TYPE Original Research

PUBLISHED 28 November 2024

DOI 10.3389/frbhe.2024.1492735

Trade-off between fitness and

OPEN ACCESS experience: insights from
historical data on ODI cricket
EDITED BY
Romain Gauriot,
Deakin University, Australia

REVIEWED BY
1 1,2 3
Sarah Jewell, Vipra Jain *, Subrato Banerjee and Priyanka Kothari
University of Reading, United Kingdom
Sundar Ponnusamy, 1
Department of Economics, Indian Institute of Technology Bombay, Mumbai, India, 2 Centre for
Monash University, Australia Behavioural Economics, Society and Technology (BEST), Brisbane, QLD, Australia, 3 Labour Market
*CORRESPONDENCE
Research Facility, School of Management and Labour Studies, Tata Institute of Social Sciences,
Vipra Jain Mumbai, India
[email protected]

RECEIVED 07 September 2024

ACCEPTED 04 November 2024 A healthy mix of younger and senior players is often desirable for any team
PUBLISHED 28 November 2024 involved in a sports-tournament. It is often argued that the experience of senior
CITATION players and fitness of the younger players could complement each other. How
Jain V, Banerjee S and Kothari P (2024) an athlete performs is determined by the accumulation of experience that
Trade-off between fitness and experience:
insights from historical data on ODI cricket.
grows with age and physical fitness that declines with age. Therefore, fitness
Front. Behav. Econ. 3:1492735. and experience are competing forces that move in opposite directions. We
doi: 10.3389/frbhe.2024.1492735 explore the age-range of peak-performance in the lifetime of a sportsperson
COPYRIGHT where the marginal benefits from experience and fitness are on balance. Using
© 2024 Jain, Banerjee and Kothari. This is an a unique data-set of all one-day international (ODI) cricket matches played
open-access article distributed under the
terms of the Creative Commons Attribution from 1971 to 2000, we find that there is an inverted U-shaped performance-
License (CC BY). The use, distribution or age curve where performance first gets better with age, then reaches a peak
reproduction in other forums is permitted, before finally declining (when experience can no longer undo the effects of
provided the original author(s) and the
copyright owner(s) are credited and that the declining fitness). This peak-performance is attained for bowlers at around 22–
original publication in this journal is cited, in 26 years and for batters at about 27–30 years. These results remain robust to
accordance with accepted academic practice. different specifications and controls. We employ a theoretical framework for
No use, distribution or reproduction is
permitted which does not comply with these policy analyses of actual selection decisions made during our period of study.
terms. These analyses build on the empirical section of our paper.

KEYWORDS

aging, experience, group performance, trade-offs, sports economics

1 Introduction and motivation of study

It is often agreed that a great sports team is one that has the right mix of youth and
experience—a team comprising only young (but fit) beginners lacking experience is likely
to fail just as that comprising only senior players who have only experience but lack physical
fitness. Best team performances are often consequences of harnessing the strengths of
diverse age groups. This idea often resonates in the real world in remarkably many contexts
of human endeavor well beyond the world of sports where team-effort is necessary to
achieve a common goal. The youth contributes energy and seniority brings in experience.
We examine this fine balance in the sport of cricket, where the interplay between
experience and fitness determines the team performance. Cricket players acquire
experience as their careers progress. This is seen in the development of tactical acumen, the
ability to strategize, analyze situations, and if needed, lead the team in difficult situations.
These qualities are essential for on-field success (Connor et al., 2020; Crowther et al., 2023).
However, experience is accumulated over time and one ages with time.
Physical fitness is expected to be inversely related with aging and
therefore poses challenges to improve performance or even sustain a peak
when (and if) attained. While experience confers strategic advantages,

Frontiers in Behavioral Economics 01 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

age-related declines in fitness can impact any player’s ability how he used his mental strength in an Indian Premier League (IPL)
to maintain previously attained performance levels. This age- match in 2011, as a 39-year old sportsman.
performance relationship has been studied in the labor market
(Kim and Lee, 2023; Borsch-Supan et al., 2021; Borsch-Supan “I was hardly in the right physical shape and tried my best to
and Weiss, 2016), for cognitively demanding tasks (Dumas, 2015; make up for it mentally. I kept visualizing my past performances.
Hertzog, 2020; Kalwij and Jaegher, 2023; Kanfer and Ackerman, It was a nerve-racking situation.”
2004; Lehman, 1953), and financial tasks (Agarwal et al., 2009;
Eberhardt et al., 2019) among other fields of research. The link Our contribution to literature is threefold. First, the existing
between age and performance has also been explored in certain studies on cricket primarily focus on the age-performance
other sports including road-cycling (Kholkine et al., 2023), hockey relationship in the different formats of the more popular Twenty-
and golf (Berry et al., 1999), tennis and boxing (Lehman, 1953), 20 cricket (Saikia et al., 2019, 2015), and test cricket—considered
running and swimming (Ericsson, 1990), 25 Olympic sports events, the gold standard of assessment of cricketer quality (Mangan,
the game of chess (Berthelot et al., 2012), and basketball (Hach et al., 2014). We focus on ODI cricket because it is mix of the pure
2023).1 The interplay between experience and fitness often shapes and the popular formats of cricket and has strategic elements of
the performance trajectory of individual players, characterized both these extremes (with twenty-20 which is about extreme hard-
by an initial ascent to peak performance followed by a gradual hitting and test cricket which is demanding enough to test the
decline, leading to inverted-U shaped performance-curve, often mettle of a cricketer). Thus, ODI cricket is a beautiful game that
also observed in various domains of human endeavor (Ericsson attains a fine balance between aggression and tact. In addition,
et al., 1993; Berthelot et al., 2019). our primary reason to specifically look at ODI cricket has to
A senior player often takes a tactical approach to the game do with the sheer volume of data in comparison to twenty-20
with strategic thought (Silva et al., 2020). Younger players are fitter international cricket (a format that is only two-decades old) and
with better reflexes that aid their approach to a sport (Schulz and test cricket which is played with a far lesser frequency in a year (in
Curnow, 1988). We econometrically look for the age (range) of comparison to ODI cricket).4 The research by Jamil et al. (2023)
peak-performance at which the fine balance between fitness and that looks at bowlers’ performances specifically in 2019 ODI cricket
experience is attained. That sportsmen attain a peak performing age World Cup is the closest to ours. We look at both bowlers’ and
before their performance dips is an observation that even applies batters’ performances for all ODI matches (including world cups)
to cricket. We look at performance data for all players in One till 2000 (for cleanest results as explained in Section 3). Second,
Day International (ODI hereafter) cricket matches played during while previous studies mostly focus exclusively either on bowling
1971–2000 to find the age-range at which this peak occurs. In the performances (Crowther et al., 2023; Thorley, 2021; Jamil et al.,
section on the description of data, we provide detailed reasons for 2023) or exclusively batting performances (Mangan, 2014), this
the choice of our period of study. study estimates the peak performing age of both batters and bowlers
Santhanam and Balasubramanian (2009) observed that in ODI cricket. Finally, we provide a formal economic framework
between the age of 32 and 34 years, the performance-levels of some to explain the trade-off between youth and experience. As players
of the greatest senior-cricketers (including Gordon Greenidge, begin their career, initially the marginal benefit from staying fit is
Garfield Sobers, Adam Gilchrist, Sunil Gavaskar, Rahul Dravid, greater than the marginal cost of inexperience. As a player ages
Vivian Richards and Sachin Tendulkar) declined significantly. initially, the marginal benefit of experience potentially outweighs
They call it the “thirty-three effect”. Even for a top player like Sachin the marginal cost of becoming less fit (and if fitness levels also
Tendulkar, Mirza (2023) notes that being fit “. . . when you’re 40 get better, then the performance unambiguously gets better since
and . . . putting your body through intense physical activity for 24 experience-gain could complement fitness in this initial age-range).
years couldn’t have been easy.” In trying to uncover these dynamics between fitness and experience
Schulz et al. (1994) and Bradbury (2009) report that using ODI cricket data, we estimate the age where the marginal
performance peaks between the ages of 27–29 years in their benefit from experience equals the marginal costs of declining
research on baseball. In addition to studies that directly relate sports fitness. This information could aid team selection committees in
performance to fitness (Cureton, 1956; Lucifora and Simmons, finding the best mix of young and senior players for performing
2003), there are studies which emphasize the role of experience, in tournaments. Experienced players know their responsibilities
mental-ability and technical skills in improving performance well and can help youngsters be strategic. At the same time, an
(Saikia et al., 2019). These characteristics are seen in senior players infusion of youngsters can improve the overall average fitness of the
who tend to better handle pressure.2 Raman (2009) analyzes how team. In general, any team in any group-sport must find a judicious
senior players employ elaborate skillsets to keep performing despite mix of youth and experience. Note that the duration of an ODI
falling fitness-levels.3 Ganguly (2018) in his autobiography explains cricket match is by far the maximum (close to 8 h) among all other
well-known team-sports.

