Cricinfo - Blogs - It Figures - Okay, Bradman is at No.1... but who is last?

Let us leave that topic aside. I have always felt that the other end of batting table presents a fascinating possibility. Who is the worst batsman who ever carried a bat and walked in. Is it Chris Martin, is it one of the Indian spinners, is it a West Indian fast bowler or an unexpected batsman out side this lot? Without further ado, let us delve in.

The first is that the batsman (okay, I know I am stretching the point) has to have played 25 Test innings, which, for a tail-ender, represents nearly 20 Tests. The next is that the career batting average should be below 10.00. These twin criteria have enabled 70 tail-end batsmen to be selected.

Let me also mention that I would not do just a simple table based on, say, Batting Average. That is something which anyone could get using the excellent Cricinfo Statsguru. I will do a composite but not complex analysis of these 70 batsmen.

I have considered three measures for analysis. These are explained below.

1. Batting Average. This is the simplest and most acceptable of all batting measures. Readers can easily identify with this measure and it reflects the batting ability very realistically, notwithstanding the "not outs" conundrum. In this particular analysis even the "not outs" do not matter since most of these batsmen remain not out on quite a few occasions. This measure will carry a weight of 20 points.

2. Dismissed Zeroes. The emphasis here is on both the words. An innings which ends at 0 means that, barring a few exceptional circumstances, very little has been contributed and another batsman, almost always a better one, has been left in the limbo. I have determined the number of dismissed zeros and determined a frequency of innings in which this has occured. The lower this figure is, the worse the batsman is. This measure will carry a weight of 15 points.

3. Average partnership runs added. This is a useful measure since it tests another facet of the tail-end batsman's skills, which is the support he provides to the senior batsmen. Basically I have computed the number of runs added while the tail end batsman was at the crease, mostly at no.10 or no.11, and determined the measure of average partnership runs per innings. This measure will carry a weight of 15 points.

I have considered (and ignored) the batsman's highest score since that does not convey any additional information. I have also not considered the "Balls played" information since that is available only for about a third of Tests. And extrapolating based on team scoring rate will not work since these batsmen are likely to take a lot more balls to score the runs.

Let us take a look at tables, first the support table.

Cty Batsman             Ins  No   Runs Avge  HS Dis  Runs  Avge
                                                0s  Added  Bpa

Zim Mbangwa M            25   8    34  2.00   8   9   171  11.0  
Nzl Martin C.S           65  30    76  2.17  12  25   663  10.9  
Win King R.D             27   8    66  3.47  12   7   275  10.2  
Bng Manjural Islam(Sr)   33  11    81  3.68  21  10   351  10.6  
Ind Chandrasekhar B.S    80  39   167  4.07  22  23   760  10.9  
Ind Maninder Singh       38  12    99  3.81  15  11   396  10.8  
Ind Doshi D.R            38  10   129  4.61  20  14   384  10.9  
Aus Reid B.A             34  14    93  4.65  13   6   262  10.8  
Ind Nehra A              25  11    77  5.50  19  10   221  10.6  
Win Valentine A.L        51  21   141  4.70  14  12   502  10.9

Now the final table.

Cty Batsman            Batting Avge  Dis 0s Freq    Avge Ptship   Total
                            (20)          (15)           (15)       (50)
Zim Mbangwa M           4.00 (2.00)   2.08 ( 2.78)   5.13 ( 6.84)  11.21
Nzl Martin C.S          4.34 (2.17)   1.95 ( 2.60)   7.65 (10.20)  13.94
Win King R.D            6.95 (3.47)   2.89 ( 3.86)   7.64 (10.19)  17.48
Bng Manjural Islam(Sr)  7.36 (3.68)   2.48 ( 3.30)   7.98 (10.64)  17.82
Ind Chandrasekhar B.S   8.15 (4.07)   2.61 ( 3.48)   7.12 ( 9.50)  17.88
Ind Maninder Singh      7.62 (3.81)   2.59 ( 3.45)   7.82 (10.42)  18.02
Ind Doshi D.R           9.21 (4.61)   2.04 ( 2.71)   7.58 (10.11)  18.83
Aus Reid B.A            9.30 (4.65)   4.25 ( 5.67)   5.78 ( 7.71)  19.33
Ind Nehra A            11.00 (5.50)   1.88 ( 2.50)   6.63 ( 8.84)  19.50
Win Valentine A.L       9.40 (4.70)   3.19 ( 4.25)   7.38 ( 9.84)  19.97

As foreseen, a dark horse has emerged. Who would have thought of a batsman who could come ahead of Chris Martin. (Mpumelelo) Pommy Mbangwa's batting is for the Gods to view. 25 innings, 8 not outs and 34 runs gives him an unbelievable average of 2.00. He has been dismissed at 0 for nearly 40% of his crease visits. He has a highest score of 8, the only one in this elite group not to have crossed 9 runs. He has always batted at no.11. His average partnership is an unbelievably low 6.8. What more do you want. I would have paid money to see Pommy bat. Note his batting sequence: 0, 2, 0, 4, 0*, 0, 0, 0, 0, 2*, 3, 2, 0, 1*, 2, 0*, 0*, 1*, 3, 0, 0, 1*, 8, 0*, 5. One fascinating string of scores.

