Batting August 17, 2009

# Comparing Test batsmen with their peers

Having done a peer comparison analysis of bowlers, it's now the turn of batsmen

I have done a lot of cricket analysis work over the past 20+ years. I love doing all this work. However once a while a new idea comes across which I consider as a watershed moment in my analytic efforts. The idea of comparing a player with peer players (the base idea of which was provided by Abdulla) is one such spark. I am very excited about this since it is one of the truest measures of a players' capabilities. This is a follow-up article to the one on Test bowlers.

The idea is to compare a player's performances with his peers. The comparison with one's own team is a limited step and is quite useful. However the real comparison is with all the peer players since it takes perfect care of the vexed question of a player playing in a very strong team. I had done this in a limited way for ODI Strike Rates. Now I have extended this to Test Players in a much more extended manner as explained below.

1. For each player, create a match subset of their career limits, in other words from their first to last Tests. For Tendulkar it is 1127(1989) to 1918(2009), a subset of 792 Tests, the longest span for any player.

2. Sum the three main data elements, Innings, Not Outs, and Runs Scored for all the players for these matches. The Batting Average is used for comparison since this is the most accepted of all measures.

3. Subtract the player's own career figures from the total for the match subset and post these figures as a database segment. Even though the players' own numbers are quite low compared to the match subsets (Tendulkar 12773 out of 749558 runs) and the impact of this subtraction is minimal, it is done to get an exact peer segment.

4. For batsmen, first the base table is created. This table compares the batsman's bating average with the composite average of all batsmen during his playing span. This covers all batsmen since separate comparisons are done for specialized batting positions such as Opening, Middle order and Late order.

I have not done a separation by period. This is a pure peer comparison, cutting across all divisions.

First let us look at the basic Batsman table.

1. Batsman Peer comparisons - Basic table

```>= 2000 Test runs

No.Batsman         Cty  Runs  Avge From- To (Mat) <------Peer-----> Ratio
Inns   Runs Avge

1.Bradman D.G     Aus  6996 99.94 1928-1948(128)  3722 113802 30.58 3.27
2.EdeC Weekes     Win  4455 58.62 1948-1958(161)  4829 138734 28.73 2.04
3.Sutcliffe H     Eng  4555 60.73 1924-1935( 91)  2600  78032 30.01 2.02
4.Pollock R.G     Saf  2256 60.97 1963-1970(126)  3900 118766 30.45 2.00
5.Walcott C.L     Win  3798 56.69 1948-1960(199)  5982 169812 28.39 2.00
6.Barrington K.F  Eng  6806 58.67 1955-1968(234)  7072 207904 29.40 2.00
7.Headley G.A     Win  2190 60.83 1930-1954(194)  5745 177352 30.87 1.97
8.Hobbs J.B       Eng  5410 56.95 1908-1930(102)  3069  88958 28.99 1.96
9.Sobers G.St.A   Win  8032 57.78 1954-1974(353) 10721 317459 29.61 1.95
10.Hammond W.R     Eng  7249 58.46 1927-1947(117)  3344 101007 30.21 1.94
11.Hutton L        Eng  6971 56.67 1937-1955(143)  4149 123572 29.78 1.90
12.Ponting R.T     Aus 11267 56.05 1995-2009(612) 18664 577309 30.93 1.81
13.Chappell G.S    Aus  7110 53.86 1970-1984(300)  8979 270067 30.08 1.79
14.Tendulkar S.R   Ind 12773 54.59 1989-2009(792) 24004 736785 30.69 1.78
15.Kallis J.H      Saf 10277 54.66 1995-2009(599) 18270 564569 30.90 1.77
16.Javed Miandad   Pak  8832 52.57 1976-1993(460) 13470 401608 29.81 1.76
17.Mohammad Yousuf Pak  7023 54.87 1998-2009(522) 16015 500382 31.24 1.76
18.Flower A        Zim  4794 51.55 1992-2002(431) 13040 384939 29.52 1.75
19.Lara B.C        Win 11953 52.89 1990-2006(661) 20051 607578 30.30 1.75
20.Sangakkara K.C  Slk  7095 55.43 2000-2009(421) 12848 411708 32.04 1.73
```
Even though the batsman peer span is shown in years, the actual computations are done for the exact match of debut onwards. The years make more sense while reading the table. The "inns" value shown on these tables is after subtracting the Not outs.

No surprise at the first placed batsmen. It would have been a shock if it had been anyone else. What is surprising is the ratio of Bradman. An amazing 3.27. Weekes is the first among 9 equals who have ratios from 1.94 to 2.04. These 10 batsmen are among the best ever, all 10 having played their game before 1970.

The batsman with the highest ratio among the contemporary players is Ponting, with a ratio of 1.81, followed by Tendulkar with 1.78 and the unheralded Kallis with 1.77. This, despite the commonly percieved notions of weaker teams, and hence cheaper runs. Note the high placement of Andy Flower.