1 See Allen and Hopkins (2015), Berthelot et al. (2015), and Marck et al.
(2019), and the citations therein. seasons (see: https://www.espncricinfo.com/story/indian-premier-league-
2 South African cricketer Graeme Smith is an excellent example (see: 2011-the-ipl-xi-517150).
https://www.espncricinfo.com/story/older-wiser-leader-360426). 4 During our period of study, about 28 test matches were played per year
3 Many experienced cricket players like Chris Gayle, Virat Kohli, Mahendra on an average. This is almost half that of the average number of ODI matches
Singh Dhoni and others have excelled in various Indian Premier League (IPL) (about 55) played per year during the same period.

Frontiers in Behavioral Economics 02 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

These ideas are remarkably applicable to any work environment

including those that specifically involve intergenerational
workforce diversity. We often see an integration of years of
experience (from Generation X and millennials) and youth full
of energy (Generation Z), where the seniors bring in the pool of
knowledge to tactfully handle or pre-empt any difficult situation
whereas the youth is more adaptable to changes and willing to
try new ideas (Lewis and Wescott, 2017). This synergy between
experienced agents and younger talent is very crucial to build a
culture of knowledge-sharing, adaptability, and collective-growth,
positioning organizations for sustained success in an ever-evolving
market landscape.

2 A model of optimal performance in

ODI cricket FIGURE 1
Talent constraint.
For simplicity, we consider two important factors in any
sport—the ability to maintain peak fitness (youth) and experience
the north-east direction) the talent constraint will be. In Figure 1,
(respectively called Y and X). The ability to draw value from
the vertical intercept (D) represents a player’s debut (maximum
experience (X) is V (x) and the ability to draw value from youth

permissible youth and no experience) and the horizontal intercept
(Y) is U y , which are essential ingredients (inputs) for sports-
(R) represents retirement (maximum permissible experience and
performance. Thus, talent is the ability to draw from experience
no youth). Time is a unidirectional variable, so players involuntarily
and/or fitness. We assume the following:
move from left to right along the talent constraint (clearly, nobody
has the option of choosing a fitness level of their 20-year-old self
a) V(0) = 0 and U(0) = 0
when they are 30), which is shown by the arrows. It is here that our
b) V ′ > 0 and U ′ > 0
dy ‘talent-budget’ differs from a traditional budget-line (or an iso-cost
c) dx
dt ≥ 0 and dt ≤ 0 (Experience increases with time (t) and
curve) where one could substitute X for Y or Y for X (i.e., move
youth (fitness) decreases)
along the budget-line in either direction) at the given market-rate
Assumption (a) simply means that there is no value
(represented by the price-ratio). An agent therefore strictly moves
from experience if one has no experience, just as there is
only from left to right along our “budget line”.
no value from youth if a player has no youth (say for he is
In addition, we impose that V ′′ ≤ 0 and U ′′ ≤ 0, to avoid
on the verge of retirement).5 According to assumption (b),
for example, the potential absurdity that an individual may never
the value derived from experience increases with experience,
need to retire if the value from experience perpetually increases
and the younger players derive greater value from youth (i.e.,
at an increasing rate (potentially leading to “convex” budget-
when age is lower). Assumption (c) simply means experience
constraints). Thus, the value derived from experience increases
increases and youth declines in time.
in experience, but at a diminishing rate. This is analogous to a
budget-constraint where the price of acquiring a good in the market
A description of ODI cricket is given in the Appendix. We
progressively increases with increased quantities of it bought.
assume that each ODI cricket player has a divine (non-negative)
Viewing the above diagram from the perspective of a single
endowment of talent or unobserved ability (denoted by A). It can
player, we abstract away from the obvious positive externalities
be argued that great players have higher endowments of talent than
associated with experience. It is clear that the experience of a senior
average players. The value that a player derives from experience (X)
player could benefit everyone in the team and not just the player
or youth (Y) (or both) depends on this endowment A, just like the
himself (who may in turn be relatively less-able to retain peak-
amount of labor or capital (or both) that a firm can employ (for
fitness in comparison to the younger players). In Figure 1, we have
their productive value) will depend on the total costs or expenditure
also abstracted away from any discontinuity by assuming that each
that the firm can afford. The greater the ability of an individual to
cricket player plays continuously without any breaks in his career.6
derive value from fitness or experience (or both), we say that the
The model discussed so far is however, rich enough to incorporate
more talented the individual is.
We thus represent the talent constraint (analogous to the iso-


cost line) as: V (x) + U y = A, for a given level of talent or 6 In the real-world such gaps are very common. Former Indian captain
ability A (shown in Figure 1). The functional forms V and U have Sourav Ganguly, for instance, debuted in international cricket in 1992 but did
a bearing on the shape and the curvature of the talent constraint. not play international cricket for next four years and rejoined the Indian team
The talent endowment A (> 0) in turn decides how high (in in 1996. However, he continued his career in the national, domestic and state
level matches during these four years. Thus, claiming that experience was

5 We are only looking at men’s ODI cricket. Female cricket was not well zero for these four intervening years is not necessarily accurate (even though

established during our period of study and hence we do not have sufficient it is unambiguously not international experience). For simplicity, we assume

number of data points for meaningful analysis. that players have no break in their career. We will later discuss why we have

Frontiers in Behavioral Economics 03 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

FIGURE 2
Talent constraint with career breaks.

