How to Solve a Problem Like Australia? Part 1 - ODIs


25th August, 2016.

Australian cricket is down in the dumps.  Arguably the most consistent cricket nation of recent times, full of pomp and arrogance, certainly since I was growing up, they are now reduced to bizarre selections and bits and pieces cricketers, and following the debacle of a 3-0 Test series whitewash to Sri Lanka, the ODI series is level at 1-1 at the time of writing.

It's worth remembering that their opponents are a Sri Lanka side which has been decimated through injuries and retirements, and recently got a hiding in all formats against England several months ago, although of course, conditions in the subcontinent are far more to Sri Lanka's liking than those of the early English summer.

However, despite the venue advantage, it's fair to say that not many saw Australia's issues coming.  Having said this, when looking at one day players since 2014, the fact that Australia have used an incredible 38 different players in this time period is quite telling that the management have no real idea who their best team is.

I ran some numbers from the domestic and international 50 over matches that these 39 players played in from 2014 to 2016 (up to and including the second ODI with Sri Lanka), with, as always, adjustments made to averages/strike rate/economy rates based on innings difficulty and the venue.  

The ODI players used by Australia are ranked below (with the exception of the late Phil Hughes), sorted by combined mean deviation of adjusted batting average and adjusted strike rate from 2014-2016 (for an explanation of the formula involved, please click here):-

Player

Adj Batting Ave

Adj Batting SR

Adj Batting Ave Mean Deviation

Adj Batting SR Mean Deviation

Combined Mean Deviation







DA Warner

49.31

102.26

1.50

1.12

1.31

SPD Smith

53.48

86.03

1.63

0.94

1.28

SE Marsh

52.00

78.45

1.59

0.86

1.22

UT Khawaja

48.46

83.61

1.48

0.91

1.20

GJ Maxwell

30.60

117.08

0.93

1.28

1.11

MR Marsh

38.66

92.65

1.18

1.01

1.10

CL White

42.35

71.82

1.29

0.78

1.04

MJ Clarke

36.36

85.84

1.11

0.94

1.02

BJ Haddin

31.01

96.47

0.95

1.05

1.00

JP Faulkner

30.86

95.07

0.94

1.04

0.99

AJ Finch

34.87

83.46

1.06

0.91

0.99

BCJ Cutting

25.46

103.13

0.78

1.13

0.95

SR Watson

32.05

84.30

0.98

0.92

0.95

NM Lyon

18.29

120.19

0.56

1.31

0.93

MC Henriques

35.54

69.82

1.08

0.76

0.92

MS Wade

29.70

84.62

0.91

0.92

0.91

DT Christian

24.93

93.74

0.76

1.02

0.89

TM Head

25.53

90.33

0.78

0.99

0.88

MG Johnson

13.72

116.13

0.42

1.27

0.84

GJ Bailey

27.98

75.58

0.85

0.82

0.84

JA Burns

27.43

72.75

0.84

0.79

0.82

JW Hastings

19.45

86.22

0.59

0.94

0.77

AC Agar

21.42

79.82

0.65

0.87

0.76

MP Stoinis

24.35

63.78

0.74

0.70

0.72

NM Coulter-Nile

10.43

99.67

0.32

1.09

0.70

A Zampa

14.90

86.89

0.45

0.95

0.70

SA Abbott

14.61

77.63

0.45

0.85

0.65

CJ McKay

13.53

76.29

0.41

0.83

0.62

PJ Cummins

18.22

60.15

0.56

0.66

0.61

KW Richardson

11.35

70.47

0.35

0.77

0.56

GS Sandhu

16.40

48.11

0.50

0.52

0.51

JL Pattinson

13.57

55.69

0.41

0.61

0.51

XJ Doherty

6.83

69.53

0.21

0.76

0.48

MA Starc

9.31

61.15

0.28

0.67

0.48

JR Hazlewood

7.05

49.06

0.22

0.54

0.38

SM Boland

1.93

41.94

0.06

0.46

0.26

JS Paris

N/A

N/A

N/A

N/A

N/A


There's quite a lot of depth in the Australian batting, with eight players (including the now retired Michael Clarke) being above the standard of the average One Day International batsman.  

Certainly, this gives Australia plenty of batting options, although the fact that of the top eleven batsman, only David Warner and Glenn Maxwell have adjusted strike rates over 100 means that scoring big totals is going to be difficult for them.  Having said that, there's definitely little reason to think that, if teams are selected correctly, that Australia should collapse.

David Warner's data indicated he is currently the best Australian ODI batsman...

However, the numbers suggest that specialist batsmen Joe Burns, George Bailey and Travis Head are not of the standard required to play one day international cricket, although I will add that Head's sample size wasn't the biggest.  Quite how Bailey has been picked so many times in the last three years is beyond me, and at almost 34 years of age he is unlikely to turn his career around.

The next question should be - was there anyone else in domestic cricket not called up who could do a good job?  

