Stats Analysis

Which players have had the worst batting and bowling slumps in Tests?

A look at batters who scored the fewest runs per Test and bowlers who took two wickets or fewer per Test, over ten Tests or more

Steve Waugh and Mark Taylor celebrate the series win, Pakistan v Australia, third Test, Karachi, day five, October 26, 1998

Steve Waugh and Mark Taylor had a stretch of ten Tests apiece where they averaged under 26 runs per match  •  Ben Radfor/Getty Images

The raging question in the minds of many Indian cricket followers is the continued poor form of Ajinkya Rahane, and, to a lesser extent, Cheteshwar Pujara. There are many who feel, justifiably, that these two players, especially Rahane, have been given long ropes and that one of the reasons for India's Test series loss in South Africa was the bull-headedness of the Indian team management in extending their support of them. It is quite possible that if the middle order had been manned by the likes of Hanuma Vihari and Shreyas Iyer, India might have won the Test series.
This article looks at the troughs of poor form in players' careers. Having said that, I will make clear here that the article does not look to find reasons for why players are not selected for any given series and try to draw conclusions. There is no verifiable data on such off-field happenings. And I do not work on anything that is picked up by hearsay or from the press.
When we talk of Rahane, everyone connected with Indian cricket will know that he was dropped for the first two Tests of India's tour of South Africa in 2018. That is a verifiable fact. However, I cannot use that in my analysis for this article, since I would not know what to do for such breaks for other players.
Faf du Plessis played ten Tests out of South Africa's 11 in 2016. He missed a Test against England in Centurion. Was he injured, dropped, or did he not play for his own reasons? Jonty Rhodes featured in ten Tests in a sequence of 18 played by South Africa. He played in the first eight Tests, the 14th and the 18th. He went on the tours to India and Pakistan. Why did he miss Tests there? Why did he miss the home Tests versus India? Was he injured? Was he not selected? These are almost totally subjective and unverifiable causes. To complicate matters, these days increasingly players are normally never "dropped" but "rested" or "rotated".
The bottom line is that I can only work with 100% verifiable data - and only player data, not team data.
After much deliberation, I have decided on the following cut-offs and decisions:
- I will consider batters and bowlers, excluding true allrounders. They are likely to perform well in either discipline in any given Test and very rarely fail in both.
- When I consider batters, I will exclude wicketkeepers. They do a very tough job and I endorse the older and more correct idea that their batting is secondary, which is how it should be. Nowadays the better wicketkeeper is often sacrificed at the altars of batting and team balance.
- For batters, the cut-off is 2500 Tests runs and an average of at least 25. Also, all players with 100-plus Test wickets are excluded from the reckoning as batters, which means players like Carl Hooper and Ravi Shastri do not qualify. For bowlers, the cut-off is 85 wickets. On average this requires around 35-40 Tests and I need that many to have a good sample size for analysis.
- I will look at poor performances over a minimum run of ten Tests. I will feature only the worst streak of any particular player's career.
- The key measure will be the Runs per Test (RpT) for batters and Wickets per Test (WpT) for bowlers. The key values are Below 30.0 for RpT and Below 2.0 for WpT values. I have considered and ignored batting average, since I do not want any batter to escape the noose because he remained unbeaten a few times. The weighted batting average (WBA) was an alternative but on balance, runs scored per Test works well. I think everyone will agree that if a batter does not compile 300 runs in ten Tests, that stretch should be considered as a gross failure.
- The streaks were determined at the level of each Test, for matches that the player played. That meant evaluating over 400,000 streaks for batters and bowlers.
I will present the data in the following five forms for batters and bowlers. For the first four tables, I will determine the worst trough for each player and include that for consideration.
1. Consecutive number of Tests of poor performances
2. Low RpTs / WpTs
3. Percentage of really poor performances (single-digit dismissals for batters, and wicketless innings for bowlers)
4. Low Performance Index (a composite of how far below the acceptable performance levels a player was and for how long)
5. Players who have multiple such troughs in their careers. For this analysis, the cut-off for batters is increased by 10% so as to have a wider group of players.
