Anantha Narayanan

First 125 years: two gladiators, 22 yards, five days

A look at the most interesting duels between a batsman and a bowler in Test cricket

Hanif Mohammad pulls the ball to the boundary, Lord's, 1967

Hanif Mohammad's 970-minute epic in Bridgetown may well be the only instance of a batsman facing 1000 balls in an innings in Tests  •  PA Photos

In my last article, I had done a single-Test analysis of the head-to-head battles between bowler and batsman. This article covered about 550 matches from Test #1546 onwards where the complete ball-by-ball data is available.

The current article is a logical follow-up to that article and I will look at matches upto Test #1545. This covers 125 years of Test cricket. As such, I feel it is probably more important than the previous article. Some of the great confrontations ever took place during this long period.

Since no ball-by-ball data is available for these matches, this is really a pseudo-analysis of the head-to-head confrontations. However, the idea for this analysis originated from the previous one.

This analysis has been done using the single source available: the scorecard.

As astute readers would have guessed by now, the analysis is a tricky and involved one. Out of these 1545 Tests, we have data available for balls played by batsman for about 900 matches.

These represent interim levels of data. For the other 650 matches or so, we do not have the balls-played-by-batsman information. So I had to adopt different techniques to arrive at reasonable results.

For the 900 matches for which the balls-played-by-batsman information is available, I did a single extrapolation. Since I knew the balls played by each batsman, I needed to only allocate these balls between the bowlers who bowled in the innings based on the balls bowled by each of them. So this is a single extrapolation.

I had to be careful when I wanted to find out the runs scored off each bowler. This had to be a separate computation based on the runs conceded by the bowlers. This would ensure that the more accurate bowlers might bowl more balls at the batsmen but concede fewer runs.

I had to adopt a two-step algorithm for the 650 or so matches for which I had bare minimum information: only the runs scored. First I had to do an estimate of the balls played by each batsman by allocating the total innings balls based on a ratio of batsman runs and team runs.

Afterwards the allocation of balls and runs to the individual bowlers followed the method adopted earlier. These calculations are summarised below with couple of examples.

Two famous triple-centuries are used: Len Hutton's 364 in Test #266 for which the balls-played information is available (847) and Hanif Mohammad's 337 in Test #446 for which balls played information is not available, even though it is a later match.

I have avoided making any further assumptions such as batsman's batting position, bowler type, when the batsman started his innings and ended (anyhow this data is not available for these matches) et al.

"Keep it simple" has been my watchword.

Formula to determine the balls played by Batsman where not available
Batsman balls = Team balls * Batsman runs / (Sum of Batsman runs)
Hanif scored 337 out of 657 (Extras 33) in 319.0 overs.
    Balls faced by Hanif = (319x6) * 337 / (657-33) = 1034.

Common to both: Bowler allocation
Formula to determine the Batsman-Bowler balls Bat-Bow balls = Batsman balls * Bowler balls / (Sum of Bowler balls) Denis Atkinson to Hanif = 1034 * (62*6+0) / (319*6+0) = 201 (Atkinson's bowling analysis 62.0-35-61-1) Bill O'Reilly to Hutton = 847 * (85*6+0) / (335*6+2) = 215 (O'Reilly's bowling analysis 85.0-26-178-3)
Formula to determine the Batsman-Bowler runs Bat-Bow runs = Batsman runs * Bowler runs / (Sum of Bowler runs) Atkinson to Hanif = 337 * 61 / (657-33) = 33 (33 extras) O'Reilly to Hutton = 364 * 178 / (903-50) = 76 (50 extras)

Let us now move on to the tables.

Analysis for Innings: Single extrapolation (Batsman balldata available)
TestYearBatsmanRunsBallsBowlerExt B-B RunsExt B-B Balls
2661938L Hutton364847Fleetwood-Smith127220
2661938L Hutton364847WJ O'Reilly76215
8401979DW Randall150498JD Higgs69203
1931930A Sandham325640OC Scott104199
2661938L Hutton364847MG Waite64182
6181967G Boycott246555EAS Prasanna87179
2801946SG Barnes234667DVP Wright61177
2801946SG Barnes234667AV Bedser55177
15262000G Kirsten180461M Muralitharan56177
2561936WR Hammond231579FA Ward75172
1601925J Ryder201461R Kilner54169
2561936WR Hammond231579WJ O'Reilly49168
13741997ST Jayasuriya340578RK Chauhan105166
7321974DL Amiss262563AG Barrett58166
10901988DC Boon184431EE Hemmings63166
13741997RS Mahanama225561RK Chauhan69161
8991981JG Wright110434DR Doshi26161
11161989Javed Miandad271465SL Boock102160
6011966RM Cowper307589FJ Titmus50160

The featured table lists all situations where the bowler bowled more than 160 balls at a specific batsman. The downloadable table lists all instances with 120 balls as cut-off.

