Turning Around Pakistan's T20i Fortunes - November 2019

8th November, 2019.

Email: Sportsanalyticsadvantage@gmail.com

Following today's comprehensive thrashing at the hands of Australia - and with it a crushing series defeat - Pakistan's new regime, under head coach and chief selector, Misbah ul Haq, are already at a crossroads.

Under former coach Mickey Arthur, Pakistan rose to world number one in the T20 international rankings, on the back of a magnificent run of results in the format.  From the end of the Champions Trophy in 2017 (which was also won by Pakistan) until the 7th August, 2019, when it was announced that Arthur was going to be leaving his role, Pakistan recorded - by some distance - the best win percentage in the format (77.27%) during this time period in matches between the traditional nine Test-playing nations - around 10% higher than rivals India and far in excess of any other nation.  

I've written before that - at least in franchise T20 where teams have level budgets - any T20 team with a long-term win percentage greater than 60% is doing incredibly well, so a huge amount of credit must be given to Arthur for achieving far in excess of this.  

The foundation of Pakistan's success was - as is often the case with successful T20 teams - a quality bowling line-up.  The graph below shows the bowling economy and bowling strike rates for these nine major international teams during this time period, in matches against each other:-

We can see here that Pakistan, in the ideal bottom-left hand corner which features low bowling economy and a low bowling strike rate, generated by far the best bowling numbers throughout this time period, and solely based on this, is difficult to dispute their ranking as the world's number one T20 international team at that point in time.  It should also be noted that the teams closest to the top-right hand corner had the worst win percentages of all the nine teams, further adding weight to the argument that the bowling ability of teams is a huge driver for their success.

The next chart below shows the batting data for each of the nine teams during this time period:-



Despite their overall success, Pakistan's batting data, although fairly competent, was not as strong as some teams, again implying that a big part of their success was bowling-driven.  In addition, their batting boundary percentage (16.04%) was third-worst among all the above teams, with only Sri Lanka and West Indies managing a lower figure.  Although Pakistan's batting wasn't the strongest, it is worth pointing out that they had two players inside the top 10 T20i runscorers in matches between these nine nations during this time period - Babar Azam (third) and Fakhar Zaman (eighth).  Shoaib Malik was also inside the top 20.

All told, though, it was clear that Pakistan under Mickey Arthur had a pretty solid strategy.  It is difficult to know the reasons as to why they adopted this strategy - I'd imagine the skill-set of the playing pool was a considerable reason - but it evidently worked.  

After Arthur left, Pakistan have played six matches and used an incredible 20 players already, which immediately hints at those in charge of selection not being aware of their best team, and/or making kneejerk reactions to isolated performances.  Pakistan are yet to win a T20 international after the departure of Arthur. 

In these matches, Pakistan have had huge issues with the bat, averaging 15.91 with a strike rate of 106.86, and boundary percentage of 11.19%.  These numbers are far below the worldwide T20 international mean figures and go some way to explaining why Pakistan have struggled under Misbah - even if a team had the five best bowlers in the world, it would struggle to defend totals with these batting numbers.  However, it's also worth pointing out that Pakistan's bowling has also been an issue under Misbah - in time period spanning the departure of Arthur, in the six matches so far, they have the worst team bowling average, the third worst economy and the worst bowling strike rate of all the nine major Test-playing nations.

So why have they struggled under Misbah?

I've already hinted at a rather volatile selection process with 20 players being used in six matches so far, and realistically, it's difficult to find statistical rationale for a number of the players selected.

As a player recruitment and strategy analyst working with various teams, I have developed an algorithm which takes a players performance across various T20 competitions and then generates expected figures for that player if they were to play a different T20 competition based on various factors, including opponent difficulty.  The basic, and logical, premise is that if a player is mediocre against mediocre quality opposition, they're going to be even worse against better opposition - so if a Pakistan batsman, for example, cannot score well against the average Pakistan domestic bowlers, they're hardly likely to thrive against the likes of Mitchell Starc, Pat Cummins, Jasprit Bumrah, Rashid Khan or Jofra Archer.  In theory, the possibilities are endless - as an extreme example, I could look at English Second XI players and see how they would be expected to perform in the IPL, while I could do the same with the Tamil Nadu T20 League in India as well.  I can also use data from the Pakistan National T20 Cup and the PSL (as well as any overseas leagues played) in order to establish how Pakistan's players would be expected to perform in T20 internationals, which is clearly useful to assess their current T20i playing pool.  