A total of 39 batters qualified for final consideration, having gone below the cut-off at least once. Each batter might have multiple poor streaks but only the worst of them is featured. Saleem Malik has 57 such streaks.
What do we see on top here? A stretch of 18 Tests in which the batter scored around 500 runs. That is a poor run indeed by Alistair Campbell. I can understand that Rhodes kept his place in the side because his fielding partly made up for when he failed with the bat. But Pakistan must really have been short of talent for Malik to have kept his place for so many matches. However, it is true that these are not necessarily unbroken series of Tests for the teams. The concerned players might have been dropped in the middle. Still, it is also true that when they came back in such instances, their poor form continued. And let me say that when I increase the cut-off to 32.5 runs per Test, Graeme Hick has a sequence of no fewer than 27 Tests.
In this list of 15 batters, there are no fewer than six from England. At the risk of generalising, let us say that this could be an indication of English selectors having been generous and giving batters long ropes. The fact that most of these batters were from the period circa 1970 through 2000 suggests potential good replacements were scarce.
Now we come to the table of low numbers of runs being scored over a minimum of ten Tests. Three of the top five positions are held by England batters of the 1980s and '90s. Nasser Hussain scored below 200 runs in ten Tests. Hick and Alan Lamb managed to cross 200 runs - in Hick's case, only just. Roshan Mahanama and Campbell come in at Nos. 2 and 3. Malik is at No. 7, followed by Sanath Jayasuriya, who had a tough time at the beginning of his career. Kraigg Brathwaite is a recent player who has had a run of really poor form. Most of us know that the first third of Steve Waugh's career was a rocky one; in the case of Mark Taylor, the bad patch came near the end of the career. All these players were at or below the 25-runs-per-Test mark. du Plessis had a poor run in his early years. Zaheer Abbas had lapses in form often, interspersed with some great batting streaks.
The table above analyses the failure percentage within a streak. These failures are determined only after a streak qualifies - that is, when fewer than 300 runs were scored over a ten-Test stretch. The perfect example of a player having a run of atrocious scores but still not making it to this table is that of Mohinder Amarnath. He had possibly the worst such run in Test history, scoring 4, 7, 0, 0, 1, 0, 0, and 0 in eight innings, totalling 12 runs. However, this stretch does not feature in this analysis because Amarnath scored 405 runs in the ten Tests of which the scores above came in the first five Tests, and 610 runs in the ten Tests ending with these five.
Hick's career was a turbulent one. He played 65 Tests but his batting average never crossed 40. Often, one good innings was followed by multiple failures. In one of his poor streaks, he amazingly failed in over 76.5% of innings he played. His sequence of scores was 2, 36, 8, 1, 6, 20, 4, 4, 6, 2, 1, 107, 0, 0, 68, 8, and 0. That is 13 failures in 17 innings. John Reid of New Zealand had a run that was nearly as poor - 50, 11, 11, 0, 3, 6, 1, 9, 7, 6, 0, 3, 1, 135, 0, and 26. That works out to 11 failures in 16 innings. Keith Fletcher's miserable run was 22, 0, 23, 8, 80, 5, 33, 20, 4, 2, 1, 28*, 1, 0, 5, 2, 0, 16, and 5. That is a sorry 12 failures in 19 innings. A veritable galaxy of top players feature on this table - Mark Taylor, Steve Waugh, VVS Laxman and Inzamam-ul-Haq among them.
The table above is based on a metric I created, called Poor Performance Index (PPI). That is a simple product of number of Tests multiplied by 40.0 minus RpT. The larger each of these two values are - indicating a long streak and a low RpT - the bigger the PPI. Campbell tops the table with a PPI of 275 (16 Tests * 17.2 RpT difference). Hussain comes next with 236 (12 * 19.7), followed by Malik and Fletcher. Allan Lamb and Hick follow next, making this a table dominated by England batters again, with four out of the top eight placements. At two extremes, Hick had a run of ten Tests but a huge RpT differential of close to 20, while Rhodes had a run of 16 Tests and a lower RpT differential.