Hutton's 364 was a monumental innings. Wally Hammond was determined to grind Don Bradman and his team into the dust and batted for nearly three days. Hutton faced 847 balls - no extrapolation needed for this, and scored 364 runs. Bill O'Reilly and Chuck Fleetwood-Smith had contrasting spells. They had spells of 85-26-178-3 and 87-11-298-1 respectively. This wide disparity in runs conceded is reflected in the numbers. Hutton faced 215 balls against Fleetwood-Smith as against 203 against O'Reilly. However he scored many more runs off Fleetwood-Smith. This shows the advantage of separate extraction processes for balls and runs. This is the only instance of a batsman facing over 200 balls off two bowlers in an innings.

Derek Randall's presence in third position is a way-out situation since he scored only 150 runs. Coming in after a rare first-ball dismissal of Geoff Boycott, Randall batted for nearly ten hours and faced a whopping 498 balls. This, coupled with the fact that Jim Higgs bowled nearly 40% of the team overs, contributed to this unlikely combination, reaching to 199 balls.

An interesting presence in this table is that of Sanath Jayasuriya and Roshan Mahanama. In that mind-numbing "Test of patience" in 1997, Sri Lanka's mammoth 952 required 271 overs to be bowled. Rajesh Chauhan bowled most, with 78 overs, and Anil Kumble and Nilesh Kulkarni followed with 72 and 70 respectively. Chauhan features here twice, and the other two hapless bowlers feature in the longer list. Jayasuriya faced 166 balls off Chauhan and Mahanama, 161. These two faced 154, 149 balls off Kumble and 149 and 145 balls off Kulkarni. Ah! My hands get tired typing out these numbers. Paralysis of the mind could very well set in.

Analysis for Test: Single extrapolation (Batsman balldata available)
TestYearBatsmanRunsBallsBowlerExt B-B RunsExt B-B Balls
1791929WR Hammond296977CV Grimmett94265
1981930H Sutcliffe215565CV Grimmett79228
8991981JG Wright143628DR Doshi33223
2661938L Hutton364847Fleetwood-Smith127220
1931930A Sandham375702OC Scott129219
1591925H Sutcliffe303871JM Gregory86219
2661938L Hutton364847WJ O'Reilly76215
7381974G Boycott211725LR Gibbs65215
1791929WR Hammond296977RK Oxenham51210
1591925H Sutcliffe303871AA Mailey95207
15302001ME Trescothick179517M Muralitharan67206
1601925J Ryder289580R Kilner72205

The featured table covers Test confrontations of 200 balls and above.

Hammond scored 119 and 177 and played a total of 977 balls. A monumental effort indeed. Clarrie Grimmett bowled over 100 overs and Hammond played 265 balls of these. This was during Bradman's debut series.

Andy Sandham scored 375 runs in the nine-day Test in which there was no play on the last two days (yes, you read it right). It is amazing that the score stood at England: 849, West Indies: 286, and England batted again. Then they did not send the first-innings triple-centurion until the score was 176 for 5. Otherwise, Sandham might very well have upstaged Graham Gooch. Sandham faced 219 balls from Tommy Scott but more importantly scored 129 runs, the highest in all these matches.

Marcus Trescothick vs Muttiah Muralitharan just missed getting into the group of Tests with ball-by-ball data. It is noteworthy that Trescothick faced Muralitharan away for over 200 balls without giving his wicket once to Murali.

Analysis for Innings: Double extrapolation (No batsman ball data available)
TestYearBatsmanRunsExt Bat-BallsBowlerExt B-B RunsExt B-B Balls
4391957PBH May285811S Ramadhin94308
5641964KF Barrington256757TR Veivers67246
4501958GS Sobers365587Fazal Mahmood116241
3261950L Hutton202654AL Valentine73233
4391957PBH May285811DS Atkinson72226
5641964RB Simpson311749TW Cartwright58225
2261933WR Hammond336595FT Badcock80225
3711953FMM Worrell237523MH Mankad96208
6311968GT Dowling239567RG Nadkarni58201
4461958Hanif Mohammad3371034DS Atkinson33201

This table also features confrontations clocking at 200 balls and above.

First, let us not forget that this is a double extrapolation method. Peter May's 891 itself is a derived figure. This was a match in which May and Colin Cowdrey engineered one of the greatest comebacks in history. After two innings the scores stood at England: 186. West Indies: 474. Then England were tottering at 113 for 3. The amateur duo of May and Cowdrey came in and added 411 runs. Sonny Ramadhin bowled 98 overs in the innings which is a record even today. May faced 308 balls off Ramadhin, and that was some feat. It is the highest faced by a batsman off a bowler. Even allowing for a 10% variation, this is around 270 balls. Let us not forget that May, the captain, declared when he was 285, thus depriving himself of a triple-century. In the bargain West Indies scored 72 for 7 and narrowly escaped a defeat. Shades of New Zealand against India during 1965.

Ken Barrington's marathon effort of 256 runs translated to 757 balls and 246 balls off Tom Veivers. Not many runs were scored, though. Garry Sobers, in his record-breaking innings of 365, which transposed to only 587 balls, faced 241 balls off Fazal Mahmood. Hanif faced maximum balls from Denis Atkinson during his 16 hour epic of 337. This was extrapolated to 1034 balls, the only time in history of Test cricket that a batsman faced over 1000 or more balls in an innings, extrapolated or otherwise. There is a fair chance that this would be correct since Hanif was an eminently defensive batsman.