temporary or permanent breaks (say) due to career-threatening or theory, we expect the ICs between youth and experience to slope
career-ending injuries. These cases are respectively covered in the downward in the region where both are productive (i.e., each
two panels of Figure 2. contributes positively to performance). If there is “too much” of
In Figure 2a, a player accumulates experience up to point B youth, (say beyond y0 in Figure 3a where the players are very young
after which there is a break in his career (say, due to an injury), so and bodies are still developing without yet any built-up stamina),
he “loses his youth” (by a magnitude of y0 – y1 ) for the duration then youth could potentially take away from performance (“over-
of his career-break before returning to play until his retirement bowled” young bowlers for instance, are likely to be prone to
at R. Note that there is no necessary addition to his experience more injuries which could prevent them from maintaining their
during his career-break. In Figure 2b, the player in question gains performance). One could make the same argument for experience
experience up to the point E after which his career comes to an if it becomes unproductive beyond the level x0 . In the region of
abrupt end either due to an injury or for some other reason (so all youth beyond y0 or experience beyond x0 , the IC slopes upward.
youth is lost and E coincides with R on the horizontal axis marking In Figure 3b, we present the possibility that too much experience,
the end of a player’s active career). One very real example where while not necessarily unproductive beyond x0 could cease to add to
individual players experienced an end of their careers involved performance in any meaningful way.
the infamous match-fixing scandal in the period 1999–2000. This In the real world, we do not usually observe a region with
period witnessed a permanent suspension of several cricketers from too much of both youth and experience since we do not expect
a few nations that eventually therefore needed a mass-addition of very young people to have a great deal of experience. For now, in
fresh players selected to fill up the team-positions rendered vacant what follows, we will focus on the region where both youth (ability
due to the suspended cricketers. This structural break around to maintain peak-fitness) and experience are “productive inputs”,
the year 2000 that caused a major disruption in the cricketing we have downward-sloping, smooth and convex performance
world (following which the continuing senior sportsmen and newly indifference curves (ICs). The higher ICs (again in the north-east
selected sportsmen played in cricket tournaments under tighter direction) are associated with higher levels of performance. Note
scrutiny) prevents any smooth data analysis from the end of the that these ICs will be vertically-biased (or flatter) particularly for
1990s to the beginning of 2000s and therefore majorly contributes bowlers (and even more so for fast-bowlers). This is because the
to our decision to only look at the time-period before 2000 for relative value of fitness is higher for fast-bowlers than it is for
a meaningful analysis under mostly similar conditions. We will batters. Similarly, performance ICs for batters will be horizontally-
elaborate more on this in the discussion of our choice of time- biased (or steeper). In Figure 4 we illustrate these ideas. Note that
period of study. more experience for a given level of youth is likely to add to
We now come to the development of performance indifference performance (indicating a movement to a higher performance IC).
curves (ICs) for cricketers. We model performance (P) as a function This explains why many players wish to avoid “resting out” through
of youth (Y) and experience (X) so that P = f (x, y). A given level tournaments and prefer continued match-play since that could
of performance can be attained with varied combinations of youth potentially help maintain form and consistency. On one occasion,
and experience shown by the “performance indifference curves” when Virender Sehwag was a newcomer, he was assured a place in
(Figure 3).7 In Figures 3a, b, we provide a general representation the team even if he underperformed batting in a position out of his
to account for various possibilities. Just as in standard economic then comfort-zone by Sourav Ganguly who was then the captain of
the Indian team.
avoided international cricketers with frequent and long career-breaks (in the It is clear that these ICs are two-dimensional projections
subsection on data description). of performance measured along the third dimension. Better
7 We can also think of youth and experience as inputs and performance as
performing cricketers achieve a higher vertical measure of
output.
performance (again, measured along the third dimension) for

Frontiers in Behavioral Economics 04 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

FIGURE 3
Performance indifference curves.

FIGURE 4
Performance indifference curves with productive youth and experience.

any given amounts of youth and experience. Finally, there exists to himself. As has often been observed, seasoned players do
a critical value of experience and youth (fitness) where the assume mentorship roles, guiding younger teammates through
performance peaks. This is the point of tangency between the their insights (Chopra, 2012)—this knowledge transfer (Funasaki
talent constraint and the performance indifference curve (shown in et al., 2024) has proved to be very important time and again.


Figure 5). In Figure 5, x∗ , y∗ is the point where the performance In addition, when opponents focus most on the seasoned
 ′ ′
of a batter peaks and x , y is the corresponding point for a players (Doshi, 2016), younger teammates get the advantage of
bowler. Under these conditions, the peak performance (P0 ) for a playing without pressure, just like Sehwag, a relatively younger
batter will take place at a higher value of experience and lower value player once said in relation to a more established senior
of youth, and conversely for a bowler (P1 ). Starting from the point player (Tendulkar):
of debut (on the vertical intercept), the arrows show that as we go
down, the benefits from experience outweigh the loss of youth, and “When I’m batting with Sachin Tendulkar, there’s very less
thus, higher levels of performance are attained till a peak P0 or P1 pressure on me because the opponents are all concentrating on
is reached, beyond which, the loss of youth outweighs the benefits getting him out”.8
from experience.
Finally, it should be noted that even after reaching the Similarly, in any organization the experienced team-members
peak performance-level, a player might not consider retirement can often take the pressure to let the youth flourish and grow to
because one’s own experience has positive externalities that other build their own experience (Argote et al., 2000). It is nonetheless
younger players (and by extension, the team as a whole) can clear that below-peak performance for those exceptionally talented
benefit from—and these positive externalities in the aggregate, senior players may still exceed the peak performances of (say)
may outweigh the effects of lower fitness (in comparison
to younger players) of the said player of experience. The
argument is easily seen in an analogous setting where the 8 This quote can be found in: https://www.bcci.tv/articles/2022/
above-mentioned experienced player would have in his own news/93456#:∼:text=When%20I’m%20batting%20with,concentrating%20
youth borrowed from the experience of those who were senior on%20getting%20him%20out

Frontiers in Behavioral Economics 05 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

FIGURE 5
Peak performance.

average or below-average young players. Thus, even if one were to better gauge the pure effect of age on the performances of both
completely abstract away from potential mentorship roles, those batters and bowlers (i.e., when the rule-changes were less frequent).
exceptionally talented senior cricketers may remain in the team Our choice of the time period between 1971 (when the first
simply on their own accord. Sachin Tendulkar for instance, may ODI match was ever played) and 2000 is primarily driven by three
not have batted in 2012 (when he was almost 39 years old) as well major reasons. First, the match-fixing scandal of 2000 needed an
as he did in his prime, but he still continued to remain in the team immediate solution that in turn required a major diversion of the
because he was still contributing significantly more runs to the team resources of the International Cricket Council (ICC), along with the
total in comparison to several younger players who eventually saw cricket boards of several countries including India and South Africa
much shorter careers. before the problem became more pronounced at a global level.
This included the announcement of temporary and permanent
suspensions of several cricketers in that year and led to an almost
3 Data and descriptive statistics immediate selection of younger cricketers in a few countries. This
was as good as a structural-break in ODI cricket (since several
It has often been argued that since 2000 cricket has cricketing nations saw an almost-simultaneous reshuffling of the
progressively become a batter’s game. Changes in the laws of mix of players in their teams as we will elaborate more on later).
the game (fielding restrictions, powerplay, free hits from illegal Second (and in some ways, a consequence of the match-fixing
deliveries called “no-balls” and limits on bouncers in an over) episode) ODI cricket saw frequent change of rules as noted in an
along with the evolution of the quality of cricket bats have created article titled “Cricket: ODI rules overwhelmingly favor batters” that
new playing conditions (Noorbhai and Noakes, 2015). All these appeared in a national daily of New Zealand that clearly states
developments in addition to flatter pitches and shorter boundaries, that “no major sport has tinkered with its rules as ceaselessly as
have mostly supported the batters. Players who played before 2000 one-day international cricket . . . Since 1999 many rules have been
can no longer relate to the modern form of the game.9 For instance, introduced and subsequently been scrapped . . . ”.10 Third, and also
when the former Indian captain Sourav Ganguly was asked to noted in the above-mentioned article, with the advent of twenty-
compare the teams that played in the early 1990s with that which 20 cricket (the first ever international match was played between
played in the early 2000s, he elaborated on the pointlessness of the Australia and New Zealand in 2005), both the other formats (ODI
comparison in the following words (Ganguly, 2018): and tests) have witnessed a marked change in players’ attitudes.
For instance, Virender Sehwag’s score of 319 runs in 304 balls
“Two different generations can never be compared. The eras (the fastest test-triple-hundred till then) in a test match against
were different. The mindsets were different.” South Africa in 2008 makes the influence of twenty-20 cricket
(where trying to hit every single ball for a boundary is the norm)
Interestingly, this difference between what Ganguly counts evident in even test cricket (let alone ODIs). Note that the talks for
as two separate eras were only about a decade apart. Given the this new format of cricket initiated well in the late 1990s possibly
regularity of additions to existing rules and changing playing explaining why players like Sanath Jayasuriya, Gary Kirsten, and
conditions since 2000, we only look at ODI matches played before Romesh Kaluwitharana revolutionized aggressive batting during
2000 under more controlled and similar playing conditions to the world cup of 1996 (following which the average strike-rate

