Batting average

Batting average is a statistic in both cricket and baseball measuring the performance of cricket batsmen and baseball hitters, respectively. The two statistics are related, in that baseball averages are directly descended from the concept of cricket averages^[1].

Cricket

See also: Cricket statistics

International cricket career batting averages (Jan 2004). Note Bradman's Test average of 99.94.

In cricket, a player's batting average is the total number of runs he has scored divided by the number of times he has been out. Since the number of runs a player scores and how often he gets out are primarily measures of his own playing ability, and largely independent of his team mates, batting average is a good statistic for describing an individual player's skill as a batsman. The number is also simple to interpret intuitively, being approximately the average number of runs the batsman scores per innings. Batting average has been used to gauge cricket players' relative skills since the 18th century.

Most players have career batting averages in the range 5 to 50:

• Between 30 and 50 is typical for specialist batsmen and genuine all-rounders. This is also the desirable range for wicket-keepers, though some fall short and make up for it with keeping skill. Players who can sustain an average above 50 through a career are considered exceptional.
• All-rounders who are in practice more prominent bowlers than batsmen typically average something between 20 and 30.
• 15 and under is typical for specialist bowlers.

Career records for batting average are usually subject to a minimum qualification of 20 innings played, in order to exclude batsmen who have not played enough games for their skill to be reliably assessed. Under this qualification, the highest Test batting average belongs to Australia's Sir Donald Bradman, with 99.94. Given that a career batting average over 50 is exceptional, and that only four other players have averages over 60, this is an outstanding statistic. The fact that Bradman's average is so far above that of any other cricketer has led several statisticians to argue that, statistically at least, he was the greatest sportsman in any sport.^[2]

Batting averages in One Day International (ODI) cricket tend to be lower than in Test cricket, because of the need to score runs more quickly and take riskier strokes and the lesser emphasis on building a large innings.

Batting averages are affected by the number of not-outs (innings in which the batsman has not been dismissed). For example Phil Tufnell, who was noted for his poor batting,^[3] has an apparently respectable ODI average of 15 (from 20 games), despite a highest score of only 5* and an overall run total of 15.

A different, and more recently developed, statistic which is also used to gauge the effectiveness of batsmen is the strike rate. It measures a different concept however - how quickly the batsman scores (nomer of runs from 100 balls) - so it does not supplant the role of batting average. It is used particularly in limited overs matches, where the speed at which a batsman scores is more important than it is in first-class cricket.

Leading Test batting averages (retired batsmen) ranked in order of average

(Source: Cricinfo Statsguru 1 March 2008)

Rank	Batsman	Tests	Innings	N.O.	Runs	Highest	Ave[4]	Career dates
1	D.G. Bradman	52	80	10	6996	334	99.94	1928–1948
2	R.G. Pollock	23	41	4	2256	274	60.97	1963–1970
3	G.A. Headley	22	40	4	2190	270*	60.83	1930–1954
4	H. Sutcliffe	54	84	9	4555	194	60.73	1924–1935
5	E. Paynter	20	31	5	1540	243	59.23	1931–1939
6	K.F. Barrington	82	131	15	6806	256	58.67	1955–1968
7	E.D. Weekes	48	81	5	4455	207	58.61	1948–1958
8	W.R. Hammond	85	140	16	7249	336*	58.45	1927–1947
9	G.S. Sobers	93	160	21	8032	365*	57.78	1954–1974
10	J.B. Hobbs	61	102	7	5410	211	56.94	1908–1930

Table shows retired players only, with at least 20 innings completed. * denotes not out.

Leading Test batting averages ranked by career total runs scored

An alternative measure of the status of batsmen is to rank them according to the number of Test runs scored in their career (Source: Cricinfo Statsguru 10 August 2008) (1886 Tests)^[4].

Rank	Batsman	Tests	Inns	n.o.	Runs	Highest	Ave	Career dates
1	S. R. Tendulkar	150	244	25	12,000+	248*	54.23	1989 –
2	B.C. Lara †	131	232	6	11,953	400*	52.88	1990 – 2006
3	A.R. Border	156	265	44	11,174	205	50.56	1978 – 1994
4	S.R. Waugh	168	240	46	10,927	200	51.06	1985 – 2004
5	R. S. Dravid †	125	216	26	10,246	270	53.92	1994 –
6	S.M. Gavaskar	125	215	16	10,122	236*	51.12	1971 – 1987
7	R.T. Ponting	119	199	26	10,099	257	58.37	1995 –
8	J. H. Kallis †	123	209	33	9,761	189*	55.46	1995 –
9	G.A. Gooch	118	215	6	8,900	333	42.58	1975 – 1995
10	Javed Miandad	124	189	21	8,832	280*	52.57	1976 – 1993
11	Inzamam-ul-Haq †	120	200	22	8,830	329	49.60	1992 – 2007
	† played for ICC XI							* denotes not out

There are two notable names on the list when sequenced in this way. Brian Lara was a great accumulator of runs at a high average, but he is only Number 22 on the list of batsmen with over 20 completed innings in order of average. Graham Gooch is not even in the Top 100 on the same list. The career totals for the Pakistan batsmen Javed and Inzamam were virtually identical, hence eleven names in the list, but Javed's average is much the better of the two.