With an expected average of 41.87 and expected strike rate 66.94, Michael Klinger had a combined mean devaition of 1.00 (high average, low strike rate) and could argue he has been unlucky not to get a chance, but his numbers aren't stellar enough to say that he's been very unfairly treated.  

Wicket-keeper batsman Tim Paine (adjusted average 38.87, adjusted strike rate 67.46) had a combined mean deviation of 0.96 (again, high average, low strike rate) and overall is a better batsman than competitor Matthew Wade, although Wade's strike rate is better.  The retired Brad Haddin had the best wicket-keeper batsman numbers.

Apart from those two possibles, the numbers didn't indicate any overwhelmingly better batting solutions than the ones at the top of the table above, so it would appear that those players picked in the last three years do represent the best batsmen that Australia have to choose from, in this format.

The table below shows the ODI players used by Australia, ranked below and sorted by combined mean deviation of adjusted bowling average and adjusted economy rate from 2014-2016:-

Player

Adj Bowling Ave

Adj Bowling Econ

Adj Bowling Ave Mean Deviation

Adj Bowling Econ Mean Deviation

Combined Mean Deviation







MA Starc

17.59

4.55

1.86

1.21

1.54

NM Coulter-Nile

21.69

4.84

1.51

1.14

1.32

MG Johnson

23.21

4.83

1.41

1.14

1.28

JR Hazlewood

26.15

4.80

1.25

1.15

1.20

SA Abbott

25.30

5.44

1.30

1.01

1.15

GS Sandhu

26.93

5.05

1.22

1.09

1.15

PJ Cummins

24.67

5.71

1.33

0.96

1.15

JS Paris

30.43

4.90

1.08

1.12

1.10

JP Faulkner

28.89

5.33

1.13

1.03

1.08

GJ Maxwell

31.52

5.34

1.04

1.03

1.03

MR Marsh

31.48

5.38

1.04

1.02

1.03

A Zampa

33.57

5.22

0.98

1.05

1.02

MC Henriques

37.80

4.75

0.87

1.16

1.01

JL Pattinson

35.01

5.11

0.94

1.08

1.01

JW Hastings

35.66

5.19

0.92

1.06

0.99

KW Richardson

34.68

5.42

0.95

1.01

0.98

NM Lyon

36.58

5.17

0.90

1.06

0.98

BCJ Cutting

34.64

5.56

0.95

0.99

0.97

XJ Doherty

51.98

4.63

0.63

1.19

0.91

DT Christian

41.83

5.54

0.78

0.99

0.89

SPD Smith

35.87

6.41

0.91

0.86

0.89

CJ McKay

46.86

5.34

0.70

1.03

0.86

SM Boland

46.04

5.68

0.71

0.97

0.84

AC Agar

55.41

5.52

0.59

1.00

0.79

AJ Finch

59.64

5.52

0.55

1.00

0.77

MP Stoinis

59.71

5.65

0.55

0.97

0.76

SR Watson

80.00

6.15

0.41

0.89

0.65

TM Head

96.20

6.35

0.34

0.87

0.60


We can see here that with a mean deviation of 1.54, Mitchell Starc is a genuine world class bowler who is the obvious choice to lead the Australian attack.  Any Australian team without Starc in it, has a much lower chance of winning an individual match.

Mitchell Starc is a world-class bowler...

What was surprising is that Nathan Coulter-Nile, who had impressive international and domestic numbers, was second, and given his high batting strike rate down the order, I am surprised he isn't a regular for Australia in this format.  

Mitchell Johnson and Josh Hazlewood are bowlers of quality, whilst other pace options Sean Abbott, Gurinder Sandhu, Pat Cummings and Joel Paris also showed up well in the data, as did the all-rounder James Faulkner.

With pacemen all at the top of the data, we have to drop down to 10th - Glenn Maxwell - to find the best spin option.  Currently, Maxwell isn't playing in the Sri Lanka series, having been dropped, and considering he's 5th in the batting rankings as well, this decision seems ludicrous.  Having an above expectation spinner who can bat with an average around 30 at over a run a ball is an incredible asset to have in a team, not someone who should be jettisoned.

Adam Zampa was the only other spinner who had a combined mean deviation over 1, indicating he is also better than the average one day international bowler.  Zampa's batting numbers - as a bowler - are far from a disaster either (expected average of just under 15, expected strike rate just shy of 90, which is well above expectation for a number 10 in this format).

Having concentrated on the best batsmen and bowlers, there are also a number of players used by Australia who didn't add value in their specialist field.  These were the following:-

Batsmen:-

George Bailey
Joe Burns
Travis Head

Bowlers:-

Scott Boland
Xavier Doherty
Nathan Lyon
Clint McKay
Kane Richardson

Lyon and Richardson had just below average numbers, but given that - in Richardson's case in particular - the pace depth is so strong for Australia, it's tough to make a case for his selection.   Lyon is not really a threat (adjusted bowling average 36.58) but his economy expectation was decent at 5.17.  However, the Australian way is to attack, not contain, and his selection seems strange given Maxwell has better bowling numbers and is obviously a much better batsman.

