Tuesday, 22 April 2014

A Brief Statistical Analysis of the Revolutionary Test Opener: Virender Sehwag

If Virender Sehwag gets the chance to play Test cricket again, it would mean that the present generations have failed big time and Indian cricket is going backward to find a solution. That’s unlikely to happen, and is not good for the health for Indian cricket as well. Whether he returns or not that won’t change anything, he has given us enough memories to cherish about.

According to John Wright, “He didn’t redefine his game because of his batting position. He redefined the position with his batting.” One cannot disagree with this simple yet strong statement. Aakash Chopra, probably the finest cricket writer amongst Indian test cricketers, goes one step further by saying, “It would be fair to say that opening in Test match cricket can be divided into two eras-pre-Sehwag and post-Sehwag.” Sehwag has literally inspired David Warner into test cricket and his performances have inspired captains and coaches to try players like Tilakaratne Dilshan and Shane Watson as test openers. Without Sehwag’s success and more importantly success with his method, it is difficult to imagine test debut of the mentioned players.

He has shown others it is possible to score more than run-a-ball in test cricket and that too for an extended period. An opener with more than 14 scores in excess of 150 or more out of 23 hundreds with a career strike rate of 82.23, Dileep Premachandran’s observation of Sehwag as a sprinter with a marathon runner’s stamina is more than appropriate.

There have been many excellent articles written about Sehwag, his method and the visual delight he gives to an audience. What I intend to do here is a mere number based comparison to show where Sehwag stands among the greats of his time. I consider the players who have scored more than 6000 runs at an average of more than 45 after Sehwag’s debut. The reason behind putting the constraint on the time period is that at least we are reducing few variables (i.e. the condition of pitches, batting equipment, rules and regulations) from the many variable space on which a measure of batting greatness possibly depends.  So, the statistical error, and the absurdity of comparison based on numbers decreases.

So, the list contains 19 players who have dominated test cricket in last 15 years. The temptation to include Inzamam-Ul-Haq, Steve Waugh and particularly Brian Lara in a list which already contains Ponting, Tendulkar, Kallis and Dravid was very hard to avoid but I stuck to the list of 19 players. The list is arranged in the descending order of batting average.

Serial No.
Name
Runs
Average
1
K Sangakkara
10368
60.27
2
J Kallis
9949
59.22
3
S Chanderpaul
8386
58.23
4
R Ponting
10548
54.93
5
M Hayden
7306
53.72
6
Y Khan
6575
53.45
7
AB de Villiers
7168
51.94
8
R Dravid
9255
51.7
9
M Hussey
6235
51.52
10
M Clarke
8240
51.5
11
S Tendulkar
9002
51.44
12
H Amla
6214
51.35
13
M Jayawardene
8853
51.17
14
V Sehwag
8586
49.34
15
VVS Laxman
7343
48.62
16
G. smith
9265
48.25
17
A Cook
8181
47.28
18
K Pietersen
8047
46.51
19
I Bell
6722
45.41
Table 1: List of all the players who have scored more than 6000 test runs at an average of 45 or more after Sehwag’s debut test match (03.11.2001).
           
As it turns out Sangakkara, Kallis and Chanderpaul lead the way with overall average while Ponting has scored more runs than any other in the mentioned time interval. In an unrelated note, there is nothing much to separate between the numbers of two Indian legends, Dravid and Tendulkar. Sehwag occupies 14th spot in the list and more importantly only one of the three openers in the list, Matthew Hayden and Greame Smith being the other two.

There are some common criticism against Sehwag’s performance which I have come across while talking with friends and following the discussion of cricket community on social platforms. This motivates me to break the numbers depending on the point of argument and compare his number with those of the mentioned successful batsmen of his generation.

(a)   Flat track bully:
For this purpose, I break down the overall average into home and away (and mutual venue when Pakistan was one of the teams) average. This makes sense to me as we can get a measure of success of a subcontinent player in pace friendly conditions and also the ability of English, South African and Australian batsmen in spinning tracks. The difference between the away and home average is calculated which shows the relative adaptability of individual player. The results are summarized in Table 2 (a). For the ease of comparison, the name of the players are re-arranged in the decreasing order of  difference in Table 2 (b).

Serial No.
Name
Home Average
(X)
Away + Neutral Venue Average
(Y)
Difference
D= (Y)-(X)
1
K Sangakkara
67.34
53.68
-13.66
2
J Kallis
63.56
54.55
-9.01
3
S Chanderpaul
66.06
52.14
-13.92
4
R Ponting
60.47
48.92
-11.55
5
M Hayden
62.29
42.52
-19.77
6
Y Khan
65.26
50.28
-14.98
7
AB de Villiers
45.85
59.85
14
8
R Dravid
51.3
51.96
0.66
9
M Hussey
61.19
41.37
-19.82
10
M Clarke
61.9
42.77
-19.13
11
S Tendulkar
46.85
55.57
8.72
12
H Amla
45.74
56.06
10.32
13
M Jayawardene
61.28
42.47
-18.81
14
V Sehwag
54.13
44.65
-9.48
15
VVS Laxman
55.75
44.88
-10.87
16
G smith
41.52
54.98
13.46
17
A Cook
52.75
41.88
-10.87
18
K Pietersen
43.55
49.93
6.38
19
I Bell
53.68
38
-15.68
Table 2 (a): Performance difference of players in home and away matches.