It should be noted that the peer averages are comparable across ages, at either side of 30. Mohommad Yousuf's peer average is the highest at 31.24. His span is 1998-2009. As also Kallis'. The lowest Peer average numbers are for the early 1950s.

Now we come to the comparison tables for specialized batting positions. These are determined by isolating the runs scored by batsmen in these specialized positions only and then comparing with runs scored in these positions by other batsmen. Opening is determined by the positions 1-2, Middle order by positions 3-7 and Late order by positions 8-11. The only question mark could be with no.7. However when you realize that top-quality batsmen such as Gilchrist, Healy, Knott, Marsh, Imran, Kapil, Botham, S Pollock, Flintoff, Boucher et al have scored over 25,000 Test runs amongst them at no.7 position, it has to belong to the Middle order classification.

First let us look at the Opening position. This time I have also shown the Batting Position Average value. This is the average of the batting position the batsman has batted in, with the opening positions being considered as no.2. Thus a value of 2.00 means that the batsman has batted in the opening positions only.

2. Batsman Peer comparisons - Opening batsmen

```>= 2500 opening runs

No.Batsman          Cty  BPos Inns Runs  Avge  <------Peer------> Ratio
Avge Out              Inns    Runs  Avge

1.Sutcliffe H      Eng  2.05  74  4522 61.11   507   18443 36.38  1.68
2.Hobbs J.B        Eng  2.15  91  5130 56.37   591   21419 36.24  1.56
3.Hutton L         Eng  2.18 119  6721 56.48   846   30900 36.52  1.55
4.Simpson R.B      Aus  3.27  66  3664 55.52  2578   94513 36.66  1.51
5.Amiss D.L        Eng  2.50  61  3276 53.70  1318   49067 37.23  1.44
6.Hayden M.L       Aus  2.00 170  8626 50.74  4339  153809 35.45  1.43
7.Gavaskar S.M     Ind  2.21 191  9607 50.30  2439   86489 35.46  1.42
8.Saeed Anwar      Pak  2.11  84  3957 47.11  2677   90241 33.71  1.40
9.Smith G.C        Saf  2.21 118  6108 51.76  2115   78959 37.33  1.39
10.Sehwag V         Ind  2.36 105  5378 51.22  2360   88396 37.46  1.37
11.Langer J.L       Aus  2.42 106  5112 48.23  4127  146726 35.55  1.36
12.Gibbs H.H        Saf  2.64 111  5242 47.23  3483  124196 35.66  1.32
13.Boycott G        Eng  2.02 168  8091 48.16  2277   82894 36.40  1.32
14.Lawry W.M        Aus  2.00 111  5234 47.15  1086   39476 36.35  1.30
15.Slater M.J       Aus  2.00 124  5312 42.84  2154   71763 33.32  1.29
16.Greenidge C.G    Win  2.03 166  7488 45.11  2684   94699 35.28  1.28
17.Boon D.C         Aus  2.85  58  2614 45.07  2131   75453 35.41  1.27
18.Hunte C.C        Win  2.00  72  3245 45.07  1082   38410 35.50  1.27
19.Stewart A.J      Eng  3.91  75  3348 44.64  3464  122407 35.34  1.26
20.Vaughan M.P      Eng  2.86  68  3093 45.49  2803  101414 36.18  1.26
```
The three great English openers lead the table. Then Simpson and another top quality English opener, Amiss, although Amiss' contemporary openers posted a high average. Hayden and Gavaskar clock in next despite the somewhat lower peer averages. It is also an indicator that more often than not Gavaskar waged a lone battle. The next three positions are held by openers from the current and immediately precding era.

Alec Stewart is one of the very few batsmen who has scored enough runs in both opening and middle order positions to qualify for both lists. His opening average is considerably better and he is in the 19th position. Readers should not forget that the runs in the table are the runs scored in the opening positions only.