Finally, we can look at all-rounders, with Australia having chosen to use a number of these players in the last three years - I refer back to the top of the article which talks about the number of bits and pieces cricketers they have used.

First of all I want to make reference to the likes of Aaron Finch and Travis Head.  These two players are not even bits and pieces cricketers, or all-rounders.  They are front-line batsmen.  Their bowling expectation is horrific from a wicket expectation perspective (and Head's economy is also atrocious) and it is an embarrassment to one of the best cricketing nations of all time that they are expected to bowl.

The table below illustrates the added value that the all-rounders used by Australia from 2014-2016 are expected to bring at this current point:-

Player

Adj Batting Ave Mean Deviation

Adj Batting SR Mean Deviation

Combined Mean Batting Deviation

Adj Bowling Ave Mean Deviation

Adj Bowling Econ Mean Deviation

Combined Mean Bowling Deviation

Combined Batting/Bowling Mean Deviation









GJ Maxwell

0.93

1.28

1.11

1.04

1.03

1.03

0.14

MR Marsh

1.18

1.01

1.10

1.04

1.02

1.03

0.13

JP Faulkner

0.94

1.04

0.99

1.13

1.03

1.08

0.07

MC Henriques

1.08

0.76

0.92

0.87

1.16

1.01

-0.07

BCJ Cutting

0.78

1.13

0.95

0.95

0.99

0.97

-0.08

DT Christian

0.76

1.02

0.89

0.78

0.99

0.89

-0.22

SR Watson

0.98

0.92

0.95

0.41

0.89

0.65

-0.40

AC Agar

0.65

0.87

0.76

0.59

1.00

0.79

-0.45

MP Stoinis

0.74

0.70

0.72

0.55

0.97

0.76

-0.52


Glenn Maxwell, Mitch Marsh and James Faulkner are overwhelmingly the best all-rounders, with Maxwell and Marsh the best batting all-rounders and Faulkner the best bowling all-rounder.  

However, only Moises Henriques and Ben Cutting, of the rest, didn't have disastrous numbers, although they neither added much from an overall batting, nor bowling perspective, although Cutting is a dangerous hitter down the order.  

Data indicates Henriques certainly isn't a good enough international all-rounder, but he isn't quite as bad as people on social media make out, although for someone that usually comes bats during the second half of an innings, his strike rate is beyond atrocious.

Data does not support the inclusion of Moises Henriques in the Australian ODI set-up...

Dan Christian, Ashton Agar, Shane Watson and Marcus Stoinis had very poor expectation numbers, with particular emphasis on the latter three.  It's difficult to understand why they would even be considered for selection in this format.

These generally poor all-rounder numbers give some insight as to why Australia have struggled - far too many bits and pieces cricketers who are neither good enough to be in the team as a batsman or bowler.  Therefore, when they bat, they do not add value to the team, and when they bowl, the produce generally below average displays.  There really is no need to have 7 or 8 decent bowling options in a side, as it will detract from the batting, and usually the extra options are worse bowlers than average. 

Based on the data, the following players should form the basis of the Australian ODI squads in the near future:-

Batsmen:-

David Warner
Steve Smith
Shaun Marsh
Usman Khawaja
Glenn Maxwell
Cameron White
Aaron Finch

All-rounders:-

Mitchell Marsh
Glenn Maxwell
James Faulkner

Wicket-keepers:-

Matthew Wade
Tim Paine

Bowlers:-

Mitchell Starc
Nathan Coulter-Nile
Mitchell Johnson
Josh Hazelwood
Sean Abbott
Gurinder Sandhu
Pat Cummins
Joel Paris
Adam Zampa

The best Australian ODI team to choose for an average match, based on data, would be the following:-

Warner, Khawaja, S Marsh, Smith, M Marsh, Maxwell, Faulkner, Wade, Johnson, Coulter-Nile, Starc.  

In spin-friendly conditions, Adam Zampa would be the obvious replacement for one of the pace bowlers.

This team is of great contrast to the current ODI team that was defeated by Sri Lanka in the second ODI - Aaron Finch, George Bailey, Moises Henriques, Travis Head and Nathan Lyon were all picked by the selectors, along with Adam Zampa, which in Zampa's case is fair given the spin-friendly conditions in the subcontinent.

The team picked on data, however, would have batting depth, and in Maxwell through to Coulter-Nile, batting depth with high strike rates to take accelerate in the second half of the innings after the top four/five batsmen give a solid start.  There are also six bowling options in Marsh, Maxwell, and Faulkner as all-rounders, and Johnson, Coulter-Nile and Starc as pace threats.

If you enjoyed reading this article, please feel free to make a donation towards the upkeep of the website.  
This will also help me to prioritise the time to write further articles and previews.  




Comments