What I have done below is assess the players used by Pakistan in T20 internationals from 2017 onwards, as well as adding a few further players who my algorithm highlighted (in red) as being worthy of consideration:

BATSMEN/ALL-ROUNDERS:-

Player

Selected by Misbah

Expected T20i Batting Average

Expected T20i Batting Strike Rate

Boundary % (All T20 Leagues & T20i, 2017+)

Non-Boundary Strike Rate (All T20 Leagues & T20i, 2017+)

Balls Faced Per Dismissal (All T20 Leagues & T20i, 2017+)








Ahmed Shehzad

Y

33.95

123.58

16.16

64.25

28.78

Asif Ali

Y

27.48

161.72

21.59

68.36

17.57

Babar Azam

Y

41.45

126.43

14.53

76.52

33.36

Fakhar Zaman

Y

24.32

133.31

18.94

60.42

18.56

Haris Sohail

Y

17.45

108.40

11.37

65.18

17.24

Hussain Talat


28.49

121.93

13.95

72.64

24.57

Iftikhar Ahmed

Y

21.81

110.11

11.14

71.36

20.95

Imad Wasim

Y

16.28

128.34

14.44

73.00

13.85

Imam-ul-Haq

Y

27.73

128.12

16.19

69.04

23.56

Kamran Akmal


28.31

141.67

24.20

45.57

20.73

Khushdil Shah

Y

27.13

122.89

16.46

63.94

21.47

Mohammad Hafeez


23.84

113.34

13.86

60.92

21.88

Mohammad Nawaz


23.46

119.34

13.16

64.88

20.84

Mohammad Rizwan

Y

24.66

112.13

11.06

74.16

23.50

Sahibzada Farhan


21.95

102.79

13.28

52.44

23.67

Sarfaraz Ahmed

Y

25.89

127.47

14.13

76.30

20.91

Shan Masood


25.74

130.05

19.17

65.22

22.36

Shoaib Malik


38.63

131.85

14.25

76.12

30.08

Sohail Akhtar


26.76

126.26

17.38

55.26

22.58

Umar Akmal

Y

19.51

116.65

15.10

60.04

17.06

Umar Amin


25.94

123.60

16.37

60.54

23.35



The table above gives immediate insight into Pakistan's batting struggles.  Only a few players selected in the last three years (Asif Ali, Fakhar Zaman and Kamran Akmal) are much above-average boundary hitters, while Sohail Akhtar - who has never been capped - and Shan Masood also could contribute in that area.

Akhtar's inclusion in this table was interesting, as he and Masood were the only batsmen not capped during this time period that my algorithm highlighted as being likely to be better than some of the existing squad, indicating that the cupboard is pretty much bare with regards to batting talent, and particularly, boundary-hitting talent in the Pakistan domestic game which is ready to translate itself to T20 international success.  Those in charge of the talent pathway in Pakistan should focus on developing boundary hitters, because the current playing pool is rather rotation-orientated and fairly ill-equipped to post 160+ totals on a regular basis.  Another point worthy of highlighting is a lack of bowling options among top six batsmen, and genuine all-rounders.  Only Imad Wasim and Mohammad Nawaz can come remotely close to this bracket, and both still look below-average batsmen at this level based on my expected data.  Mohammad Hafeez is a bowling option in the top six, but his batting data is unimpressive, which should be a concern at the age of 39.

There are a number of batsmen in this list who are shown to be likely below-average on the international stage, and there certainly isn't much statistical rationale behind selecting a number of batsmen in the current squad.  I'll re-iterate again - it is both illogical and unreasonable to expect players to improve on their domestic numbers against better bowlers in the international arena.