9 Also, the audience today looks forward to new batting records rather 10 The article can be accessed through the following link: https://www.
than new bowling records since the former are more frequent than the nzherald.co.nz/sport/cricket-odi-rules-overwhelmingly-favour-batters/
latter and the global audience has formed its expectations accordingly PKS5IIBDZGADBP7XWIYWACM4ZE/#:$\sim$:text=Since%201999%20many
(see: https://www.financialexpress.com/archive/on-the-elements-of-anti- %20rules%20have%20been%20introduced,given%20the%20option%20of
marketing/1163105/). %20when%20to%20implement

Frontiers in Behavioral Economics 06 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

TABLE 1 List of variables used.

observed during 1998–2000 was 65.7 runs per 100 balls which is
almost 20 runs higher than the average strike-rate of 45.9 observed
Dependent Description
during 1971–73). In addition we observe clear structural breaks for variables
instance, in the average number of boundaries (fours and sixes).
Runs Total runs scored by a batter in a particular match
The average number of boundaries per match during 1971–99
was 32.55. Within this period, the average number of boundaries Strike-rate Average runs scored for every 100 balls

during the last 6 years (1994–99) was 35.45 (only 2.9 boundaries Wickets Total wickets taken by a bowler in a particular match
per match higher on an average in comparison to the entire period). Economy-rate Average number of runs conceded for every over bowled
The average number of boundaries per match during 2001-03 was
Independent Description
40.21 (almost five boundaries more than the average during just the
variables
preceding 5 years).
For our study, we focus on 210 cricketers who played ODI Age Age of the ith player in his jth match

matches during 1971–2000.11 We consider only cricketers who Country A dummy for the country represented by the ith player
have played more than 25 matches and have careers spanning Opponent A dummy for the opponent country an ODI match is played
at least 5 years with at most one break-period of at least 1 year against
(in the Appendix we argue that our selected sample is fairly Batter A dummy for whether the ith player is a batter
representative of the population of cricketers that played during our
Bowler A dummy for whether the ith player is a bowler
study period—see Supplementary Figure A2). The reason for our
Wicket-keeper A dummy for whether the ith player is a wicket-keeper
criterion for at most one-break period of over a year of international
cricket is that we want the experience data to be as continuous as All-rounder A dummy for whether the ith player assumed at least two of
the above roles
possible (the previous section models experience as a continuous
variable). The data consists of 22,114 (14,112 for batters and 8,002 Fast-bowler A dummy for whether the ith bowler is a fast-bowler

for bowlers) observations and has been compiled from Wisden, ODI workload Number of ODI matches played by the ith player in the year
Cricinfo and ESPN. Table 1 gives the description of variables used of the jth match

(a brief description of ODI cricket is provided in the Appendix). Test workload Number of test matches played by the ith player in the year
For any given match we look at the number of runs scored of the jth match

and the number of wickets claimed respectively as key performance Balls faced Number of balls faced by the ith batter in his jth match
indicators for batters and bowlers. Other measures such as the Balls bowled Number of overs bowled by the ith bowler in his jth match
strike-rate (a measure of how quickly a batter scores his runs
Venue A dummy for the venue country of a match
and is defined as average runs scored per 100 balls faced) and
Year of match A dummy for the year a match was played in
the economy-rate (the number of runs given away for every six
deliveries bowled) had only become more relevant towards the very Experience ODI matches played by the ith player before his jth match
extreme end of our period of study when the rate at which runs
were scored or given away actually began to matter in this limited-
over format. In any case, we provide results on these measures in TABLE 2 Descriptive statistics.
the Appendix. Table 2 provides the descriptive statistics. We are
interested in the following hypothesis. Variables Average Min Max
H 0 : The average performance of an ODI cricket player is Runs 28.08 0 194
independent of age of the player.
Strike-rate 63.67 0 400
We begin with a preliminary look at the average batting and
bowling performances [measured respectively on the vertical axes Wickets 1.23 0 7

of panels (a) and (b)] by age (on the horizontal axis). In Figure 6a, Economy-rate 4.19 0 24
each point in gray represents the batting average of a player at a Age (in years) 28.34 14 42
given age. For instance, if player i scored m runs and was dismissed
ODI matches played by a cricketer 8.88 0 41
k times during all the matches he played when he was 21 years old, in a year
then his batting average at the age of 21 is m/k which is represented
Test matches played by a cricketer 5.16 0 18
by one dot in gray. We similarly look at other batters’ averages when in a year
they were 21 and represent them with other gray dots. The average Proportion of batters (%) 38.10 - -
of all these average runs scored by all players in our data at their age
Proportion of bowlers (%) 33.80 - -
of 21 is uniquely represented by the black dot—this represents the
expected batting average of a batter at the age of 21. We similarly Proportion of all-rounders (%) 28.10 - -

look at the expected batting averages at other ages for all players
in our sample and immediately see that there are potential reasons
to reject our null hypothesis H 0 since the batting averages clearly out the expected bowling average for each age-value (number of
vary with time. We do a similar analysis for bowlers and work runs conceded to the opponent divided by the number of wickets
claimed) and display the same in Figure 6b—we continue to see
potential reasons to reject H 0 .
11 The first ever ODI match was played on January 5, 1971 between Note that higher batting averages for batters and lower bowling
Australia and England in Melbourne. averages for bowlers represent better performance. As a first step,

Frontiers in Behavioral Economics 07 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

FIGURE 6
Batting and bowling averages (individual and overall).

FIGURE 7
Crude fit of batting averages and bowling averages.

we do a crude smooth-curve-fitting exercise on the exact data in jth match. For bowlers, the key dependent variable is the number
Figure 7. of wickets claimed. Xij is the vector of player-specific and match-
It seems that average batting improves with age till it reaches a specific covariates associated with the coefficient vector λ. uij
peak-age before declining as a player ages further. This is also true denotes the stochastic error term. The key variable of interest is Age.
for average bowling performance. To further look into the specific We also include Age2 to account for the non-linear effects of age on
ages of peak performances, we now turn to regressions. the performance of a player based on our observations in Figures 6,
7. For now, we do not take experience as a separate independent
variable because for any given player, age and experience are likely
4 Analysis of results to be correlated (experience could only increase as a player ages).
Further, as pointed out in text footnote 6, some players might have
We begin with the regression (Equation 1), below a break in their career, defining experience for such players may
not be free from ambiguity. We however, also note that players
yij = α0 + α1 Ageij + α2 Age2ij + λXij + uij (1)
who start their careers earlier are expected to have more experience
Here, yij denotes the performance outcome of player i in his at a younger age relative to others who start their careers later
jth match. For batters, we look at runs scored by batter i in his (for instance, the Indian cricketer Sachin Tendulkar started his

Frontiers in Behavioral Economics 08 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

TABLE 3 Determinants of runs scored.

career in 1989 at the age of 16 whereas the Sri Lankan cricketer
Sanath Jayasuriya who also debuted in 1989 was 20 years old—
Independent Least Fixed- Fixed- Fixed-
hence we will expect Tendulkar to have more experience than variable squares effects effects effects
Jayasuriya at a comparable age). We exploit this variation across (1) (2) (3) (4)
players and separately include both age and experience in some of Age 3.0235∗∗∗ 5.8535∗∗∗ 5.7432∗∗∗ 0.9919∗
our regression specifications (mostly deferred to the Appendix) that (0.5499) (0.7748) (0.7645) (0.5534)
largely lead us to very similar conclusions (as we will discuss later). Age-squared −0.0499∗∗∗ −0.1000∗∗∗ −0.0983∗∗∗ −0.0184∗∗
In the above regression (1) we also need to account for the fact (0.0095) (0.0134) (0.0133) (0.0072)
that some players are high-ability performers while others not so All-rounder −1.7905∗∗∗ – – –
much in terms of their intrinsic abilities (Baker and Wattie, 2018). (0.5924)
Menon (2012) defines what it is to have sheer talent:
Balls faced – – – 0.7456∗∗∗
(0.0133)
“It is what you can’t learn, hone or teach. . . a skill can
Constant −16.3148∗∗ −55.811∗∗∗ −50.7993∗∗∗ −14.414∗∗∗
be learned, talent can’t. . . talent is Garry Sobers, talent is
(7.8487) (11.0134) (10.7075) (13.8646)
Jimi Hendrix. . . ”
Observations 14,112 14,112 14,112 14,112