I can see the New Zealand readers having mixed feelings. They would dearly love to have Chris Martin head this table because they love his batting. I can only suggest that if you increase the number of innings to 30, Chris Martin will be at the top. Let us see Martin's exploits. 65 innings, 30 not outs, 76 runs giving Martin a slightly higher average of 2.17 as compared to Pommy. He has crossed single figures once in his career, an unbeaten 12 against Bangladesh when he outscored O'Brien. He has 25 dismissed zeroes, the most frequent amonst all these batsmen. But his partnership average is a healthy 10+. Only twice has Martin batted at no.10 when Shane Bond and Cummings could not bat. Let me add, I would also pay money to see Chris Martin bat.

Reon King is next. Not as great a fast bowler as some of the other greats such as Walsh or Ambrose, but equally inept a batsman.

Then comes Manjual Islam, followed by three Indian spinners. Reid of Australia separates these three from Ashish Nehra, another rabbit of a batsman. Alf Valentine is last in this table.

Fidel Edwards, who is 11th in the table is the only other batsman wiith a sub-5.00 batting average. However he has recently batted very well, saving West Indies twice at Antigua and Napier.

To view the complete list please click here.

Ananth, Bradman being best is far from the truth. It is a myth propagated by Western media. Look at your own detailed analysis of all time best Test bowlers - not a single one of them ever bowled a ball to Bradman. Where in fact, five of them - Ian Bishop, McGrath, Donald, Pollock have bowled to Tendulkar (not counting Waqar and Akram). Add to that poor fielding and no third umpires (so benefit of doubt always went to the batsmen). You put together all these factors - lack of best bowlers, poor fielding and umpiring, how can you say that Bradman is the best? Thanks.
[[
I thought of not publishing this comment since it has no relevance to the article. Then I decided that this has been prompted by my first para and since I opened the gate I have to let the comment in.
I disagree completely with Prasad. There is no western myth to say that Bradman was the best batsman who ever played. One day I will do a complete analysis of Test batsmen to dispel this myth about myth.
This topic, while responding to this specific article, is closed. No comments on this track will be published. Please reserve your comments for the subject matter at hand.
Ananth:
]]

Good article. But I think it should have been about "batsmen," and not about the poor souls who are forced to be at the receiving end at the pitch! Clearly, these bowlers are picked for their prowess with the ball, and if they contribute anything with a bat, it's a bonus.

What about the specialist batsmen who repeatedly fail to perform? Consider Mike Brearley - isn't he the only player to be picked exclusively as a captain (Leeds, 1981)?

For this list, we should disregard the batting average, since a high unbeaten score can skew it higher. We could use median score per inning. Another yardstick: number (or ratio) of scores under 20 - clearly, if a batsman does not reach 20, he has failed.

Another factor: % of runs scored by the batsman while he's at at the crease, or the inning total. (Low = bad)

Strength of opposition may be a good factor, but is impossible to be quantified with a scientific (or mathematical!) wrapper.

This should have some interesting revelations :)

Good stuff - but I have a couple of points.

A weighting of 15 for the partnership element seems a little high - for the real rabbits, the length of the partnership is more likely to reflect the ability of their partners to farm the strike, rather than telling us anything about the ability of no.11 to stand his ground.

And I'm with Siva from Singapore in not understanding your method of scoring. Please can you explain how multiplying one of the factors by 2, and the others by 0.75, squares with your weighting of these factors by 20, 15 and 15? Thanks.
[[
Could as well have been 10 for partnership.
Continuing with the points related explanation.
1. Average ranges from 2.00 to 10.00. Hence multiplying this measure by 2 gives the required range of 4 to 20 with 4 being the "top" in this analysis.
2. The Zeros frequency runs from 2.60 to 17.00. Multiplying by 0.75 fits this into the required range of 1.7 to around 13, with 1.7 being the "top".
3. Similarly the average partnership value runs from 6.84 to 18.64. Multiplying this by 0.75 gives a range of around 5 to 14, perfect, with 5 being the "top"..
Ananth:
]]