For more comprehensive statistics, see List of cricket batting averages.

Baseball

See also: List of Major League Baseball batting champions and Baseball statistics

In baseball, the batting average (BA) is defined as the ratio of hits to at bats.

Henry Chadwick, an English statistician raised on cricket, was an influential figure in the early history of baseball. In the late 19th century he adapted the concept behind the cricket batting average to devise a similar statistic for baseball. Rather than simply copy cricket's formulation of runs scored divided by outs, he realised that hits divided by at bats would provide a better measure of individual batting ability. This is because of an intrinsic difference between the two sports; scoring runs in cricket is dependent almost only on one's own batting skill, whereas in baseball it is largely dependent on having other good hitters in your team. Chadwick noted that hits are independent of teammates' skills, so used this as the basis for the baseball batting average. His reason for using at bats rather than outs is less obvious, but it leads to the intuitive idea of the batting average being a percentage reflecting how often a batter gets on base, whereas hits divided by outs is not as simple to interpret in real terms.

In modern times, a season batting average higher than .300 is considered to be excellent, and an average higher than .400 a nearly unachievable goal. The last player to do so, with enough plate appearances to qualify for the batting championship, was Ted Williams of the Boston Red Sox, who hit .406 in 1941, though the best modern players either threaten to or actually do achieve it occasionally, if only for brief periods of time.

Ty Cobb holds the record for highest career batting average with .366, 8 points higher than Rogers Hornsby who has the second highest average in history at .358. Cobb's career batting average record will probably never be broken, since even the best of modern hitters find it difficult to hit higher than .360 in more than one or two seasons, let alone consistently throughout their entire careers. The record for lowest career batting average for a player with more than 2,500 at-bats belongs to Bill Bergen, a catcher who played from 1901 to 1911 and recorded a .170 average in 3,028 career at-bats. The modern-era record for highest batting average for a season is held by Napoleon Lajoie, who hit .426 in 1901, the first year of play for the American League. The modern-era record for lowest batting average for a player that qualified for the batting title is held by Rob Deer, who hit .179 in 1991. The highest batting average for a rookie was .408 in 1908 by Shoeless Joe Jackson.

For non-pitchers, a batting average below .250 is often considered poor, and one below .200 is completely unacceptable. This latter level is known as "The Mendoza Line", named either for Mario Mendoza, a stellar defensive shortstop who hit .215 during his Major League career, or for Minnie Mendoza, also a shortstop, who was a long-time minor-league player who finally reached the majors briefly in 1970 at the age of 36, and hit .188 in 16 games. The league batting average in Major League Baseball for 2004 was just higher than .266, and the all-time league average is between .260 and .275.

Sabermetrics, the study of baseball statistics, considers batting average a weak measure of performance because it does not correlate as well as other measures to runs scored, and because it has little predictive value. Batting average does not take into account walks or power, whereas other statistics such as on-base percentage and slugging percentage have been specifically designed to measure such concepts. Adding these statistics together form a player's On-base plus slugging or "OPS". This is commonly seen as a much better, though not perfect, indicator of a player's overall batting ability as it is a measure of hitting for average, hitting for power and drawing bases on balls.

In 1887, Major League Baseball counted bases on balls as hits. The result of this was skyrocketed batting averages, including some near .500, and the experiment was abandoned the following season.

Qualifications for the batting title

The Major League Baseball batting average championship (often referred to as "the batting title") is awarded annually to the player in each league who has the highest batting average. Ty Cobb holds the MLB (and American League) record for most batting titles, officially winning 11 in his pro career.^[1] The National League record of 8 batting titles is shared by Honus Wagner and Tony Gwynn. Most of Cobb's career and all of Wagner's career took place in what is known as the Dead-Ball Era, which was characterized by higher batting averages and much less power, whereas Gwynn's career took place in the Live-Ball Era.