Analysis for Test: Double extrapolation (No batsman balldata)
TestYearBatsmanRunsExt bat-BallsBowlerExt B-B RunsExt B-B Balls
4391957PBH May315891S Ramadhin99340
3241950C Washbrook150679AL Valentine48264
3241950C Washbrook150679S Ramadhin57262
5211962Hanif Mohammad215733GAR Lock81258
3711953FMM Worrell260608MH Mankad104250
5641964KF Barrington256757TR Veivers67246
3771953GO Rabone175732HJ Tayfield47242
3261950L Hutton204663AL Valentine76241
4501958GS Sobers365587Fazal Mahmood116241
5641964RB Simpson315761TW Cartwright58227
2261933WR Hammond336595FT Badcock80225
6651969MG Burgess178628Intikhab Alam75224
4391957PBH May315891DS Atkinson73223
3371951EAB Rowan296836R Tattersall46223
3391951AJ Watkins177657MH Mankad36221
4461958Hanif Mohammad3541079OG Smith57221
3391951AJ Watkins177657SG Shinde83220
3371951EAB Rowan296836MJ Hilton93210
2891947B Mitchell309728R Howorth67203

The cut-off for this is also 200 balls.

May's figures were derived separately for each innings and added. Since he played a reasonable number of balls in the first innings, the total comes to 891 and May faced a mind-boggling tally of 340 balls off Ramadhin. This is by far the maximum number of balls faced by a batsman off a single bowler. Look at how far off this is from the Cyril Washbrook v Alf Valentine and Washbrook v Ramadhin numbers.

I am happy to see the presence of Allan Watkins' defensive classic during England's tour of India, with a virtual "B" side. Hutton, May, Cowdrey, Fred Trueman, Denis Compton, Trevor Bailey and Alec Bedser were missing. Watkins handled Vinoo Mankad very effectively. This is nostalgia for me since the first cricket book I ever read was the one on this tour.

Comparison of extrapolated values with actual values for selected combinations
Single innings
Test Year Batsman          Bowler          Ext Actual Diff (Act to Est)
1810 2006 DPMD Jayawardene N Boje          201  221  - 9.1%
1563 2001 JH Kallis        RW Price        188  189  -0.05%
2034 2012 Azhar Ali        MS Panesar      164  163  +0.07%
1641 2003 SP Fleming       M Muralitharan  160  185  -13.6%
1696 2004 BC Lara          GJ Batty        150  161  - 6.9%
Single Test
Test Year Batsman          Bowler          Ext Actual Diff
1641 2003 SP Fleming       M Muralitharan  249  265  - 6.1%
1572 2001 BC Lara          M Muralitharan  213  240  -11.3%
1952 2010 HM Amla          Harbhajan Singh 188  177  + 6.2%
1743 2005 Younis Khan      A Kumble        185  208  -11.1%
1562 2001 A Flower         CW Henderson    176  202  -12.9%

In this table I have compared the actual head-to-head values as determined in my previous analysis with values extrapolated using the single extrapolation method. The differences are of the order of 15% on either side. In general, the actual values tend to be higher than the extrapolated values.

Finally, a question might arise as to what degree of confidence I have on the extrapolated results. I would say around 80% for the single extrapolation and 70% for the double extrapolation method. In other words I would expect values to be around 10% either side for single extrapolation and 15% on either side for the double extrapolation. Just a gut feeling. That is all. I agree that this is an estimate. However, when we have figures like May scoring 285 runs out of a total of 547 in 258 overs, we are not going to be far off if we say, with 70% degree of confidence, that May would have faced 811 balls (1548*285/547).

Readers are requested to come up with suggestions on how the ball-by-ball data available for the recent 550 matches can be used. I have one idea, provided by Ashwin Krishnamurthy, which is to look at how players played during the period close to reaching landmarks, such as hundreds.

I have created a document file containing all the qualifying performances. There are about 380 Innings-level selections and 270 Test-level combinations, shown in four tables. To download/view this document, please CLICK HERE.

I will not comment on the South African tour happenings since ESPNcricinfo does not like outsiders using strong comments on these matters in their space, and my choice of words cannot be mild.

However, I cannot but pass a comment or two on the Indian bowling woes attributed to the new rules. The sardonic smile you see on the horizon is mine on hearing that the new rules are not helping MS Dhoni. How do the same rules that help the India team plunder 120 runs in the last ten overs become unfair when it comes to the India bowlers? Without the new rules, India might well have finished at 270, in Mohali. And how do the new rules bring down R Ashwin's skills, someone who has opened the bowling in all formats of the game regularly, and who is not a guileful Erapalli Prasanna, by a long mile. And Ravindra Jadeja's flat fastish slow left arm is not going to be affected a lot by a fresher ball. The plain truth is that, barring Mitchell Johnson (and Mohammad Shami at Ranchi), the bowling across the board has been awful. And that is the difference in the series scoreline.

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