To the extent that more talented players derive greater benefit R-squared 0.0026 0.0018 0.0048 0.8197

from experience (borrowing from our theoretical model), one Player fixed-effects No Yes Yes Yes
would expect the main regressors (including age) to be correlated Country/ opponent No No Yes Yes
with unobserved ability or talent (i.e. more talented players evolve dummies
better with age in comparison with average players). In order
Year of the match No No No Yes
to difference out the effects of these unobserved abilities (or dummies
talent), we run a fixed-effects regression shown in the specification
Experience No No No Yes
(Equation 2).
Workload No No No Yes
yij = α0 + α1 Ageij + α2 Age2ij + λXij + Ai + uij (2) ∗ , ∗∗ , ∗∗∗
report significance levels at the 10%, 5%, and 1% respectively. Robust standard errors
are reported in the parentheses.
Where Ai denotes the intrinsic talent/ability of player i
(our player-specific fixed effect). This is where the fixed-effects
specification (Equation 2) becomes useful. The term Ai above Since different countries have varied cultural emphasis on
represents the player-specific fixed-effects and represents the talent different sports, we also include country (represented by the
endowment (or unobserved ability) discussed in Section 2. We now player) dummies and opponent-country dummies in columns 3
discuss the regression results. and 4. For instance, Holland has a weaker culture for cricket in
comparison with Australia just as Brazil has a more pronounced
culture for football in comparison with India (Adelman, 1986;
4.1 Effect of age on the performance of Kirsch, 1991). Home-field advantages may affect the performance
batters of players through many factors like familiarity with the home-
ground and home-conditions (Barnett and Hilditch, 1993; Moore
Table 3 reports the regression results where the dependent and Brylinsky, 1995; Pollard, 2002; Morley and Thomas, 2005).
variable is the runs scored by the ith batter in his jth match. Column To account for these effects we also included venue-dummies as
1 reports the crude least-squares estimates for Equation 1, and the regressors. These venue-country dummies however, were found to
regression results for Equation 2, are reported in columns 2, 3, and be highly correlated with (a linear combination of) the country-
4 (with progressively more controls). dummy or the opponent-country dummy. This is because in
We see that Age is associated with a significantly positive any ODI cricket match, more often than not one of the two
coefficient in each of the four columns above and the coefficient competing countries is also the host country (more specifically,
for Age2 is (again significantly) negative. We therefore infer an country dummy + opponent dummy = country host dummy +
inverted U-shaped relation between age and performance just as opponent host dummy). World cup matches where the host is
seen in Figure 7. We reject the hypothesis H 0 and infer that often a third country are exceptions to this rule, but the world cup
performance first increases to reach a peak with respect to age tournament happens once in 4 years and this relative infrequency
before it eventually diminishes as fitness-levels significantly decline does not significantly reduce the strength of the linear relation
(Kalen et al., 2021; Rey et al., 2022). We also include information discussed here.
on whether a batter also contributed to bowling or wicket-keeping During our period of study (1971–2000), the game of ODI
as a preliminary control for general workload (in column 1). cricket experienced a few regulatory changes (introduction of
The significantly negative coefficient of the “all-rounder” dummy day-night cricket, changes in number of overs in 1987, fielding
variable (as defined in Table 1) suggests that involvement in a game restrictions introduced in 1996, and the Duckworth and Lewis
beyond batting, generally has a crowding out effect on the ability method in 1998). These changes may have affected the outcomes
to bat. Note that there are exceptions to this rule since Kapil Dev, (Silva et al., 2015). We include year-of-match dummies to account
Richard Hadlee, Ian Botham and Imran Khan were exceptionally for these effects (column 4). The significant coefficients on the age
gifted all-rounders. variable tell the same story.

Frontiers in Behavioral Economics 09 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

TABLE 4 Determinants of wickets claimed.

In column 4, we also further include controls for experience
and workload. Experience is measured in terms of the number of
Independent Least Fixed– Fixed– Fixed–
ODI matches played by the ith batter leading up to his jth match. variable squares effects effects effects
We also include an experience-squared term to account for any (1) (2) (3) (4)
non-linear effects. Both of these regression terms are statistically Age 0.104∗∗∗ 0.0915∗∗ 0.0928∗∗ 0.1287
insignificant and do not contribute to the main results (possibly due (0.028) (0.0412) (0.0414) (0.0847)
to the correlation of experience with age). For workload, we first Age–squared −0.0019∗∗∗ −0.0018∗∗ −0.0018∗∗ −0.0029∗∗
individually included the number of test and ODI matches played (0.0004) (0.0007) (0.0007) (0.0013)
in the year of player i’s jth match (see Table 1). While the number of All–rounder −0.0944∗∗∗ – – –
test matches has no significant effect on batting-performance, the (0.0296)
number of ODI matches played is associated with better batting
Overs bowled – – – 0.1564∗∗∗
performance (captured in a significant positive coefficient). This (0.0065)
may hint that only relevant experience in ODI cricket matters
Constant −0.0796 0.1294 0.5465 0.4887
in ODI performance. However, there is a potential endogeneity (0.3998) (0.5708) (0.5986) (1.8743)
issue associated with the inclusion of ODI workload. Performance
Observations 8,002 8,002 8,002 8,002
in a particular match (and by extension, in a given tournament)
could determine the chances of selection (and hence workload) R–squared 0.0043 0.003 0.0005 0.0198

in the following matches (or tournaments) up to an immediate Player fixed–effects No Yes Yes Yes
time-period. In the fixed-effects regression of column 4 therefore,
Country/ opponent No No Yes Yes
only the test-match related workload is retained (which remains dummies
insignificant). Finally, note that the R-squared value in column 4
Year of the match No No No Yes
is significantly higher than it is in the previous three columns— dummies
the number of balls faced contributes most of the explanatory
Workload No No No Yes
power to observed variation in runs scored (also associated with
a significantly positive coefficient). Experience No No No Yes

When a batter starts his ODI career, his age-performance Bowler–type No No No Yes
relationship is positive. Overtime this performance-curve reaches ∗ , ∗∗ , ∗∗∗
report significance levels at the 10%, 5%, and 1% respectively. Robust standard errors
a peak and the performance dips thereafter. After the peak, are reported in the parentheses.
though the batter is more experienced, his physical capabilities
decline (Saikia et al., 2019). Based on our regression coefficients performance for fast bowlers is not qualitatively different from that
that suggest an inverted U-shaped relationship, it is possible to for spinners. Experience and workload (introduced in Table 1) in
determine the age at which this peak occurs by finding the age column 4 are treated in a manner similar to the previous set of
consistent with the global maximum of the performance-curve. regressions reported in Table 3. The results remain largely robust
According to our regressions for runs, the age at which the peak to the introduction of new variables and alternate specifications.
hits is between 27 and 30 years. Our findings are consistent with
a top international batter, Rahul Dravid’s self-assessment. He once
said (Ugra, 2012): “. . . I was playing my best cricket, at 28–29.” We 4.3 A final robustness check
next examine the results for bowlers.
So far, we have only looked at exactly one measure for batting
and bowling performances. We now extend our analysis to cover
strike rate (for batters) and economy rate (for bowlers). These
4.2 Effect of age on the performance of variables are defined below.12
bowlers
(Runs scored by batter i in match j)×100
Table 4 shows the effect of age on the number of wickets StrikeRateij = (Number of deliveries faced by batter i in match j)
claimed. We generally observe the same results. Age is associated
with a statistically significant positive coefficient and age-squared is
associated with a statistically significant negative coefficient.
EconomyRateij =
Based on this, we infer that on an average, bowlers attain a
Runs Conceded by bowler i in match j
peak at a younger age relative to batters—somewhere between (Number of deliveries bowled by bowler i in match j)/6
22 to 26 years. This might be because bowling (especially fast
bowling) requires speed, power and other skills that typically peak In the Appendix, we look at a joint-estimation of regressions
in an individual’s early twenties (Chomik and Jacinto, 2021; Noakes with runs and strike-rate as dependent variables on our exogenous
and Durandt, 2000). So, the complementarity between fitness and
experience is higher for bowlers than for batters. These results 12 Note that economy rate is defined as the average runs given away per
are also consistent with the economic framework developed in ‘over’ bowled. Since the definition of an “over” was standardized in the latter
Section 2. We also include a dummy variable called Fast-Bowler part of the 1970s (to be equal to six consecutive deliveries by a bowler), we
(= 0 for spinners) to account for bowler-type in column 4, but suitably modify our definition of economy-rate as above so that these figures
the insignificance of that variable indicates that the age of peak remain comparable for all bowlers in our study period.