3. Batsman Peer comparisons - Middle order batsmen

```>= 4000 middle order runs

No.Batsman          Cty  BPos Inns Runs  Avge  <------Peer------> Ratio
Avge Out              Inns    Runs  Avge

1.Bradman D.G      Aus  3.65  70  6996 99.94  1841   64844 35.22  2.84
2.EdeC Weekes      Win  4.16  75  4399 58.65  2388   79001 33.08  1.77
3.Sobers G.St.A    Win  5.09 128  7658 59.83  5363  185285 34.55  1.73
4.Barrington K.F   Eng  4.07 113  6604 58.44  3512  122194 34.79  1.68
5.Hammond W.R      Eng  3.70 120  6934 57.78  1628   57387 35.25  1.64
6.Chappell G.S     Aus  4.04 132  7110 53.86  4450  156700 35.21  1.53
7.Compton D.C.S    Eng  4.34 114  5805 50.92  2569   86396 33.63  1.51
8.Ponting R.T      Aus  3.85 201 11267 56.05  9177  344014 37.49  1.50
9.Javed Miandad    Pak  4.24 167  8789 52.63  6639  234403 35.31  1.49
10.Tendulkar S.R    Ind  4.28 233 12758 54.76 11806  437913 37.09  1.48
11.May P.B.H        Eng  3.66  96  4525 47.14  2593   83403 32.16  1.47
12.Kallis J.H       Saf  3.80 188 10277 54.66  8981  336648 37.48  1.46
13.Sangakkara K.C   Slk  3.09 121  6845 56.57  6328  246703 38.99  1.45
14.Harvey R.N       Aus  3.65 126  6147 48.79  3651  122850 33.65  1.45
15.Lara B.C         Win  3.78 223 11828 53.04  9833  359979 36.61  1.45
16.Dravid R         Ind  3.27 191 10334 54.10  8859  332724 37.56  1.44
17.Mohammad Yousuf  Pak  4.71 128  7023 54.87  7884  300580 38.13  1.44
18.Waugh S.R        Aus  5.42 211 10910 51.71  9473  341102 36.01  1.44
19.Flower A         Zim  5.03  93  4786 51.46  6408  230728 36.01  1.43
20.Border A.R       Aus  4.70 220 11116 50.53  5914  209290 35.39  1.43
```
The middle order table shows no surprises. Again Mohammad Yousuf's peer batsmen batting average is quite high, only exceeded by Sangakkara's peer average. The early 50s show the lowest middle order batsman averages.

4. Batsman Peer comparisons - Late order batsmen

```( >=500 late order runs and BPos avge >8.0)

No.Batsman          Cty  BPos Inns Runs  Avge <------Peer------>  Ratio
Avge Out              Inns    Runs  Avge

1.Johnson M.G      Aus  9.03  22   762 34.64   695   11199 16.11  2.15
2.Strang P.A       Zim  8.17  25   737 29.48  2546   36143 14.20  2.08
3.Vettori D.L      Nzl  8.34  98  2959 30.19  4851   73245 15.10  2.00
4.Symcox P.L       Saf  8.44  23   668 29.04  1781   25879 14.53  2.00
5.Broad S.C.J      Eng  8.03  20   628 31.40   635   10389 16.36  1.92
6.Reiffel P.R      Aus  8.40  34   936 27.53  1855   26951 14.53  1.89
7.Blignaut A.M     Zim  8.31  30   835 27.83  1944   29804 15.33  1.82
8.More K.S         Ind  8.33  44  1180 26.82  1458   22140 15.19  1.77
9.Smith I.D.S      Nzl  8.34  60  1667 27.78  2418   38154 15.78  1.76
10.Boje N           Saf  8.10  42  1125 26.79  2843   43787 15.40  1.74
11.O'Keeffe K.J     Aus  8.06  23   606 26.35  1076   16462 15.30  1.72
12.Nash D.J         Nzl  8.82  30   729 24.30  3147   44928 14.28  1.70
13.Vaas WPUJC       Slk  8.09 109  2783 25.53  5557   83365 15.00  1.70
14.Chandana U.D.U   Slk  8.29  21   519 24.71  2567   38534 15.01  1.65
15.Verity H         Eng  8.52  28   620 22.14   506    7101 14.03  1.58
16.Ghavri K.D       Ind  8.53  41   900 21.95  1281   18099 14.13  1.55
17.Wasim Akram      Pak  8.14  97  2160 22.27  4784   70503 14.74  1.51
18.Madan Lal S      Ind  8.18  30   669 22.30  2577   38789 15.05  1.48
19.Wardle J.H       Eng  8.10  26   568 21.85  1197   18002 15.04  1.45
20.Allen D.A        Eng  8.63  34   805 23.68   973   16025 16.47  1.44
```
This is a very interesting table. The additional qualification of Batting position average ensures that only genuine late order batsmen are compared. Mitchell Johnson has recently started batting at no.8. Hence his entry into this table. Soon he will go out of the table as he builds more innings at no.8 and possibly no.7.

Johnson is on top with a ratio of 2.15. The others are good quality late order batsmen. Anyone who has a ratio of greater than 1.4 should be classified as a top quality late order batsman.

If readers want different cut-offs for the tables, they are welcome to suggest the same.

Since the tables cover, with almost no exception, all the top batsmen of the world with variable career spans, I have given below the extreme peer average values in various classifications. The base table shows maximum spread, 10.7% on either side of 28.65, since it includes all batsmen, batting at 1-11. The Opening batsmen table has a spread of 7.4% on either side of 33.78. The Middle order table has a spread of 9.3% on either side of 32.69.

```Base table (All batsmen)
Low:  24.58 1890-1912 S.E.Gregory
High: 32.71 2005-1009 Mike Hussey

Opening batsmen
Low:  33.24 1950s C.C.Mcdonald
High: 38.47 1970s Fredericks

Middle order batsmen
Low:  32.16 1951-1961 Peter May
High: 39.34 2005-2009 Kevin Pietersen

Late order batsmen
Low:  14.03 1930s Verity
High: 16.47 1960s D Allen
```

Anantha Narayanan has written for ESPNcricinfo and CastrolCricket and worked with a number of companies on their cricket performance ratings-related systems