Solving the Impossible: How to contain MS Dhoni in death overs

26th April, 2018.

Email: Sportsanalyticsadvantage@gmail.com

Yesterday in the IPL, we again saw captain MS Dhoni haul Chennai Super Kings to another victory, this time against the Royal Challengers Bangalore, chasing 205 with several balls to spare.  

His stats now for this year's IPL make for incredible reading, having accumulated 209 runs from three completed innings (average 69.67) at a strike rate of 165.87, and on the surface, it would appear that reports of Dhoni's demise have been grossly exaggerated.

However, closer analysis of those 209 runs in the current IPL is worthwhile - he's struck at an incredible 224 strike rate against pace, but a mere 121 versus spin, and this large discrepancy is a trend that has beset Dhoni in recent years.

In fact, from 2016 onwards in T20 (IPL and T20 internationals), Dhoni has struck at 157 per 100 balls faced against pace bowling, but just 110 against spin, so across a big sample of data, there is a huge difference in his skill-sets against various types of bowling.

However, in this time period, just short of 39% of the balls that Dhoni has faced in T20 have been against spinners, so it would appear that teams are not necessarily taking advantage of Dhoni's issues scoring quickly against spin bowling.

With Dhoni often used in the role of a finisher, this isn't necessarily surprising, as he faces a lot of his balls faced in the death overs, which tend to be bowled by pace bowlers.  In fact, during this time period, he's faced 39.3% of his deliveries in overs 17-20, a figure likely to be one of the highest in world cricket, and of these death balls faced, 85.90% were bowled by pace bowlers.

Generally, T20 teams don't like to bowl spin at the death, with the perception that it is more likely to leak runs than pace bowling, so the fact that Dhoni has faced such a high percentage of pace bowling in the death overs isn't a surprise.  However, what is more of a shock is that teams haven't often taken this opportunity to attempt spin in order to combat his strike rate weaknesses against spinners.

I ran through Dhoni's data from overs 17-20 from 2016 onwards, and this analysis generated the following figures:-



Balls

Runs

SR

IPL, Overall


192

367

191.15

T20i, Overall


191

321

168.06

2016+, Overall


383

688

179.63

IPL, vs Pace


161

338

209.94

IPL, vs Spin


31

29

93.55

T20i, vs Pace


168

287

170.83

T20i, vs Spin


23

34

147.83

Overall, vs Pace


329

625

189.97

Overall, vs Spin


54

63

116.67


First off, we can see that Dhoni has excelled at the death in the IPL in particular, recording a strike rate of just below 200 runs per 100 balls, and his death strike rate in T20 internationals is still solid, at a lower 168.  

However, we can see that in the IPL in particular, there is an incredible difference between his strike rates at the death between pace and spin, where he has struck at almost 210 against pace, but just 94 versus spin.  While there is a smaller difference in T20 internationals, Dhoni still demonstrates a bias towards pace bowling, and it's worth noting that with this T20 international spin data, he scored a combined 20 off 8 balls against Sri Lankan spinners Dilruwan Perera and Milinda Siriwardene - players arguably not of franchise league quality.

All told, from 2016 onwards, Dhoni has struck at 189.97 against pace at the death and just 116.67 against spin bowling at the death, and some pace bowlers in particular have been singled out for brutal treatment by the CSK captain:-

Bowler

Role

Balls

Runs

SR






Paterson

Pace

11

29

263.64

Siraj

Pace

11

29

263.64

Bhuvneshwar Kumar

Pace

20

49

245.00

Tye

Pace

12

28

233.33

Milne

Pace

10

23

230.00

Mohit Sharma

Pace

14

30

214.29

Kaul

Pace

10

21

210.00

Morris

Pace

13

23

176.92

DJ Bravo

Pace

25

44

176.00

Thisara Perera

Pace

15

26

173.33

McClenaghan

Pace

11

17

154.55

Bumrah

Pace

19

26

136.84

Jordan

Pace

19

25

131.58

Mustafizur

Pace

11

13

118.18

Mills

Pace

10

11

110.00


The list above illustrates the bowlers who have bowled at least 10 balls against Dhoni at the death from 2016 onwards, and just four - Jasprit Bumrah, Chris Jordan, Mustafizur and Tymal Mills - have managed to restrict him to a strike rate below 150.  

Given all of this data, it seems absolutely logical to think that when Dhoni is batting at the death, teams would want to combat his prowess against pace bowling by bowling spin, so why don't they?

Certainly, it is difficult to understand why they don't, but I believe a great deal stems from the fact that too many teams are run by ex-players who are still far too restricted to traditional, conventional thinking.  How many times has it even crossed their mind to bowl spin to Dhoni at the death?

It is my belief that it will take an analyst to have a real voice in the IPL, or become a team's General Manager, for this conventional thinking to cease.  Currently, I am not convinced that any IPL analyst has a huge impact, or the final say, on team selection, recruitment and retention decisions for any T20 franchise around the world, and until this changes, teams will continue to make baffling decisions such as bowling Corey Anderson at the death against MS Dhoni.  It is difficult to think that the talented young spinner, Washington Sundar, could have been any worse.

Finally, I want to leave you with one final thought.  In his excellent article yesterday, the superb cricket journalist Tim Wigmore discussed the lack of accountability that T20 coaches face.  For this article, I provided Tim with some data about Daniel Vettori, the Royal Challengers Bangalore coach, which demonstrated that Vettori has a very poor win percentage across his T20 coaching career, and below is his current win-loss record at the franchises that he has coached:-


Completed Matches

Wins

Win %





Brisbane Heat, 2015-16

8

3

37.50

Brisbane Heat, 2016-17

9

5

55.56

Brisbane Heat, 2017-18

10

4

40.00

Brisbane Heat, Overall

27

12

44.44

Middlesex, 2017

12

5

41.67

Royal Challengers Bangalore, 2014

14

5

35.71

Royal Challengers Bangalore, 2015

14

8

57.14

Royal Challengers Bangalore, 2016

16

9

56.25

Royal Challengers Bangalore, 2017

14

3

21.43

Royal Challengers Bangalore, 2018

6

2

33.33

Royal Challengers Bangalore, Overall

64

27

42.19

All Teams

103

44

42.72



We can see that at none of the Brisbane Heat, Middlesex or Royal Challengers Bangalore that he's managed an overall win percentage of over 45%, and only in three of nine seasons coached has he managed a win percentage over 50%.  

Perhaps this explains the bizarre decisions taken by Royal Challengers Bangalore on a regular basis - we've seen in many sports, notably baseball and soccer, that being a good, high-profile ex-player is not close to being a guarantee that they will make a good manager, and it would appear that cricket, and perhaps RCB as well, will learn this the hard way.
Comments