Frontiers in Behavioral Economics 10 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

regressors for the batters. For bowlers, we look at a joint-estimation

of regressions where the dependent variables (wickets and the
economy-rate) are regressed on all exogenous variables. The
idiosyncratic error-terms from the regressions with runs scored and
strike-rate as dependent variables could potentially be correlated.
This is because a batter who wishes to score high, must also
score runs quickly (i.e., have a high strike-rate) given a limited
number of overs. The approach to batting is often player-specific.
Similarly, bowlers do not mind giving away more runs as long as
they can get quick wickets. On the other hand, tighter bowling
(i.e., with a lower economy rate) is often less rewarding in terms
of wickets claimed because of the low-risk approach. The approach
to bowling depends on an individual bowler. As a robustness check,
we perform a joint estimation of runs-scored and strike-rate on all
other exogenous variables (column 1), and also of economy-rate
and wickets (column 2). The joint estimation results are reported FIGURE 8

in the Appendix, which remain consistent with the previous Selection of Sourav Ganguly in 1996 was a great decision.

findings—runs scored and wickets taken continue to peak within

the age-ranges discussed previously. Strike-rate peaks at around 30
years. It is believable that economy-rate gets better with age since
bowlers who are more experienced tend to develop tighter ways of
bowling (with experience) and introduce the necessary variation in
their bowling to keep the run-rate of the opposition low.

5 A concluding remark with a

retrospective policy-analysis
Finally, borrowing from the theory and the empirical sections
of this paper, we do a policy analysis of actual selection decisions
taken during our study period. Former Indian captain Sourav
Ganguly had only one (international) career-break during our
period of study that exceeded 1 year (and hence he is included in
our sample of cricketers as per our sample-selection criteria). More FIGURE 9
specifically, he made his ODI debut in 1992 and was immediately Imran Khan’s comeback was key to the success of Pakistani cricket.
dropped by the selectors from subsequent tournaments. He faced
a career-break for about 4 years before making an international
comeback in 1996 after which he remained on the side for each
of the 5 years till the end of our sample period (1996, 1997, 1998, y1996 , and eventually to y2000 along the Youth axis—one’s youth
1999, and 2000). In fact, he captained the Indian team in the period declines with increasing years. Thus, in this case the youth level
1999–2005. Following the suspension of a few players (including at (say) 1992 was much higher than the youth level at 2000 so
the previous captain) on account of match-fixing, Ganguly was y1992 > y2000 . Ganguly’s key role in mentorship and the process of
entrusted with the task of building a new team with fresh and team building could be seen in the careful selection of new players
young players who were eventually instrumental in winning the and the decision to back them with the assurance that they will be
2011 world cup for India (although Ganguly himself retired a few retained despite some initial performance setbacks (if any). This
years before 2011). This included players like Yuvraj Singh (the aided the new players to express themselves in their sport freely
player of the tournament in this 2011 world cup) and MS Dhoni without the fear of being thrown out of the team. The end result
(who captained the world-cup winning Indian team in 2011) along was a more aggressive Indian team that was far different from the
with other great players such as Virender Sehwag (who was known underdogs of the early 1990s. Even younger players who were not
for his explosive batting style). directly mentored by Ganguly found him inspiring—they pushed
Based on Figure 8, it is clear that the decision to re-introduce each other and bettered their own performances and this culture
Ganguly into ODI cricket in 1996 was a great move on the part of eventually evolved into positively-reinforcing cohort-effects—to the
the Indian selectors. Although Ganguly’s personal peak occurred extent that some of the Indian cricketers today openly take pride in
in 2000 when he touched a batting average of almost 50 runs (he maintaining their fitness levels.
was 28 at that time—consistent with our empirical findings and The story of Imran Khan (Figure 9) was not so significantly
represented by a tangency point of peak-performance in Figure 8), different. He decided to retire in 1987 but was requested by the
he continued to play (mostly as the captain of the team) beyond (then) President of Pakistan to resume his play from 1988. He
his personal peak. Note that the youth level declines from y1992 to ended up captaining the Pakistan cricket team that won the world

Frontiers in Behavioral Economics 11 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

cup in 1992. Note that his personal peak bowling was observed “It is the confidence of doing well consistently, and the
in 1975 (represented by the tangency in Figure 9—right before maturity gathered over the years. With experience you learn to
he turned 24), but under his leadership, the careers of younger trust your game more than you did as a youngster. I have looked
players like Wasim Akram and Waqar Yunis flourished—these at all areas of my game and worked hard on all of them. I have
were players who eventually went on to make great contributions really worked on my physical fitness.”
to Pakistani cricket. Wasim Akram was the “man of the match” in
the final match of world cup 1992.
Since Sourav Ganguly was primarily a batter, his performance
ICs are steeper than that of Imran Khan who was primarily a
bowler (the steepness of ICs can be compared between Figures 8, Data availability statement
9). Consistent with the discussion of Section 2 and the empirical
analyses that follow, we see that Imran Khan’s performance peaked The raw data supporting the conclusions of this article
at a younger age than Sourav Ganguly’s. What is interesting to note will be made available by the corresponding author, without
here is a key difference—Sourav Ganguly made a comeback before undue reservation.
his performance peaked in ODI cricket, but Imran Khan’s made
a comeback after his performance peaked in ODI cricket. Both of
these case studies are interesting in their own right.
The study shows that both batters and bowlers playing ODI
cricket attain a peak in their careers. This peak is where the trade- Author contributions
off between fitness and experience balance each other out. Bowlers
have a different performance trajectory in comparison with batters VJ: Conceptualization, Data curation, Formal analysis,
and retire at an earlier age (Brown et al., 2021). The physical Investigation, Methodology, Resources, Software, Validation,
demands required for a bowler are much more as compared to a Writing – original draft, Writing – review & editing. SB:
batter. Thus, the deterioration of fitness also happens quicker for Conceptualization, Formal analysis, Methodology, Resources,
bowlers. Moreover in recent years, with more batter-favoring rules, Supervision, Validation, Visualization, Writing – original draft,
the bowler’s margin for error has been reduced even further. This Writing – review & editing. PK: Conceptualization, Investigation,
has necessitated the bowlers to upscale their skills (Douglas and Methodology, Project administration, Resources, Supervision,
Tam, 2010). Validation, Visualization, Writing – original draft, Writing –
This research offers insights for cricket team selection review & editing.
committees. We make a case for a healthy mix of young and senior
players with aligned interests. The Australian cricket team that has
won the maximum number of world cups has clearly employed
an optimal mix of both youth and experience. More specifically, Funding
there has always been a significant proportion of players who have
previously won a world cup and a healthy proportion of younger The author(s) declare that no financial support was
players who are both fit and talented. For every subsequent world received for the research, authorship, and/or publication of
cup, the former newcomers among the previous world cup winners this article.
become the experienced seniors who guide the new players. This
story continues with the Australian team to date making it the
most successful ODI team (with six world cups) among all the
cricket-playing nations.
Performance involves the cognizance of the relevant trade- Acknowledgments
offs involved (Careau and Wilson, 2017). In the labor market,
senior employees may lack physical and mental fitness for mundane The authors are grateful to Ayush Sudarsan for assisting Vipra
daily jobs (Salthouse, 2012) which could be costly to employers. Jain with the herculean task of compiling/cleaning the entire
However, they possess the appropriate experience required for dataset from various sources over a period of several months—this
advisory or managerial roles and can also facilitate knowledge aided the quality of empirics done by Vipra Jain. The remaining
transfer. Similarly, senior cricket players, though less fit, can authors were responsible for the research idea and the theory.
provide the right performance space to younger players. They can The authors thank Marie Claire Villeval, Matthew Rablen, Ronald
transfer game-specific knowledge to the younger players to boost Wendner, Johan Gustafsson, Arti Kalro, Souvik Banerjee, Sarthak
their confidence by sharing their own experiences. One of the best Gaurav, Shubh Verma, Soumita Ganguly, and Pooja Sanjay Kadam
Indian batters, Rahul Dravid, in an interview with Sambit Bal in for their valuable inputs. Sincere gratitude is extended to two
2004 summarizes the above thought in the following words13 : referees who provided valuable feedback to give this research
article its final shape. This article is fondly dedicated to Kaushik
Basu, not only for the volume of his experience from which
each author of this article has immensely benefited, but also for
13 See https://www.espncricinfo.com/story/the-dravid-method-134464 the youthful enthusiasm with which he contributed ideas to this
for reference. research agenda.

