column is the difference between the two.

They do not record workload management. A bowler kept back for the death, or held for the second new ball, has their figures shaped by the captain's plan as much as by their own skill.

The more granular measures on a modern statistics page start to fill these gaps: dot ball percentage, boundary percentage conceded, phase-by-phase economy, average by batting position dismissed, and false shot or control rates induced. Where those are available, they are usually more informative about what a bowler actually does than any of the three classic figures.

Reading a bowling line properly

A short routine gets most of the value from the numbers.

• Start with the format and the role. Death bowler, powerplay bowler, Test strike bowler and Test container are four jobs with four different correct-looking figures.
• Check the balls bowled before believing anything. All three measures are volatile over small workloads.
• Read strike rate and economy as a pair, then look at the average. The pair tells you what kind of bowler this is; the average only tells you the combined price.
• Convert part-overs before doing any arithmetic yourself. Sixths, not tenths.
• Remember that the runs column excludes byes and leg byes. If you are reconciling a card, that is where the missing runs are.
• Do not compare figures across formats. Compare Test with Test, T20 with T20, and preferably phase with phase.

None of the three numbers is a verdict, and each is most useful when it is read as the answer to its own narrow question rather than as a score out of ten. The bowler with the best average in a squad is not automatically the best bowler in it, and in a Twenty20 side, they are often not even the most important one.

The batting equivalent of this piece is batting average vs strike rate. If the underlying columns are unfamiliar, how to read a cricket scorecard explains where each figure comes from, and the cricket statistics glossary defines the terminology used throughout. You can also browse player records to compare bowling figures across formats.