Serial No.
Name
Difference
D= (Y)-(X)
1
AB de Villiers
14
2
G smith
13.46
3
H Amla
10.32
4
S Tendulkar
8.72
5
K Pietersen
6.38
6
R Dravid
0.66
7
J Kallis
-9.01
8
V Sehwag
-9.48
9
VVS Laxman
-10.87
10
A Cook
-10.87
11
R Ponting
-11.55
12
K Sangakkara
-13.66
13
S Chanderpaul
-13.92
14
Y Khan
-14.98
15
I Bell
-15.68
16
M Jayawardene
-18.81
17
M Clarke
-19.13
18
M Hayden
-19.77
19
M Hussey
-19.82
Table 2 (b): Re-arrangement of Table 2(a) in the decreasing order of difference (D) between away and home matches.
           
What the difference column shows us that only 6 players in the list have a positive difference i.e. better away record. The list is headed by 3 South Africans (de Villiers, Smith and Amla) and the 3 bottom places are occupied by 3 Australians (Hussey, Hayden and Clarke). Rahul Dravid is the batsman whose numbers vary the least with respect to conditions. Sehwag is placed at No. 8 in the list. But based on the difference it would be misleading to say he is a better away performer than Sangakkara or Ponting, his absolute away average is still lower than them. What it shows the relative measure of a player’s own performance in home and away matches. But to suggest that Sehwag is a flat track bulley is no more meaningful than to suggest Hussey and Clarke are lesser player of spin bowling.

(b)   First innings specialist:
From outset it is easy to expect that Sehwag’s or most of the player’s second innings average will be less than their first innings average. In order to get a quantitative picture, the overall average is broken down into 2 innings and the comparison is shown in Fig. 3(a). Like the earlier case, we calculate the difference between 2nd innings and 1st innings average, and rearrange the players in the descending order of difference in Table 3(b).

Serial No.
Name
1st Innings Average
(X)
2nd Innings Average
(Y)
Difference
D= (Y)-(X)
1
K Sangakkara
62.21
57.32
-4.89
2
J Kallis
58.53
60.45
1.92
3
S Chanderpaul
64.64
49
-15.64
4
R Ponting
59.96
46.73
-13.23
5
M Hayden
51.2
57.53
6.33
6
Y Khan
54.69
51.64
-3.05
7
AB de Villiers
56.38
44.81
-11.57
8
R Dravid
59.72
39.67
-20.05
9
M Hussey
55.12
45.87
-9.25
10
M Clarke
57.79
41.54
-16.25
11
S Tendulkar
58.44
39.86
-18.58
12
H Amla
48.68
55.35
6.67
13
M Jayawardene
60.93
36.1
-24.83
14
V Sehwag
62.5
30.25
-32.25
15
VVS Laxman
48.01
49.77
1.76
16
G smith
49.63
46.19
-3.44
17
A Cook
53.49
38.38
-15.11
18
K Pietersen
47.28
45.43
-1.85
19
I Bell
51.04
37.16
-13.88
Table 3 (a): Performance difference of players in 1st and 2nd innings of match.

Serial No.
Name
Difference
D= (Y)-(X)
1
H Amla
6.67
2
M Hayden
6.33
3
J Kallis
1.92
4
VVS Laxman
1.76
5
K Pietersen
-1.85
6
Y Khan
-3.05
7
G smith
-3.44
8
K Sangakkara
-4.89
9
M Hussey
-9.25
10
AB de Villiers
-11.57
11
R Ponting
-13.23
12
I Bell
-13.88
13
A Cook
-15.11
14
S Chanderpaul
-15.64
15
M Clarke
-16.25
16
S Tendulkar
-18.58
17
R Dravid
-20.05
18
M Jayawardene
-24.83
19
V Sehwag
-32.25
Table 3 (b): Re-arrangement of Table 3(a) in the decreasing order of difference (D) between 2nd innings and 1st innings average

Hashim Amla leads the way this time, followed by Matthew Hayden and Jack Kallis. It comes as no surprise that VVS Laxman, the crisis man, has the highest difference as well as the highest absolute 2nd innings average (49.77) amongst the Indian batsmen. He is roughly 10 runs clear ahead of Dravid and Tendulkar. Sehwag is at the bottom of the list, both in terms of 2nd innings average and the difference. In fact his first innings average (62.5) is more than twice that of his 2nd innings average (30.25). It’s not really a bad thing to do the bulk of scoring in the first innings and set the game up for your team. The numbers really confirm the notion that Sehwag is not the man for epic 2nd innings chases, memorable rear-guard performances.

           This is by no means a complete analysis. In fact, what I believe, a fractionally better idea to test (only statistically) the consistency of a player, is to break his numbers into several sections, based on his performances
·         Against all the oppositions
·         On all the countries
·         During each year of his carrier

Say for the third heading, we will get Sehwag’s average in each of his seasons. Then we can have an idea about the standard deviation of his all year averages and the mean is of course his career average. A player with maximum mean supposed to be the best (only statistically) of his time (in this case Sangakkara), while a minimum standard deviation is a measure of his consistency. Though most of the players go through purple patches and lean patches, and the height and depth of these crests and troughs will affect the standard deviation. This can be repeated for all the 3 headings mentioned earlier. But, for the time being I restrain myself from such a crazy and even meaningless number crunching.

It’s true to certain extent that numbers don’t lie but at the same time they don’t tell the complete story either. For people of our generation will fail to understand why with a career average of 39.62, Victor Trumper is considered as the first great batsman from Australia and almost always preceded by the word “immortal”. The sheer joy, the anxiety, the air of unpredictability and the frustration when they got out, makes player like Sehwag inaccessible to numbers. Our generation can consider ourselves very fortunate to see Virender Sehwag’s batting, that is an experience to tell our grandchildren. Long live Virender Sehwag, your memories will remain with us.

2 comments: