Why a Batting Average Divides by Dismissals, Not Innings

Cricket's averages are some of the oldest statistics in sport, and they are still the first numbers quoted about any player. They are also regularly worked out in ways that give the wrong answer — usually by dividing by the wrong thing. This guide goes through each figure, what it divides by, and why.

Batting average: runs per dismissal

A batting average is the runs a batter has scored divided by the number of times they have been out. Not the number of innings — the number of dismissals. An innings in which the batter was not out adds its runs to the total but adds nothing to the divisor.

Runs	Innings	Not outs	Dismissals	Average
1,450	38	6	32	45.31
1,450	38	0	38	38.16

The same runs in the same number of innings give an average of 45.31 with six not-outs and 38.16 with none. The first batter is not lucky: they were simply not dismissed six times, and an average measures how many runs a batter makes each time they are out.

The batter who was never out

Follow that logic and a batter who has never been dismissed has no average: runs divided by zero is not a number. Scorecards often show such a batter's average as a dash, and the cricket average calculator says "no dismissals yet" rather than inventing a figure. Treating the runs themselves as the average — a common fudge — overstates nothing for one innings, but it breaks the moment the batter is out for the first time.

Why finishers' averages need care

Batters who come in late, especially in limited-overs cricket, finish unbeaten far more often than openers. That lifts their averages, and it is not a flaw: they genuinely were not out. But it does mean that averages across batting positions are not like for like.

Role	Runs	Innings	Not outs	Average	Strike rate
Top order	1,150	30	2	41.07	82.14
Finisher	620	30	14	38.75	137.78

The finisher has scored barely half the runs, yet the averages are close, because nearly half the finisher's innings ended unbeaten. The strike rate — runs per hundred balls — shows the rest of the story: the finisher's job was to score quickly at the end, and they did.

Strike rate: runs per hundred balls

A batting strike rate is runs divided by balls faced, times a hundred. It measures pace, not reliability, and in T20 cricket it often matters more than the average. A batter averaging 30 at a strike rate of 150 can be more valuable to a side than one averaging 45 at 110, depending on what the team needs from that position. Selectors usually read the two together: an average says how often a batter gets out, a strike rate how quickly they score before they do, and a batter who excels at one while falling well short on the other tends to be picked for a specific job.

The three bowling figures

Bowlers have three figures, and each answers a different question.

Bowling average is runs conceded per wicket: how expensive each wicket was. Economy rate is runs conceded per over: how hard the bowler was to score from. Strike rate is balls bowled per wicket: how often the bowler takes one.

A bowler who concedes 612 runs from 140.3 overs and takes 24 wickets averages 25.50, has an economy of 4.36 and a strike rate of 35.1.

The overs trap again

Economy has the same trap as a run rate: overs are written in base six. 7.4 overs is seven overs and four balls, 7⅔ overs. A bowler who concedes 32 from 7.4 overs has an economy of 4.17, not 4.32. The calculator converts to balls first, which also makes the strike rate exact, since a strike rate is counted in balls anyway.

Attacking bowlers and economical ones

Bowler	Runs	Balls	Wickets	Average	Economy	Strike rate
Attacking	520	720	26	20.00	4.33	27.7
Economical	450	900	15	30.00	3.00	60.0

The attacking bowler goes for more runs an over but takes a wicket every 27.7 balls, so their average is lower. The economical bowler concedes less and takes wickets less often. In a Test match, where wickets win games, the first is usually more valuable; in a T20, where every run matters, the second can be. Neither figure is enough on its own, which is why all three are quoted.

Comparing across formats and eras

Averages depend on the conditions they were made in. Pitches, the ball, the length of the game and the style of the era all move them. An average of 40 in Test cricket and 40 in T20 describe very different batters, and a bowling average from an era of uncovered pitches is not directly comparable with one from today. The calculator gives the figures exactly; interpreting them still needs the context of where they were made.

Sample size

A small number of innings or wickets makes any of these figures unreliable. A batter with three innings and two not-outs can have a spectacular average that means almost nothing. Before drawing conclusions from an average, check how many dismissals — or wickets — it rests on. That count is the real denominator, and the calculator shows it alongside the result so it is never hidden.

Reading the inputs from a scorecard

Scorecards mark a not-out innings with an asterisk, so 87* is 87 runs without a dismissal. A batter who retires hurt or ill is treated as not out for the average, because they did not lose their wicket; a batter who retires out, a rare choice made in some limited-overs matches, is counted as dismissed. An innings in which a batter was listed but did not bat is not an innings at all and is left out of both the count and the divisor. For bowlers, the overs in a bowling analysis are already in cricket notation — 9.3 overs is nine overs and three balls — and can be entered as they are printed. Get those conventions right, and the calculator's figures will match the ones in the record books.