BOWLING:-

Player

Selected by Misbah

Expected T20i Bowling Average

Expected T20i Bowling Economy

Expected T20i Bowling Strike Rate






Faheem Ashraf

Y

20.78

8.17

15.26

Hasan Ali


20.37

7.93

15.41

Imad Wasim

Y

27.84

6.61

25.27

Mohammad Amir

Y

19.60

7.24

16.24

Mohammad Hafeez


25.06

7.02

21.42

Mohammad Hasnain

Y

21.45

9.06

14.21

Mohammad Irfan

Y

26.57

7.43

21.46

Mohammad Nawaz


27.65

7.79

21.30

Muhammad Musa

Y

26.04

10.89

14.35

Rumman Raees


25.83

8.52

18.19

Shadab Khan

Y

20.91

7.14

17.57

Shaheen Shah Afridi


21.61

8.09

16.03

Sohail Tanvir


25.97

8.06

19.33

Usman Shinwari

Y

20.93

8.11

15.48

Wahab Riaz

Y

22.34

7.61

17.61

Waqas Maqsood


33.35

9.73

20.57

Amad Butt


19.99

9.16

13.09

Haris Rauf


25.16

8.66

17.43

Umar Khan *


16.62

7.65

13.04

Mohammad Ilyas


24.53

8.68

16.96


While Pakistan have considerable problems finding above-average batsmen in their talent pool, they don't have a shortage of bowling options.  The vast majority of the 16 frontline bowlers used by Pakistan from 2017 onwards have excellent expected T20i data and they have a number of real elite T20i options, including numerous young players with extremely high potential who should be able to form the spine of their bowling attack for years to come.  In addition, there's plenty of variety there with plenty of quality left-arm quicks as well as right-arm options, and legspin and offspin choices as well.

Not only this, there are four further bowling options which my algorithm highlighted as further potential bowling options (Umar Khan has a small sample size of data) which further bolsters an already rich talent pool from a bowling perspective. 

It's also worth noting that I consider Pakistan to over-promote young bowling talent before it is ready.  The numbers above show that Mohammad Hasnain and Muhammad Musa are still likely to contribute worse than average economy on the international stage, although it's obvious that they have high future potential.  From 2017 onwards, nine players aged below 20 played T20i cricket for one of the nine main Test-playing nations and four were from Pakistan - I'd recommend that they ensure that their players have a robust domestic sample size of data (either in Pakistan or franchise leagues abroad) prior to picking them against higher quality opposition.

So how do they turn this around?

The first area to debate must be - What criteria do we judge success for Pakistan currently?  

It is quite rare to get a straight answer to questions along this line from most team coaches or decision makers regardless of the format or the sport but the three areas below perhaps summarise what I would consider reasonable criteria:-

1) Preparation of a competitive and consistent squad with settled selection capable of performing well in pace-orientated conditions at the T20 World Cup in Australia next year.
2) Reaching long-term goals via giving young players exposure.
3) Obtaining a high win percentage in numerous continents and conditions, as well as in the subcontinent.

Focusing on the first option would seem like a logical current priority, while being mindful of the second option - perhaps making 'line calls' in favour of younger players when data suggests there isn't much between a younger player and an older option.  It should be hoped that with strong selection criteria, number three will be a simple natural consequence.

This strong selection criteria required would be an obvious starting point for turning the situation around - based on the above numbers, it is clear that a number of batsmen selected currently are not likely to be above-average batsmen in the international arena and it is difficult to make a case for quite a few at the time of writing.  The following batsmen (listed in alphabetical order) look like likely to have the most success at in T20i cricket currently and should comprise a shortlist for batting roles, although it's fair to suggest that I don't consider many to be particularly above-average:-

Ahmed Shehzad
Asif Ali
Babar Azam
Fakhar Zaman
Hussain Talat
Imam-ul-Haq
Kamran Akmal
Sarfraz Ahmed
Shan Masood
Shoaib Malik
Sohail Akhtar

Selecting batsmen from this group and using them in their best roles should be able to get Pakistan regularly up to the 150-160 mark in T20 internationals which should be capable of being defended well by a world-class bowling attack (assuming they get selection right in this area as well).  This brings us back to a similar playing style to under Mickey Arthur, which was so successful, and follows the general blueprint of T20 success in both T20 internationals and the various worldwide T20 franchise leagues.

If this article has given you insight into the data that Sports Analytics Advantage can offer cricket teams around the world in formulating team strategies, selection, draft or auction plans, or any other work, please feel free to enquire at sportsanalyticsadvantage@gmail.com.
Comments