This last batting table lists batters who had multiple separate poor streaks in their career. There are no overlaps among these streaks. In this table, I have provided additional information such as career Tests (C-Tests) and the numbers of the Tests within the batters' careers, so the readers have an idea about the stage of the player's career when the streak occurred. The RpT cut-off has been increased to 33.0 only for this analysis.
Jayasuriya is the only batter in history to have had three such streaks, though in an admittedly long career. They came at the beginning, middle and end of his career. A total of 14 other batters had two such poor streaks in their careers. Of special interest will be Zaheer, Steve Waugh, Mike Atherton, Hick and Rahane. Marvan Atapattu had two such streaks in the first 26 Tests he played. Then he flourished. Rahane is something of an aberration in that his two streaks of poor performances have come not at the beginning of his career, but at a stage when a player is expected to do well. Contrast this with the early career failures of Waugh, Malik and Atapattu. Rahane's last streak ends with the Kanpur Test last year, since he scored 136 runs in the three Tests in South Africa.
In closing, it seems clear that the longest of all ropes were extended to the English batters of the 1980s and 1990s. It is possible that injuries were also involved, and players not being selected. However, it is clear that as and when they came back, they continued to fail. If they came and performed better, they would not have ended up on these tables.
Now to the bowlers. The cut-off for the bowlers is a rather steep 2.0 WpT. Thirty-four bowlers qualify. Readers may justifiably ask why I have set the bowling cut-off this high, considering a wicket is normally weighed as being equal to about 25 runs in Tests. The batting cut-off is at a lowly 30.0. The reasons are explained below.
1. Normally only four or five bowlers take all the wickets that fall in a match - an average of 15-16 per Test. So, if for the purposes of argument all bowlers in a side are performing about equally well, they are liable to take a little over three wickets a game.
2. In many cases, the batting team has either one innings or one and a bit to bat in matches they win. However, to win a match, teams normally need to take all 20 wickets.
3. In a score of 500 for 3, a No. 5-6 batter might not even get a look in. It is very rare that a bowler does not get to bowl. Yes, Nathan Lyon did not bowl a ball at the MCG recently. But that is a rare exception.
Hence 2.0 seems reasonable. Like with the batters, it must be understood that only the worst streak a bowler had is considered here. These bowlers will possibly have had many other poor streaks, overlapping or not, in their careers. For instance, Pramodya Wickramasinghe had 219 such streaks, albeit many of them overlapping. Ishant Sharma had 63.
Wickramasinghe had a chequered career. What is most amazing is not that he took only 85 wickets in 40 Tests but that he played that many matches. It is a mark of how short Sri Lanka were of pace-bowling talent in the 1980s and '90s. Wickramasinghe had a 26-Test streak in which he took only 39 wickets. A six-wicket haul broke that streak. Shannon Gabriel was similarly lacklustre in the first half of the 2010s, running a 21-Test streak with under 2.0 wickets per Test. It is amazing that Maninder Singh had a 20-Test streak with little to show. However, it must be remembered that these streaks do not account for matches the player missed on account of being dropped, injured, or for other reasons.
Ishant, Shivlal Yadav, Venkatesh Prasad and S Venkataraghavan form an Indian quartet of bowlers who had streaks of over 17 Tests with low productivity. It is understandable why Venkataraghavan had such lows. He was often the third bowler in India's spin troika. The two bowlers who surprise me by their presence this list are Derek Underwood and Makhaya Ntini - both very effective bowlers overall.
Wickramasinghe and Maninder had streaks of ten Tests in which they took one wicket per Test. Ntini and Ishant did only slightly better. Kemar Roach and Suranga Lakmal are surprises here. And Craig McDermott's lean period of 1.5 WpT, is a surprise considering Australia's tough selection policies. There seems to be a wide representation of all the teams in the bowler tables.
For bowlers, I have defined a failure as an innings in which they failed to take a wicket. Yes, I can anticipate some questions on that. Classifying a three-over wicketless innings as a failure could be considered unfair to the bowler because of the extremely short bowling effort. However, that could be compensated by a similar classification of a 30-over wicketless innings.
Yadav had 11 failures out of 18 - a failure rate of over 60%. Phil Edmonds and Lakmal follow next. It can be seen that these bowling lists have a high proportion of spinner. And a recurring theme is that the teams these bowlers belonged to were not really strong teams; the average number of wickets taken by these bowlers' teams per match might only have been around ten or 12.
The Poor Performance Index table seems to follow along the lines of the first two tables. The index is determining by multiplying the Tests played by 3.0 minus the WpT. Wickramasinghe is the leader here too, with a PPI value of 33.0 (26 Tests * 1.27 WpT differential). Maninder follows next with a PPI of 23.0 (13 * 1.77). Note how Ntini comes in though he had a streak of only ten Tests, because he fell way behind on the WpT value - a low 1.2. It is important to mention that Ntini had a wonderful career overall, taking 390 wickets in 101 Tests - a WpT value of 3.86.
Unlike batters, very few bowlers go through troughs multiple times in their careers. While we had 15 batters with two or more poor streaks, we only have seven bowlers who had two lows apiece in their career. It is understandable why Venkataraghavan had two such troughs: his opportunities were limited, and in any case, he was also a limited bowler. Underwood was a surprise here too. He was England's leading spinner in his time. It is quite possible that the pitches back then did not help spinners much. Such might also have been the case with Edmonds - though unlike Underwood, he was a limited bowler. Prasad was not all that effective on Indian pitches, and it was the same with Ishant in the early part of his career. Fernando had to bowl on the unhelpful pitches in Sri Lanka. It can be seen that most of these bowlers had poor finishes to their careers.
In summary, the bowling tables are largely dominated by roughly the same set of bowlers. Like England in batting, India seem to be a significant presence in these bowling tables, giving plenty of rope to bowlers struggling to deliver.
Conclusion
Going back to one of the points raised in the preamble, it is clear that Rahane is going through a poor extended phase and it is costing India. It is a matter of concern that in a career of 82 Tests, he has had two awful streaks of ten Tests each. And the final damning number is that Rahane averages 60 runs per Test - way below what is expected of a middle order batter. When a team's results are strongly influenced by the poor form of one player, it is time a firm decision is taken.
Coming to Pujara, it is quite different. His poorest streak was the recent one in which he scored 378 runs. Not great but reasonable. I would say. Maybe he deserves a home series.
Anticipating the queries of many readers, I have listed below the poorest ten-Test streaks of a few top batters and bowlers. The complete data file is available for download here.
Batters: Runs
Don Bradman: 997 (a miss, a la 99.94)
Brian Lara: 479
Sachin Tendulkar: 337
Kevin Pietersen: 503
Ricky Ponting: 288 Garry Sobers: 419
Sunil Gavaskar: 446
Viv Richards: 391
Jack Hobbs: 675
Steven Smith: 579
Seymour Nurse: 649
Joe Root: 548
Bowlers: Wickets
Sydney Barnes: 48
M Muralidaran: 29
Shane Warne: 22
Dale Steyn: 25
Glenn McGrath: 30
Ravi Ashwin: 35
Saqlain Mushtaq: 38
Richard Hadlee: 36
Dennis Lillee: 26
Bill O'Reilly: 47
Kapil Dev: 16
Andrew Flintoff: 7
Talking Cricket Group
Any reader who wishes to join the general-purpose cricket ideas-exchange group of this name that I started last year can email me a request for inclusion, giving their name, place of residence, and what they do.
Email me your comments and I will respond. This email id is to be used only for sending in comments. Please note that readers whose emails are derogatory to the author or any player will be permanently blocked from sending in any feedback in future.

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