Alex Hales, Gary Ballance & James Vince - Good Enough for England?


Despite England's recent success in Test Cricket, there has been plenty of speculation that several of their batsmen are far from secure in their positions.

Certainly England look to have plenty of depth in the pace bowling department, which will usually be good enough to win the majority of matches on home soil in conditions which suit them.  However, to most observers they don't have this depth in their batting.  Captain Alastair Cook and the world-class Joe Root are the mainstays of the batting line-up, but Nick Compton was dropped after a series of mediocre performances whilst Alex Hales, James Vince and Gary Ballance are all under pressure to deliver after inconsistency and poor scores have beset the starts of their Test careers.  

The question that is often asked is, are they good enough?  Will they become good enough?  A further alternative point of view, from some corners, is the following question - are there better players in county cricket more suitable for the England team?

Actual and Expectation Data makes answering those questions much easier than any subjective judgement ever will.  The problem with subjective judgement (e.g. visually scouting players) is that it is difficult to avoid a pre-conceived idea or bias, whilst data, when analysed correctly, is absolutely impartial.

So let's look at the data to solve the questions above.  Before playing 5-day Test Cricket, it is typical for a player to succeed in 4-day County Cricket first, so it is possible to see whether these players proved that they could be good enough for Test Cricket in the 4-day format.

The table below shows the County Championship average of Hales, Vince and Ballance from the start of the 2014 season until 28/7/16:-

Player

Division

Completed Innings

Runs

Ave

BF

SR








AD Hales

1

40

1989

49.73

3006

66.17

AD Hales

2

2

78

39.00

136

57.35

AD Hales

Overall

42

2067

49.21

3142

65.79

GS Ballance

1

36

1514

42.06

2936

51.57

GS Ballance

Overall

36

1514

42.06

2936

51.57

JM Vince

1

33

1018

30.85

1822

55.87

JM Vince

2

25

1525

61.00

1984

76.86

JM Vince

Overall

58

2543

43.84

3806

66.82


As can be seen from the data above, none of the three players has averaged over 50 in 4-day County cricket since the start of 2014.  This is a critical number as England's players, in 2014 and 2015 combined, had a Test average of 0.8x their domestic average in the same time period.  Therefore a player that has averaged 50 in the English County Championship has an expectation to average 40 in Test Cricket (50*0.8).

Using this basic measure of three year averages in the County Championship, we can say that none of this trio have an expectation to average 40 in Test Cricket - the mark of a solid Test batsman.

For most people, a quick glance at the table's overall averages and strike rates would lead people to assuming that Hales is better than Vince, who is better than Ballance.

Alex Hales has the best County Championship average of the three batsmen under pressure...

However, Hales (albeit just two) and Vince have played innings in Division Two of the County Championship and this also must be accounted for.  

We can use adjustment factors to take into account the relative ease of run-scoring in Division Two, compared to Division One.  Certainly Vince, with a Division Two batting average almost double that he recorded in Division One, enjoyed the weaker bowling in the lower level of competition.  Ballance hasn't had that chance in recent years.

When I looked at the difference in domestic averages compared to Test averages between Division One and Division Two in 2014 and 2015 there actually wasn't a huge discrepancy between the two divisions.  However, with the England selectors preferring to pick Division One players, the sample for Division Two was small and somewhat dominated by Alastair Cook and Moeen Ali.  

Therefore a better measure was to look at the players from 1/1/14 to 28/7/16 who had played in both divisions of the County Championship, of which there was a total of 125 (1967 completed innings).  This sample of completed innings was much more substantial.  

These 125 players recorded a Division One average of 0.8x (where x is known to me, but kept to myself due to sensitivity) multiplied by their Division Two average, and 0.9x (again, the same for x here) multiplied by their Division Two strike rate.  With this, we now have evidence that run-scoring is significantly easier in Division Two than Division One - entirely logical as it is a lower level - and this must be factored into the average calculations for Hales, Ballance and Vince.

When these adjustment factors are introduced, we can find expectations for Division One average and strike rate, and the same data for Test average and strike rate:-

Player

Adjusted Division One Expected Average

Adjusted Division One Expected Strike Rate

Test Expected Average

Test Expected Strike Rate






AD Hales

48.90

65.56

39.12

57.04

GS Ballance

42.06

51.57

33.65

44.87

JM Vince

39.38

63.21

31.50

54.99


The table above shows that due to the lower level of bowling faced in Division Two, Vince's Division One expected average drops below Ballance's, and the Test expected averages are also all below 40, with Ballance's and Vince's not much above 30.  Despite Ballance's current career Test average in the 40s, this data makes it very difficult to classify Ballance or Vince as likely to succeed at Test level in the long term.  It is more than possible that Ballance's career average is flattered by playing 22 of his 30 Test innings at home, with the majority of only eight away innings (six) coming against the West Indies.  Certainly Vince's mediocre data in Division One domestic cricket makes it baffling as to why he was even selected for the Test team in the first place.

At this point, it's also worth talking about the trading implications of this data.  There are likely to be times in the near future (I don't think we'll be this lucky in the mid/long term) that Ballance and Vince will be batting together at Test level.  Given their statistical expectations are likely to be lower than market expectations, it is reasonable to conclude that picking spots to back England's opponents in-play with this duo at the crease will be an entry point with positive expectation.

Gary Ballance's Test expected average of 33.65 makes it very difficult for him to be a long-term batsman for England... 

