Economy, Strike Rate and Average: Reading Bowling Figures Guide

Every bowler on a scorecard gets four numbers, always in the same order: overs, maidens, runs, wickets. A line like 26.3-9-74-10 reads left to right, and nothing else needs explaining once you know which slot is which.

Overs is how much bowling the bowler actually did, counted in overs of six balls. A wide or a no-ball does not count as one of the six, so it has to be bowled again. Maidens counts the overs in which the bowler conceded nothing at all. Byes and leg byes are not charged to the bowler, so an over that leaks a couple of leg byes can still be a maiden; a wide or a no-ball is charged, so that over cannot be.

Runs is the runs conceded: everything off the bat, plus the wides and no-balls the bowler bowled. Byes and leg byes are left out. Wickets counts only the dismissals credited to the bowler: bowled, caught, leg before wicket, stumped and hit wicket. A run out is credited to nobody's figures, which is why a side can lose ten wickets while the opposing bowlers show eight between them.

Here is the one thing that trips up almost everyone. The overs column is not a decimal. The digit after the point counts balls, not a fraction of an over, and it can only run from 1 to 5.

Take Jim Laker's innings at Old Trafford in 1956, printed as 51.2-23-53-10. The 51.2 means 51 completed overs and 2 balls of the next one: 51 times six plus 2, which is 308 balls. As a genuine decimal that is 51.33 overs. Same spell, two numbers, and only one of them can be divided into anything.

The gap looks trivial until you divide by it. Jim Laker conceded 53 runs, so his real economy rate is 1.03 an over; treat 51.2 as a decimal and you get 1.04. On a short spell the error is far larger. Wanindu Hasaranga's 5.5-1-19-7 is 35 balls, or 5.83 overs, and his economy is 3.26; read the printed 5.5 as a decimal and you would report 3.45. Every rate below is therefore taken from the ball count. Convert first, divide second.

Economy rate is runs conceded per six balls bowled. It answers one question, how expensive the bowler was, and nothing about wickets enters into it.

Work it through on Anil Kumble's 26.3-9-74-10. He conceded 74 runs from 159 balls. Divide the balls by six and you get 26.5 overs; divide 74 by 26.5 and the answer is 2.79 runs an over. Had you divided by the printed 26.3 instead, you would have said 2.81.

Bowling strike rate is balls bowled per wicket taken: how quickly the bowler got people out. Anil Kumble bowled 159 balls and took 10 wickets, so his strike rate for that innings was 15.9 balls a wicket, a shade under three overs each.

This is the most confused number in cricket, because strike rate means the opposite thing for a batter and a bowler. A batter's strike rate is runs per hundred balls faced, and a big one is good. A bowler's is balls bowled per wicket, and a small one is good. They are not two versions of one statistic: they are not in the same units and they point in opposite directions. Anything a bowler is charged for, balls or runs, reads low when it reads well.

Bowling average is runs conceded per wicket taken, the price of a wicket. Anil Kumble conceded 74 runs for 10 wickets, so each one cost him 7.4 runs.

Average is the other two rearranged: runs per ball, times balls per wicket, gives runs per wicket. So a bowler can reach a good average from two directions. One is hard to score off and takes wickets slowly; another goes for plenty but strikes constantly. Both can average the same, and they are not the same bowler. A bowler who takes no wickets has neither average nor strike rate, because you cannot divide by zero: Mick Lewis's 10-0-113-0 is that kind of spell, 113 runs and nothing to show for them.

Across the 49,117 Test spells in the archive, the mean economy rate is 2.82 runs an over. Across the 59,368 one-day international spells it is 4.78. A run is priced 1.70 times higher per over in the shorter game, and that ratio explains most of what separates the two crafts.

The formats ration different things. A one-day innings is exactly 300 balls, and no bowler may bowl more than a tenth of them. Balls are the scarce resource, so what a captain buys is cheapness: an over costing four instead of six saves two runs the batting side can never get back, because there is no extra over in which to recover them. Economy is the headline number.