Frontiers in Behavioral Economics 12 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

Conflict of interest organizations, or those of the publisher, the editors and the
reviewers. Any product that may be evaluated in this article, or
The authors declare that the research was conducted in the claim that may be made by its manufacturer, is not guaranteed or
absence of any commercial or financial relationships that could be endorsed by the publisher.
construed as a potential conflict of interest.

Supplementary material
Publisher’s note
The Supplementary Material for this article can be found
All claims expressed in this article are solely those of the online at: https://www.frontiersin.org/articles/10.3389/frbhe.2024.
authors and do not necessarily represent those of their affiliated 1492735/full#supplementary-material

References
Adelman, M. L. (1986). A Sporting Time: New York City and the Rise of Modern Crowther, R. H., Renshaw, I., Spratford, W. A., Sayers, M. G., and Gorman,
Athletics Champaign, IL: University of Illinois Press, 1820–1870. A. D. (2023). Previous experiences can influence individual cricket spin bowlers’
perceptions of affordances in the performance landscape, Psychol. Res. 87, 1729–1742.
Agarwal, S., Driscoll, J. C., Gabaix, X., and Laibson, D. (2009). The age of reason:
doi: 10.1007/s00426-022-01769-6
financial decisions over the life cycle and implications for regulation. Brookings Pap.
Econ. Act. 2, 51–117. doi: 10.1353/eca.0.0067 Cureton, T. K. (1956). Relationship of physical fitness and athletic
performance and sports. JAMA. 162, 1139–1149. doi: 10.1001/jama.1956.02970290
Allen, S. V., and Hopkins, W. G. (2015). Age of peak competitive
035010
performance of elite athletes: a systematic review. Sports Med. 45, 1431–1441.
doi: 10.1007/s40279-015-0354-3 Doshi, A. (2016). Tendulkar in Wisden: An Anthology. London:
Bloomsbury Publishing.
Argote, L., Ingram, P., Levine, J. M., and Moreland, R. L. (2000). Knowledge transfer
in organizations: learning from the experience of others, Organiz. Behav. Human Deci. Douglas, M. J., and Tam, N. (2010). Analysis of team performances at
Proc. 82, 1–8. doi: 10.1006/obhd.2000.2883 the ICC World Twenty20 Cup. Int. J. Perform. Analys. Sport 10, 47–53.
doi: 10.1080/24748668.2010.11868500
Baker, J., and Wattie, N. (2018). Innate talent in sport: separating myth from reality.
Curr. Issues Sport Sci. 3, 1–9. doi: 10.36950/2018ciss006 Dumas, J. A. (2015). What is normal cognitive aging? Evidence from task-
based functional neuroimaging. Curr. Behav. Neurosci. Reports 2, 256–261.
Barnett, V., and Hilditch, S. (1993). The effect of an artificial pitch surface on
doi: 10.1007/s40473-015-0058-x
home team performance in football (soccer). J. Royal Statist. Soc. Series A, 156, 39–50.
doi: 10.2307/2982859 Eberhardt, W., de Bruin, W. B., and Strough, J. (2019). Age differences in financial
decision making: the benefits of more experience and less negative emotions. Behav.
Berry, S. M., Reese, C. S., and Larkey, P. D. (1999). Bridging different eras in sports.
Deci. Making 32, 79–93. doi: 10.1002/bdm.2097
J. Am. Statist. Assoc. 94, 661–676. doi: 10.1080/01621459.1999.10474163
Ericsson, K. A. (1990). Successful Aging: Perspectives From the Behavioral Sciences.
Berthelot, G., Johnson, S., Noirez, P., Antero, J., Marck, A., Desgorces, F. D.,
Cambridge: Cambridge University Press.
et al. (2019). The age-performance relationship in the general population and
strategies to delay age related decline in performance. Arch. Public Health 77, 1–9. Ericsson, K. A., Krampe, R. T., and Tesch-Römer, C. (1993). The role of deliberate
doi: 10.1186/s13690-019-0375-8 practice in the acquisition of expert performance. Psychol. Rev. 100, 363–406.
doi: 10.1037//0033-295X.100.3.363
Berthelot, G., Len, S., Hellard, P., Tafflet, M., Guillaume, M., Vollmer, J. C.,
et al. (2012). Exponential growth combined with exponential decline explains lifetime Funasaki, K., Inoue, Y., Takahashi, Y., Shirai, K., and Okada, Y. (2024). Knowledge
performance evolution in individual and human species. GeroSci. Offici. J. Am. Aging sharing among coaches: expert power and social cognitive theory perspectives. Eur.
Assoc. 34, 1001–1009. doi: 10.1007/s11357-011-9274-9 Sport Managem. Quart. 2014, 1–20. doi: 10.1080/16184742.2023.2299253
Berthelot, G., Sedeaud, A., Marck, A., Antero, J., Schipman, J., Sauliere, G., Ganguly, S. (2018). A Century is Not Enough: My Roller Coaster Ride to Success.
et al. (2015). Has athletic performance reached its peak? Sports Med. 45, 1263–1271. London: Juggernaut Books.
doi: 10.1007/s40279-015-0347-2
Hach, B. R., Gaina, L. I., Stanescu, D., and Gusita, B. (2023). “The impact
Borsch-Supan, A., Hunkler, C., and Weiss, M. (2021). Big data at work: of age on NBA player’s performances: a data mining approach,” in Institute of
age and labor productivity in the service sector. J. Econ. Ageing 19:100319. Electrical and Electronics Engineers 17th International Symposium on Applied
doi: 10.1016/j.jeoa.2021.100319 Computational Intelligence and Informatics (SACI) (Timisoara: Institute of
Borsch-Supan, A., and Weiss, M. (2016). Productivity and age: evidence from work Electrical and Electronics Engineers), 65–70. doi: 10.1109/SACI58269.2023.10158
teams at the assembly line. J. Econ. Ageing 7, 30–42. doi: 10.1016/j.jeoa.2015.12.001 605