To determine which players are eligible to win the batting title, the following conditions have been used over the sport’s history...^[5]

• Pre-1920 – A player had to appear in 60% of his team's games to qualify for the league title.
• 1920-1937- A player must appear in 100 games.
• 1938-1944 – A player had to appear in 400 at bats to qualify in AL but the NL stayed with 100 games. The lone exception was 1938: By order of the AL president, Jimmie Foxx (.349, in 149 games and 565 at-bats) was awarded the batting title over rookie Taffy Wright (.350, in 100 games and only 263 at-bats).
• 1945-1956 – A player needed 2.6 at bats per team game. (With the 154-game schedule of the time, that meant a rounded-off 400 at-bats.) Note that from 1951–1954, if the player with the highest average in a league failed to meet the minimum at-bat requirement, the remaining at-bats until qualification (for example, 5, if the player finished the season with 395 ABs) were hypothetically considered hitless at-bats; if his recalculated batting average still topped the league, he was awarded the title.

• 1957 to the present – A player has needed 3.1 plate appearances per team game; thus, players were no longer penalized for walking so frequently or benefited for walking so rarely. (In 1954, for example, Ted Williams batted .345 but had only 386 ABs, while topping the AL with 136 walks. Williams thus lost the batting title to Cleveland’s Bobby Avila, who hit .341 in 555 ABs.) In the 154-game schedule, the required number of plate appearances was 477, and since the era of the 162-game schedule, the requisite number of PAs has been 502. (Adjustments to this 502 PA figure have been made during strike-shortened seasons, such as 1981 and 1994.)

Also note that from 1967 to the present, if the player with the highest average in a league fails to meet the minimum plate-appearance requirement, the remaining plate appearances until qualification (for example, 5 PA's, if the player finished the season with 497 plate appearances) are hypothetically considered hitless at-bats; if his recalculated batting average still tops the league, he is awarded the title. (This policy was invoked in 1981, securing Bill Madlock his third NL batting crown, and in 1996, when NL titlist Tony Gwynn finished the year with only 498 PAs.)

The decline of the .400 hitter

A point of interest to baseball followers is that hitting .400 was a special and rare feat in the early 20th century. It was accomplished 13 times between 1900–1941 by 8 players, but has not occurred at all since 1941. Many people have expounded theories on why this is the case.

One theory of particular interest was proposed by biologist and baseball fan Stephen Jay Gould, in his book Full House: The Spread of Excellence From Plato to Darwin.^[6] According to Gould, the disappearance of the .400 batting average does not indicate a decline of baseball skill, but, quite the contrary – an improvement in skill. He suggests that instead of looking at the extreme values (the best and worst hitters), we should be looking at the statistical distribution of the batting average of all hitters. If we do this, Gould notes that the league average of batting averages has stayed constant over the last century (mostly due to rules being changed whenever this average started to change), but the variance has been on a continuous decrease, as all major league baseball players have become better and better. As a result of this decreasing variance, the best and worst batting averages came closer to the league average, and the best batting average dropped below .400.

Since a batter's batting average isn't determined just by the batter's individual skill (as is the case in, say, track and field records), but rather the batter's success against opposing players, the gap in skills of an at-bat narrowed. In the early 20th century, the variance of baseball player skills was still high, so when the top batters played, they had the opportunity to be opposed by both very good and by mediocre players, and as a result had an opportunity to achieve very high batting averages. As baseball became a more professional "industry", variance in player skill came down, and the best batter found himself opposed by consistently very good players, and as a result was not able to achieve as high a batting average as was possible a century earlier.

It is also important to note that pitching strategies have changed dramatically since the era of the .400 hitter. Since the 1950s, pitchers have increasingly tried to strike out hitters, rather than get the hitter to put the ball in play. Hitters also more frequently try to hit home runs, which leads to more strikeouts, but in many cases greater offensive production. Also, it is more acceptable to pitch around strong hitters, and to stop throwing strikes after the first two are thrown in a plate appearance, to try to get the hitter to swing at a ball. Lastly, managers now use many more relievers in an average game. This means that hitters see the same pitcher fewer times in a game (losing the advantage of familiarity), and are more likely to face a fresh pitcher, and even a specialist pitcher brought into a game just to get that specific hitter out.

In general, all of these factors either increase strikeout or walk totals, both of which make it much more difficult to achieve a high ratio of hits to at bats, relative to earlier eras of baseball.

Other possible factors are the increase of the number of games played at night (batting average for day games are higher) and the replacement of a number of older "hitter's paradise" ballparks by stadiums of more uniform dimensions.

All-time leaders

Other contexts

Following from usage in cricket and baseball, batting average has come to be used for other statistical measures of performance.

An example is the Internet Archive, which uses the term in ranking downloads. Its "batting average" indicates the correlation between views of a description page of a downloadable item, and the number of actual downloads of the item. This avoids the effect of popular downloads by volume swamping potentially more focused and useful downloads, producing an arguably more useful ranking.

References