A Test innings has no ball limit. The only way to end it is to take ten wickets, and to win the match, twenty, which is why the mean is so much lower and why nobody wins a Test by being tidy. Strike rate is the Test measure that matters most, and average is what careers are ranked by, because over a long career it folds both in. The practical consequence: going at 2.82 an over is exactly average in a Test and outstanding in an ODI, while 4.78 is exactly average in an ODI and a disaster in a Test. Same number, different currency.

These are the best innings figures in each archive, ranked the way a scorecard ranks them, most wickets first and fewest runs as the tie-break. The overs are in cricket notation; the three derived measures come from the ball count.

The three measures are doing separate jobs here. Jim Laker and Anil Kumble both took all ten, but Jim Laker did it over 308 balls at 1.03 an over while Anil Kumble needed 159 and went at 2.79. By strike rate, Anil Kumble's was the faster work, 15.9 balls a wicket against 30.8; by economy and average, Jim Laker's was the more controlled. Neither number is the answer on its own.

The cheapest five-wicket hauls are freakish rather than instructive, which is what makes them useful. Ernie Toshack's 2.3-1-2-5 in a Test cost 2 runs; Courtney Walsh took 5 one-day wickets for 1 run in 27 balls. Figures like these come from collapses on helpful pitches and tell you more about the conditions than about the bowler: an economy rate of 0.22, Courtney Walsh's, is not a skill anybody can reproduce.

The most expensive spells are filtered here to long ones, at least 20 overs in a Test and at least 8 in an ODI, so a single savaged over cannot qualify. Chuck Fleetwood-Smith conceded 298 runs from 87 overs for 1 wicket: an average of 298 and a strike rate of 522 balls. His economy, though, was 3.43, barely above the Test mean of 2.82. He was not expensive by the over; he bowled an enormous number of overs without taking wickets, and only the strike rate catches that. The one-day equivalent is the opposite shape: Bas de Leede's 10-0-115-2, 115 runs from a full allocation at 11.5, well over double the ODI mean, with the damage entirely in the economy column.

Maidens are the other extreme. Bobby Peel bowled 56 in one Test innings, 102.1-56-78-3, conceding 78 runs from 613 balls at 0.76 an over; Phil Simmons's 10-8-3-4 is the one-day version, 8 maidens at 0.3. A high maiden count describes the texture of a spell rather than its result.

Figures are an account of one bowler in isolation, and cricket is not played in isolation. The biggest thing they miss is pressure built at the other end: a bowler who sends down six maidens in a row often gets his wickets credited to his partner, because the batter who has not scored for four overs eventually goes after someone, and it is rarely the man who tied him down.

They say nothing about who was batting, either. Four wickets against the top order on a moving ball is not four against the tail with the game gone, but both read as four. Nor do they record the match situation: an economy of six in the last ten overs of a one-day innings, when everyone is swinging, is better bowling than four in the quiet middle overs.

Then there is the length of the spell. A four-over burst and a forty-over shift are not comparable, whatever the rates say: the short spell is bowled flat out, the long one by a tiring bowler holding an end so someone else can attack from the other. Rates reward the burst and conceal the shift. And figures cannot see the ball that beat the bat. Read them, then read how they were got.

Every bowling line quoted here comes from a scorecard you can open. Try a few in the Test archive and the ODI archive, converting the overs to balls as you go. If the rest of the card is still unfamiliar, the companion guide on how to read a cricket scorecard covers the batting side of it. Then go and generate a set of figures of your own in The Follow-On.

• MCC Laws of Cricket, Law 17: The over, and Law 18: Scoring runs
• CricArcade Test match archive
• CricArcade ODI archive
• Figures computed with scripts/build-guide-facts.ts over the archive's scorecards

• How to Read a Cricket Scorecard: A plain-English guide to the batting card, bowling figures, extras, fall of wickets and the result line, with what every abbreviation actually means.
• How DRS Works: Who can review, what ball-tracking actually shows, and why umpire's call is not a bug: the Decision Review System explained, with what the archive says it changed.
• How Duckworth-Lewis-Stern Works: Why rain does not cost both sides the same, how overs and wickets in hand become a revised target, and what the method cannot know.
• Fielding Positions Explained: The map of a cricket field, from slip to deep backward square leg, and the modifier system that lets you decode a position you have never heard named.
• Browse every guide, or pick a game to play.