Bradbury, J. C. (2009). Peak athletic performance and ageing: evidence from Hertzog, C. (2020). Aging and peak human performance: a glance back. J. Gerontol.
baseball, Journal of Sports Sci. 27, 599–610. doi: 10.1080/02640410802691348 Ser. B: Psychol. Sci. 75, 1621–1624. doi: 10.1093/geronb/gbaa057
Brown, T. W., Gough, L. A., and Kelly, A. L. (2021). Performance trajectories of Jamil, M., Harkness, A., Mehta, S., Phatak, A., Memmert, D., and Beato,
bowlers and batters from youth level to senior professional status in cricket. Int. J. Perf. M. (2023). Investigating the impact age has on within-over and death bowling
Analy. Sport 22, 1–15. doi: 10.1080/24748668.2021.1993028 performance in international level 50-over cricket. Res. Sports Med. 31, 171–180.
doi: 10.1080/15438627.2021.1954515
Careau, V., and Wilson, R. S. (2017). Performance trade-offs and ageing in the
‘world’s greatest athletes’. Proc. Royal Soc. B 284:20171048. doi: 10.1098/rspb.2017.1048 Kalen, A., Ferreiros, A. P., Costa, P. B., and Rey, E. (2021). Effects of age on physical
and technical performance in National Basketball Association (NBA) players. Res.
Chomik, R., and Jacinto, M. (2021). “Peak performance age in sport,” in Sports Med. 29, 277–288. doi: 10.1080/15438627.2020.1809411
The Australian Research Council Centre of Excellence in Population Ageing
Research (CEPAR) (Sydney, NSW: The Australian Research Council Centre of Kalwij, A., and Jaegher, K. D. (2023). The age-performance relationship for a
Excellence in Population Ageing Research (CEPAR)). Available at: https://cepar.edu. cognitive-intensive task: Empirical evidence from chess grandmasters. Sports Econ.
au/sites/default/files/peak-performance-age-sport.pdf Rev. 2:100010. doi: 10.1016/j.serev.2023.100010
Chopra, A. (2012). “A cricketer most evolved,” in Rahul Dravid: Timeless Steel, ed. Kanfer, R., and Ackerman, P. L. (2004). Aging, adult development, and work
S. Bal (1st ed) (Mumbai: The Walt Disney Company (India) Private Limited), 174. motivation, Acad. Managem. Rev. 29, 440–458. doi: 10.5465/amr.2004.13670969
Connor, J. D., Renshaw, I., and Farrow, D. (2020). Defining cricket batting Kholkine, L., Latre, S., Verdonck, T., and de Leeuw, A. W. (2023). Age
expertise from the perspective of elite coaches. PLoS ONE 15:e0234802. of peak performance in professional road cycling. J. Sports Sci. 41, 298–306.
doi: 10.1371/journal.pone.0234802 doi: 10.1080/02640414.2023.2208998

Frontiers in Behavioral Economics 13 frontiersin.org

Jain et al. 10.3389/frbhe.2024.1492735

Kim, H., and Lee, B. S. (2023). Aging workforce, wages, and productivity: Raman, W. V. (2009). There’s no substitute for skill. Sportstar 31:24. Available
do older workers drag productivity down in Korea? J. Econ. Ageing 24:100444. at: https://sportstar.thehindu.com/magazine/theres-no-substitute-for-skill/
doi: 10.1016/j.jeoa.2023.100444 article29706006.ece4
Kirsch, G. B. (1991). Massachusetts Baseball and Cricket, Sports in Massachusetts: Rey, E., Lorenzo-Martinez, M., Lopez-Del, C. R., Resta, R., and Lago-
Historical Essays. Westfield, MA, Institute for Massachusetts Studies, 1840–1870. Penas, C. (2022). No sport for old players. A longitudinal study of aging
effects on match performance in elite soccer. J. Sci. Med. Sport 25, 535–539.
Lehman, H. C. (1953). Age and Achievement, Princeton NJ: Princeton
doi: 10.1016/j.jsams.2022.03.004
University Press.
Saikia, H., Bhattacharjee, D., Bhattacharjee, A., and Lemmer, H. (2015).
Lewis, L. F., and Wescott, H. D. (2017). Multi-generational workforce: four
Longitudinal linear mixed effect model: an application in analyzing age effects in
generations united in lean. J. Busin. Stud. Quart. 8, 1–14. Available at: https://www.
Twenty20 cricket. Thailand Statist. 13, 223–42. Available at: https://ph02.tci-thaijo.org/
proquest.com/docview/18803658772
index.php/thaistat/article/download/37774/31385/
Lucifora, C., and Simmons, R. (2003). Superstar effects in sport: evidence from
Saikia, H., Bhattacharjee, D., and Mukherjee, D. (2019). Cricket Performance
Italian Soccer, J. Sports Econ. 4, 35–55. doi: 10.1177/1527002502239657
Management. Cham: Springer Nature.
Mangan, J. (2014). “Ageing and Test cricket averages: a statistical note,” in
Salthouse, T. (2012). Consequences of age-related cognitive declines. Annual Rev.
Australian Institute for Business and Economics Working Paper Series (Brisbane, QLD:
Psychol. 63, 201–226. doi: 10.1146/annurev-psych-120710-100328
Australian Institute for Business and Economics).
Santhanam, V., and Balasubramanian, S. (2009). If Cricket is Religion, Sachin is God.
Marck, A., Antero, J., Berthelot, G., Johnson, S., Sedeaud, A., Leroy, A., et al. (2019).
Gurgaon: HarperCollins Publishers India.
Age-related upper limits in physical performances. J. Gerontol.: Series A, 74, 591–599.
doi: 10.1093/gerona/gly165 Schulz, R., and Curnow, C. (1988). Peak performance and age among superathletes:
track and field, swimming, baseball, tennis, and golf. J. Gerontol.: Psychol. Sci. 43,
Menon, S. (2012). “The talent myth,” in S. Bal (Ed.), Rahul Dravid: Timeless Steel (1st
113–120. doi: 10.1093/geronj/43.5.P113
ed), (Mumbai: The Walt Disney Company (India) Private Limited), 174.
Schulz, R., Musa, D., Staszewski, J., and Siegler, R. S. (1994). The relationship
Mirza, S. (2023). “Beyond cricket,” in Sachin @ 50: Celebrating a Maestro, ed. B.
between age and major league baseball performance: implications for development.
Majumdar (1st ed) (Noida: Simon and Schuster India), 124.
Psychol. Aging 9, 274–86. doi: 10.1037//0882-7974.9.2.274
Moore, J. C., and Brylinsky, J. (1995). Facility familiarity and the home advantage.
Silva, A. F., Conte, D., and Clemente, F. M. (2020). Decision-making in youth
J. Sport Behav. 18, 302–311.
team-sports players: a systematic review. Int. J. Environm. Res. Public Health 17:3803.
Morley, B., and Thomas, D. (2005). An investigation of home advantage and doi: 10.3390/ijerph17113803
other factors affecting outcomes in English one-day cricket matches. J. Sports Sci. 23,
Silva, R. M., Manage, A. B. W., and Swartz, T. (2015). A study of the
261–268. doi: 10.1080/02640410410001730133
powerplay in one-day cricket. Eur. J. Operat. Res. 244:3. doi: 10.1016/j.ejor.2015.
Noakes, T. D., and Durandt, J. J. (2000). Physiological requirements of cricket. J. 02.004
Sports Sci. 18, 919–929. doi: 10.1080/026404100446739
Thorley, J. (2021). Bowled over or over bowled? Age-related changes in the
Noorbhai, M. H., and Noakes, T. D. (2015). Advances in cricket in the 21st century: performance of bowlers in Test match cricket. Int. J. Sports Sci. Coach. 16, 1138–1151.
science, performance and technology. Afric. J. Phys. Health Educ. Recreat. Dance 21, doi: 10.1177/17479541211001726
1310–1320. Available at: https://hdl.handle.net/10520/EJC1822983
Ugra, S. (2012). “When you’ve played at the top, it’s hard to settle for second-best,”
Pollard, R. (2002). Evidence of a reduced home advantage when a team moves to a in Rahul Dravid: Timeless Steel, ed. S. Bal (1st ed) (Mumbai: The Walt Disney Company
new stadium. J. Sports Sci. 20, 969–973. doi: 10.1080/026404102321011724 (India) Private Limited), 174.

Frontiers in Behavioral Economics 14 frontiersin.org