The Roar
The Roar



The stats that prove we should pick our best XI regardless of age

DanTheStatsMan new author
Roar Rookie
23rd January, 2022
Autoplay in... 6 (Cancel)
Up Next No more videos! Playlist is empty -
DanTheStatsMan new author
Roar Rookie
23rd January, 2022
1095 Reads

As a sport, cricket lends itself particularly well to statistical analysis.

Such analysis is often riddled with oversights – insufficient sample size, spurious correlations, and inappropriate statistical tools for a given question.

When done properly, however, it can give powerful insights into commonly debated questions.

One that tickled my curiosity is the importance of investing in promising young batters by selecting them in the Test team.

This has been common practice over recent decades – think back to 2016 when Matt Renshaw, Peter Handscomb and Nic Maddinson, all in their early-to-mid 20s, were brought into the Test team.

This was arguably more as an investment in the future than to reward their performances to date.


But is this a worthwhile practice?

I decided on the following process in an attempt to answer this question. I identified every Australian Test batter who debuted within the last 30 years and has now (beyond reasonable doubt) completed their Test career.

I then divided these players into two groups – those who played at least ten Test innings before their 28th birthday (who I will refer to as early starters) and those that debuted after their 28th birthday (the late starters).

Ultimately, there were 12 batters in the early starters group and 19 in the late starters group.

I then looked at their later career Test run-scoring, defined as the period after their 28th birthday.

Michael Hussey of Australia celebrates scoring the winning runs

Michael Hussey (Photo by Tom Shaw/Getty Images)

So, how do the numbers look?

First, I took the total Test runs for each group over their collective later career (the period after each player’s 28th birthday).


This totalled 49,261 runs at 46.34 for early starters and 29,364 runs at 40.56 for late starters – 5.78 runs per innings more for those that had early exposure to the Test team.

The main confounder here, however, is that better players will generally debut earlier, meaning the early starter group will be of a higher quality.

While there is no perfect method to correct for this, I would suggest first-class batting average as the best measure of the quality of a player over their professional career.

This is 47.92 for the early starters compared to 43.97 for the late starters a difference of 3.95.

Therefore, if we correct later career Test performance relative to first-class average, the benefit of early Test exposure reduces to about two runs per innings.

There’s another way to run the numbers, however. The above analysis was based on aggregate runs, thereby weighting heavily towards a handful of particularly prolific players.

To gauge how a particular individual will fare, it is better to weight the performances of each player equally by simply taking the mean of their Test averages.

Steve Smith is bowled by James Anderson.

(Photo by Robert Cianflone/Getty Images)


By doing this, the mean Test batting average in the later career of early starters is 42.26 compared to 41.87 for late starters a difference of just 0.39 runs per innings.

Additionally, if you were to standardise this to the mean first-class batting averages of each group (as I did before), early exposure actually confers a loss of about two runs per innings.

So, what to make of this?

Statistically, the benefit of giving batters early exposure to the Test team seems limited – somewhere between a loss of two runs to a gain of five runs per innings over their later career, depending on how the analysis is conducted.

I suspect the true figure lies somewhere in the middle – a benefit of a couple of runs or so.

Additionally, gaining this benefit is predicated on early exposure to the Test team, which historically has not been a productive period.

In this analysis, the early starters demonstrated a mean average of 35.05 by their 28th birthday. Many, including the likes of Matthew Hayden, Damien Martyn and Justin Langer, averaged below 30.

Justin Langer and Matt Hayden

(Photo by Hamish Blair/Getty Images)


Therefore, whatever small benefit is ultimately obtained from early exposure needs to be offset against the (often relatively limited) immediate fruits of that period.

The direct application of this analysis is whether to give young players an opportunity in the Test team as a form of investment in their future output.

A current example would be exciting young openers Bryce Street and Henry Hunt, who have produced respectable if not spectacular output at domestic level.

While I share in this excitement, the statistics suggest giving them Test exposure for the sake of experience would only make a fairly minor difference in their later career.

While not strictly within the scope of this analysis, I wonder whether this concept could be applied to players of all ages and levels of experience.

In other words, whether Test exposure at any age makes a significant difference to future output.

Consider the current situation with Marcus Harris and Usman Khawaja.

Usman Khawaja on the way to his ninth century

(Photo by Cameron Spencer/Getty Images)


Most pundits would agree Khawaja is currently the superior red-ball player, however Harris was initially selected due to his younger age and longer future in the game.

This analysis, however, lends weight to selecting Khawaja given he is the superior current player, with little opportunity cost in not selecting Harris in terms of his future run-scoring.

I appreciate readers may be incredulous at this point. Those that watch, play and/or appreciate the game would assert that Test cricket is an entirely different beast to first-class and other domestic cricket, giving players an entirely different environment to advance their game.

I am sympathetic to this view – it is simply that the numbers don’t bear this out.

There is one concession, however: a population-level analysis such as this cannot necessarily be applied to individuals.

While early Test exposure gives little benefit on average, there will be variability between individual players.

Some may have benefited enormously, others less so, and for some it may even have been detrimental (such as due to a loss of confidence or increased pressure on returning to the team).

Marnus Labuschagne

(Photo by Bradley Kanaris/Getty Images)

Perhaps identifying which players will benefit from early exposure is the key.

Having a national selector like George Bailey, who is younger and knows the current players more personally, may be beneficial in this respect.

I suspect he sees Marcus Harris as an individual for whom current Test exposure is necessary to increase his future Test output, particularly given his longstanding dominance of domestic cricket and gradual improvement over the most recent Ashes series.

Sports opinion delivered daily 


Generally, however, it would appear the best approach is to select the best players regardless of age.

While antithetical to the future-orientated views of Greg Chappell and others, the careers of players including Michael Hussey, Adam Gilchrist, Brad Hodge and Adam Voges demonstrate immediate success at Test level is possible for those at a later stage of their career.

In a way, this is unsurprising – these players had had years to fine-tune their technique through domestic cricket and, in the twilight years of their career, had a more mature temperament and worldly perspective.

So long as selecting these players doesn’t come at the cost of developing younger talent, which this analysis suggests is not the case, we can simply pick the best XI at the time, even if it leaves us with something of a Dad’s Army.