It's also difficult to specify a reason as to why these players have been selected.  A push by the media, being the 'next cab off the rank', having a face that fits, ambitious subjective assessment by a selector(s) or a lack of better alternatives could all be realistic explanations.  However, the last reason - a lack of better alternatives - could be addressed by data too.  

The table below shows the expected Division One averages and strike rates (or real Division One averages if the player has only played Division One cricket in the sample period) for English qualified players using data from 1/1/14 to 28/7/16, as well as these players' Test expectation averages and strike rates. Only players with over 2000 runs in the time period were included in the sample:-

Player

Adjusted Division One Expected Average

Adjusted Division One Expected Strike Rate

Adjusted Test Expected Average

Adjusted Test Expected Strike Rate






JM Bairstow

71.50

73.93

57.20

64.32

EC Joyce **

50.52

54.90

40.42

47.76

AD Hales

48.90

65.56

39.12

57.04

A Lyth

48.74

53.07

38.99

46.17

ME Trescothick

48.07

55.86

38.46

48.60

JC Hildreth

46.92

61.36

37.54

53.38

LJ Wright

46.91

64.89

37.53

56.45

DJ Malan

46.45

53.95

37.16

46.94

CMW Read

45.89

58.40

36.71

50.81

SG Borthwick

45.59

57.23

36.47

49.79

SA Northeast

43.42

53.84

34.74

46.84

MH Wessels ***

39.76

59.05

31.81

51.37

JM Vince

39.38

63.21

31.50

54.99

SD Robson

39.38

54.93

31.50

47.79

WL Madsen

39.10

46.63

31.28

40.57

NRD Compton

39.00

46.09

31.20

40.10

CD Nash

38.87

55.85

31.10

48.59

PD Trego

38.68

69.80

30.94

60.73

SJ Croft

38.32

47.32

30.66

41.17

NLJ Browne

38.09

46.22

30.47

40.21

AZ Lees

37.71

48.71

30.17

42.38

JJ Roy

37.34

79.97

29.87

69.57

SM Davies

37.24

57.25

29.79

49.81

PD Collingwood

37.00

55.98

29.60

48.70

MD Stoneman

36.56

65.80

29.25

57.25

BM Duckett *

36.39

62.31

29.11

54.21

DKH Mitchell

36.21

41.89

28.97

36.44

KK Jennings *

35.88

46.38

28.70

40.35

RJ Burns

35.83

47.56

28.66

41.38

MA Carberry

35.60

46.54

28.48

40.49

V Chopra

34.20

49.78

27.36

43.31

BA Godleman

34.03

46.28

27.22

40.26

WR Smith

33.93

38.92

27.14

33.86

LWP Wells

33.90

42.72

27.12

37.17

SR Patel

32.50

61.97

26.00

53.91

T Westley

32.17

49.78

25.74

43.31

MJ Richardson

32.04

52.58

25.63

45.74

PJ Horton

31.97

46.12

25.58

40.12

JS Foster

31.73

51.11

25.38

44.47

CDJ Dent

31.50

47.25

25.20

41.11

JHK Adams

31.30

42.34

25.04

36.84

DJ Bell-Drummond *

31.20

43.79

24.96

38.10

WD Bragg

27.59

40.90

22.07

35.58

AJ Robson

27.52

46.97

22.02

40.86

* These players had significantly improved statistics in 2016
** Joyce has played for both England and Ireland 
*** Wessels should be able to play for England from 2016

Looking at the data above, just two players had an expected Test average of over 40 - Jonny Bairstow, who is currently in the test team, and the veteran Ed Joyce, who played white-ball cricket for both England and Ireland in the past.

Hales is third in the list and with better alternatives thin on the ground, the data suggests that he should be persevered with.  However, Ballance, who was not included in the table due to not scoring 2000 domestic runs in the time period, would be ranked just 11th in this list for expected Test average (33.65) whilst Vince was equal 13th with the discarded opener Sam Robson.  

Other players above Ballance and Vince who are unlikely to be available for selection include another veteran, Marcus Trescothick (5th), who has retired from Test cricket, fellow Somerset batsman James Hildreth (6th), who was extremely highly rated when younger and Luke Wright (7th), who again was positively treated by data, as he was in T20 analysis, but persistently ignored by selectors.  It would be absolutely justifiable for Wright to complain that he didn't get enough chances at international level.

Adam Lyth is the batsman that the data highlighted that should be recalled to the Test team instead of Ballance or Vince, and given that it is unlikely that the selectors will pick others above him, Dawid Malan also looks to have claims.  Scott Borthwick, whose bowling ability could also prove useful, and Sam Northeast also merit consideration.  Northeast was also highlighted in the T20 articles as an under-rated batsman and data suggests that the Kent captain is a better batsman than some of those in the England set-up currently across both formats.

Northeast's team-mate Daniel Bell-Drummond, along with Ben Duckett and Keaton Jennings, boast much improved data in 2016 and this trio, with years on their side, could be viable future alternatives.  Certainly if they start 2017 strongly, their three-year rolling average and strike rate expectations would increase signficantly and they'd leap up this table.

Nick Compton (16th) was dropped following a poor start to the Sri Lanka series and the data shows that there were many other batsmen who should have been given a chance when Compton was recalled.  The 2014-2016 domestic data generated a Test expectation average of 31.20 with an extremely low expected strike rate of 40.10 for Compton, and this isn't very far removed at all from Compton's real career Test average of 28.70 and strike rate of 36.04.  Usage of data would have saved England this particular situation...

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